Berlin Workshop Series 2006 35918 Equity and Development Edited by Gudrun Kochendörfer-Lucius and Boris Pleskovic Equity and Development Berlin Workshop Series 2006 Equity and Development Edited by Gudrun Kochendörfer-Lucius and Boris Pleskovic THE WORLD BANK Washington, D.C. © 2006 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. 1 2 3 4 09 08 07 06 The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work with- out permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete infor- mation to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street, NW, Washington, DC 20433, USA, fax 202-522- 2422, e-mail pubrights@worldbank.org. ISBN-13: 978-0-8213-6105-4 eISBN-10: 0-8213-6106-6 ISBN-10: 0-8213-6105-8 DOI: 10.1596/978-0-8213-6105-4 ISSN 1813-9442 Edited by Gudrun Kochendörfer-Lucius and Boris Pleskovic Contents ABOUT THIS BOOK VII INTRODUCTION Gudrun Kochendörfer-Lucius and Boris Pleskovic 1 OPENING ADDRESS Michael Hofman 7 OPENING ADDRESS Inequality of Opportunity and Economic Development Francisco Ferreira and Michael Walton 11 KEYNOTE ADDRESS Distribution, Equity, and Development François Bourguignon 29 What Is Equity? What Is the Role for Governments in the Promotion of Equity? Does the Role Differ between Developed and Developing Countries? Promoting Equity through Social Cohesion: The Council of Europe Strategy Gabriella Battaini-Dragoni 39 Equality of Opportunity and Other Equity Principles in the Context of Developing Countries Dennis Cogneau 53 What Is Equity? Stephan Klasen 69 Should Poor People Care about Inequality? Martin Ravallion 79 V VI | CONTENTS Equity-Enhancing Social Transformation: Historical Evidence from Europe and Transition Countries Inequality and Political Regimes Carles Boix 95 Income Inequality, Economic Transformation, and Growth in Russia Evgeny Gavrilenkov 113 Justice and Transition Vladimir Gligorov 127 Building Efficient Welfare States: Lessons Learned Welfare State, Redistribution, Growth, and Poverty Reduction Rolph van der Hoeven 149 Global Inequalities: What Can Be Done to Reduce International Inequities? Rights-Based Development Jeremy Hobbs 165 Rising Labor Force Participation as a Source of Pro-Poor Growth Andrew Warner 173 What Will Greater Integration Mean for Inequalities between and within the Richer and Poorer Countries of the "New Europe"? Social Solidarity in an Expanding Europe Kemal Dervis¸ 191 Integration and Inequality in the EU Accession Countries Jan Svejnar 197 APPENDIXES Appendix 1: Programme 203 Appendix 2: Participants 209 About this Book The World Bank and InWEnt (Capacity Building International, Germany) hold a Development Policy Forum each September in Berlin. This meeting, known as the "Berlin Workshop," provides a forum for the European research community to con- tribute its perspectives to early discussions in preparation of the World Bank's annual World Development Report. The Workshop offers new ideas and distinctive per- spectives from outside the World Bank. Participants in the Workshop come from a range of academic, governmental, think­tank, and policymaking institutions in Europe, the United States, and the Russian Federation, as well as from the World Bank and the German development institutions. Conference papers are written by the participants and are reviewed by the editors. Participants' affiliations identified in this volume are as of the time of the conference, September 6­8, 2004. The planning and organization for the Workshop involved a joint effort. We extend our special thanks for the support of François Bourguignon, senior vice president for development economics and chief economist of the World Bank, and Michael Wal- ton and Francisco Ferreira, codirectors of the World Bank's World Development Report 2006. We wish to thank Aehyung Kim, Klaus Kruger, and Joachim Müller for their advice and suggestions. We would also like to thank conference coordinators Theresa Bampoe, Marianne Donda, Irene Federwisch, and Judith Klemmer whose excellent organizational skills kept the Workshop on track. Finally, we thank the edi- torial staff, especially Stuart Tucker and Mark Ingebretsen from the Office of the Publisher, Kim Kelley, and Grit Schmalisch, for all of their work on this volume. VII Introduction GUDRUN KOCHENDÖRFER-LUCIUS AND BORIS PLESKOVIC The articles in this volume were presented at the seventh annual Berlin Workshop, held in Berlin on September 6­9, 2004. The workshop was sponsored by InWEnt (Capacity Building International), Germany, and the World Bank. The workshop provides a forum for the European research community to con- tribute its perspectives to early discussions in preparation of the World Bank's annual World Development Report (WDR). The Workshop offers new ideas and perspec- tives from outside the World Bank giving fresh and enriching impulses to the devel- opment of the WDR. Participants of the 2004 workshop came from a range of aca- demic, governmental, think­tank, and policymaking institutions in Europe, the United States, and the Russian Federation, as well as from the World Bank and the German development institutions. Michael Hofmann in his opening address emphasizes the importance of equity for poverty alleviation and sustainable development in developing countries. He also notes that equity is vital for European societies, especially those in the process of tran- sition or enlargement. He argues for strengthening research on equity issues to explore different cultural endowments and their implications for addressing inequality by quoting David S. Landes in The Wealth and Poverty of Nations: "If we learn anything from the history of economic development, it is that culture makes all the difference." Hofmann supports country-specific policies and welcomes the recent, more pluralis- tic approach of the World Bank. At the same time, he reiterates that development and equity warrant more than technocratic fine-tuning. Structural change with equity implies political interest and the political will to serve people in need. In their opening address Francisco H.G. Ferreira and Michael Walton introduce the topic of World Development Report 2006 (WDR): the interrelationship between equity and development and the respective development strategy as well as Gudrun Kochendörfer-Lucius is Managing Director of InWEnt­Capacity Building International, Germany. Boris Pleskovic is Research Manager, Development Economics, at the World Bank. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 1 2 | GUDRUN KOCHENDÖRFER-LUCIUS AND BORIS PLESKOVIC the consequences of inequality on a country and global level. The workshop was divided into five sessions following a keynote address. In his keynote address, François Bourguignon examines the conceptual foundation of the workshop sessions by discussing the definition of equity itself. What do we mean by equity and how does equity differ from equality? Whereas equity is commonly asso- ciated positively with impartiality and justice, economists understand equality as an idealistic and unattainable goal often linked to socialism and communism. Interestingly enough, the connotation is different with respect to the negation of the terms (inequity and inequality): Less inequities but rather inequalities become political and economic focal points, and overcoming inequalities are taken as condition for equity. Thus, the terms are sometimes used simultaneously and interchangeably, with equity as the goal and inequalities as the impediment to be overcome. The terminological twins equity/equality, however, can be conceptualized in highly diverging ways with differ- ent consequences for development strategy. The discussions throughout the workshop mirror the controversial positions of international discourse on the topic. Equity and equality can point to varying dimensions; for example, with respect to outcome, opportunities, and capabilities. Each of these dimensions entails different responsibilities for political action. Whereas equality in outcome implies an egalitar- ian perspective, economic studies on inequality in outcome mostly take into account the results of actions and conditions such as unequal incomes. Equality of opportu- nities then includes the bases of actions: do we all start from the same level? In con- trast, the concept of equality in capabilities takes us even one step further. Even when we start from the same level, we may still lack the capabilities to effectively seize the opportunities offered in this equal starting position. Equality in capabilities requires us to focus on the structural and procedural bases of agency, such as the functions and capabilities required to empower everyone to really take advantage of the (equal) opportunities a society offers. Session I, on what is equity? What is the role for governments in the promotion of equity? Does this role differ between developed and developing countries? Gabriela Battaini-Dragoni embeds the concept of equity within a broader framework and strategy for social cohesion. Family, community, and civil society are portrayed as essential promoters of equity; thus, the author proposes a redefinition of the clas- sical role of the state as the only responsible actor for ensuring welfare. Public agen- cies should therefore follow a partnership approach and cooperate with social part- ners, nongovernmental organizations, and the private sector. The second paper by Dennis Cogneau demonstrates that no single concept of equality serves as best solu- tion for all aspects of life. Equity or equality in real life may require a complemen- tary mix of equality of opportunity; of meritocratic, efficient allocation; and of equal- isation of primary functionings. Dennis Cogneau suggests that each of these dimensions may best serve different domains of social life. Moreover, he stresses that these three dimensions imply different mechanisms that should reinforce each other. Stephan Klasen in the third paper of the session provides an in-depth analysis of the different dimensions of equity and equality by referring to philosophical arguments presented by Nozick, Rawls and Sen. He concludes that before operationalizing and INTRODUCTION | 3 measuring inequity, the concept itself has to be clarified. Klasen argues in favour of equality--specifically Amartya Sen's "equality in capabilities"--as the goal for the WDR. Inequalities in functionings, procedures and means can then be measured as helpful proxies for approaching the concept. The last paper by Martin Ravallion elaborates on the significant role of inequality in poor countries. Even though we might agree on the premise that takes absolute poverty reduction as the overriding goal for development policies in poor countries over reducing inequality per se, Ravallion questions, however, that the crucial issue in reducing absolute income poverty is the economic growth rate alone. Equity and growth do not necessarily rep- resent a trade-off in general (even though they might so on the policy level), nor should redistributive policies come at the expense of lower long-term living standards for poor people. Rather the potential for "win-win" policies correlates with the fact that some of the factors that impede growth also imply that the poor share less in the opportunities unleashed by growth. Therefore, Ravallion argues that one future chal- lenge for development policy is precisely to combine growth-promoting policies with policies that assure that the poor can fully participate in these opportunities that growth offers. Session II, on equity-enhancing social transformation and historical evidence from European and Transition Countries, focus on policies that impact equity. Looking at Europe from a historical perspective shows that changes in the degree of equity in a society are possible. This provides an encouraging example for countries in transi- tion. What specifically can we learn from this historical experience? Theories of political regimes, as presented by Carles Boix in the first paper, help to understand the occurrence of democracies and authoritarian regimes by highlighting the distri- bution of economic assets and the political balance of power among different social groups as key factors. Moreover, Boix points out that economic development fosters democracies by equalizing conditions and reducing the weight of immobile assets. However, Evgeny Gavrilenkov argues that many studies prove that most economies in transition experienced a rapid increase in inequality, especially in income inequal- ity. This was the case in Russia, for which he presents detailed data on the current economic situation including monetization and consumption. Gavrilenkov recom- mends that an expected rise in pensions should be accompanied by an active restruc- turing of the state in the sectors administration, national security, legal system, and social policies. The last paper by Vladimir Gligorov deals with the issue of justice in transition countries. Even though these countries experienced an increase in inequal- ity and show a significant scale of disadvantaging redistribution of resources, open social conflicts are not characteristic for them. He argues that this conundrum can be explained by the dominance of allocative over distributive justice. With respect to the future integration into the European Union, however, the institutionalization of a system of social justice in these transition countries will become a crucial issue. Session III, on building efficient welfare states and lessons learnt, discusses the task of formulating policies that foster both efficiency and equitable social welfare. Does the relationship between efficiency and equity necessarily have to be a zero-sum game? European welfare states have plenty of experience in tackling this question. 4 | GUDRUN KOCHENDÖRFER-LUCIUS AND BORIS PLESKOVIC Most of them are currently undergoing major reforms of their welfare systems to increase the efficiency while fostering competitive economy. The session thus asks if some of the lessons learnt by the European countries can be of value to the develop- ing world. The central question of the presented paper by Rolph van der Hoeven is whether the current system of welfare states and redistribution in industrialized countries is applicable in developing countries. Could it contribute to faster poverty reduction in these countries? Whereas the economic literature used to conceptualize equality and growth as a trade-off-relationship, recent studies including van der Hoeven's paper challenge this restrictive view and seek to combine redistributive policies with poverty reduction. He analyses different redistributive instruments with respect to their effects on growth and efficiency. The results indicate that greater dis- tributional equality provides for a favourable initial condition for sustainable and rapid growth. Moreover, redistribution of current income and assets or of an econ- omy's growth increment seems to be the most effective form of poverty reduction. Thus, van der Hoeven recommends a new agenda for growth and equity which should imply the integration of both distributional and poverty reducing policies with a priority for the latter. Session IV, on international inequalities and what can be done to reduce them, focuses on the global level, whereas Session III concentrates on equity issues at the national level. Jeremy Hobbs presents the Oxfam perspective on this issue pledging for a rights-based development approach. This entails a broader concept of equity includ- ing, besides income distribution, the provision of health and education as well as land, cultural, minority, and democratic rights. Whereas the Doha Round and the Millen- nium Development Goals provided some useful opportunities to address inequity and poverty, Hobbs points out that it will be a challenge for the World Bank, political lead- ers, and civil society to raise public awareness and build sufficient momentum for implementing the necessary measures. With respect to the domain of trade, one neces- sary measure would be the firm agreement to end agricultural export subsidies and export credits. With respect to aid and the Millennium Development Goals (MDG), all DAC countries (member countries of OECD's Development Assistance Committee) and the EU should set timetabled targets to reach 0.7 percent of their GDP in foreign aid by 2013. Regarding debt, a shift of focus on debt relief should be sought. This should include the proposition that any calculation of debt sustainability has to be linked to the poverty needs of a country, in particular the financing gaps it is facing in reaching the MDG. Andrew M. Warner, author of the second paper, stresses that growth does not invariably reduce poverty. He studies the channels through which eco- nomic growth effectively reduces poverty and impacts income distribution. As com- pared to other possible channels of poverty reduction, such as a rise in the average wage within sectors, employment shifts from low- to high-wage sectors or a rise in profits per employed person, research results and cross-country data prove that rising rates of employment represent a significant and highly effective way by which growth reduces poverty and income inequality. Therefore, raising labor force participation rep- resents an effective and most promising source of pro-poor growth. INTRODUCTION | 5 Session V, on what will greater integration mean for inequalities between and within the richer and poorer countries of the New Europe, draws a very differenti- ated picture. Starting with an intercountries analysis of the EU, the first paper by Kemal Dervis¸ provides insights from Turkey as one of the EU potential candidates. Studies show that inequalities with respect to per capita income will increase between the newly accessing countries and the older Europe. However, looking at the produc- tivity per worker versus per capita income, the demographic structures of the access- ing countries in Eastern Europe vary, presenting Turkey among the top if compared to these countries. It is also suggested that Turkey as a potential future candidate for joining the EU could offer the ageing population of Europe a demographic bonus and help address the pension problem in the EU. Jan Svenjar in the second paper of the session analyses the impact of EU integra- tion on the economy and on inequality in the accession countries from Central Europe and the Baltics. He discusses the negative development with respect to income and consumption inequality. For example, the overrepresented groups among the long-term unemployed are the same as in Western European countries: mainly unskilled workers, young and older people, minorities and handicapped peo- ple. Women disproportionately dropped out of the labour market. In contrast to Russia and the other CIS countries (Commonwealth of Independent States), how- ever, the evolution of income and consumption inequality in the Central European and the Baltic states indicates a different path during transition. These countries put in place comprehensive social safety nets including unemployment compensation, social security benefits and a functioning pension system in order to address inequal- ity issues. Russia and the other CIS countries, on the other hand, installed a system of taxes and transfers that actually reinforced the effect of market forces resulting in a highly unequal income distribution. According to Svejnar, this provides for a prin- cipal lesson and positive prospect: Although newly introduced market forces in the formerly centrally planned and state-owned economies in transition generally resulted in greater inequality, there is still room for government policies to mitigate these effects. The workshop participants conclude by outlining key issues that need to be addressed in the World Development Report. Equity is a widely debated and com- plex concept. It thus asks for a clear definition in order to firmly guide action and development policies. Studies revealed the close link between equity and pro-poor growth and discussed equity as a necessary correlative condition for sustainable eco- nomic growth. It becomes clear that inequality matters to welfare not only intrinsi- cally whenever people have a social preference for equality. But inequality also mat- ters instrumentally because studies showed that progressive redistribution accelerates poverty reduction or, that growth elasticity of poverty reduction falls with inequal- ity. Hence, equity and efficiency do not necessarily represent a trade-off-relationship-- they can complement and foster each other. It is important to carefully analyse dif- ferent redistributive instruments with respect to their effects on growth and efficiency and vice versa. 6 | GUDRUN KOCHENDÖRFER-LUCIUS AND BORIS PLESKOVIC In addition to conceptual clarifications and further analysis of development poli- cies, these policies should be embedded within a broader institutional framework. It was highlighted that equitable institutions can be an important building block for sustainable social and economic development. Therefore, with respect to institutions and policies that enhance equity and fight poverty, we need a twofold framework. On the one hand, this framework must focus on distributions of assets and opportu- nities and distribution of incomes as well as of voice and agency. On the other hand, the framework needs to include the analysis and adaptation of economic institutions and political and sociocultural institutions as well as institutions of governance in order to serve equity as well as sustainable pro-poor growth. As pointed out by François Bourguignon, traditional analyses of inequality focussed on income distribution or the distribution of consumption per capita. The approach of the World Bank in its current WDR, however, shifts the focus from inequality of outcome to inequality of opportunities in order to guarantee that every- body starts more or less from the same level. Thereby the WDR seeks to widen the inequality lens: Instead of looking only at poverty and the gap between the poor and the rich, it looks rather at the full range of distribution explicitly, including the top end. Thus, the WDR takes both ends as complementary and necessary for develop- ment analysis: like two sides of the same coin. All in all the workshop made it clear that the 2006 WDR demarcates a shift in the World Bank analysis keeping important continuity at the same time: The continuity lies in the emphasis on poverty and growth, and the innovative perspective is the focus on distributional aspects--thereby including the other side of the coin. Equity and development, therefore, should not be thought as contradictory but rather be understood as complementary and potentially reinforcing. It is one of the merits of this workshop to have contributed to such a diversified view on the interrelationship between equity and development within and between countries worldwide and to constructively discuss the potential risks and chances. Opening Address MICHAEL HOFMANN It is a pleasure for me to welcome you to this workshop in preparation for the World Development Report 2006--Equity and Development, which is jointly hosted here in Berlin with the Development Policy Forum of InWEnt, the Federal Ministry for Economic Cooperation and Development (BMZ), and the World Bank. I am grateful to the Delegation of the European Union (EU) for hosting this ses- sion. We try to be good Europeans, even though we are perhaps not the easiest part- ners. No doubt, we have to improve German development assistance as it also applies to EU aid. In Rome we all agreed to work on harmonization, and we will reiterate this in Paris. We try to integrate bilateral and multilateral assistance--and seek co-financing for working together. But we must become better service providers for the developing countries, which should be in the driver's seat for their own development. I hope their governments are driving toward development with equity. If develop- ing countries want to achieve sustainable development, they must make sure that income distribution is less uneven than it is in too many countries. Recently (in early September 2004), I read in the Financial Times: "A stone's throw from some of the poorest and most violent shanty towns in the world, customers sip champagne and eye the latest imported diamond rings in an exquisite jewellery store." And this arti- cle on Brazil stated that ". . . the coming to power of Ignacio Lula da Silva, Brazil's first working-class president, in January 2003, did little to curb the ability or desire of Brazil's wealthy to indulge in extravagance." I mention this because we are, indeed, constantly asked for increases in develop- ment assistance. We discuss the International Financial Facility, and there are other ideas for doubling the development assistance to finance the achievement of the Mil- lennium Development Goals (MDGs). It is important that those in the industrialized countries who have to provide the financial means do not get the impression that the Michael Hofmann is Director General in the Federal Ministry for Economic Cooperation and Development (BMZ) in Bonn/Berlin, Germany. Berlin Workshop Series 2006 © 2006 The International Bank for Reconstruction and Development/The World Bank 7 8 | MICHAEL HOFMANN poorer segments of our society should transfer money to the wealthy in developing countries. It is imperative to make clear that we are striving for structural, and hence, distributional change in developing countries. Hence we touch on political ground, which is normally not the mandate of the World Bank. If finance ministers and executive directors are taken into account, word- ing must be used in the WDR that is understood by these economic actors. However, you should reach beyond this audience. If you want to reach more people in the devel- oped countries and in developing countries, you have to speak out clearly--and in polit- ical terms--to say that a highly uneven income distribution can no longer be tolerated. Let me remind you of one of François Bourguignon's predecessors at the World Bank, Hollis Chenery, who 30 years ago coauthored the famous book Redistribution with Growth. I again read C.L.G. Bell's article in that volume, on the political frame- work for redistribution, with a positive reference to Cuba and Vietnam. When the World Bank hails achievements in poverty alleviation, it still has to refer to these countries or to China--but tends not to mention their socialist background. Why is it that China or Vietnam can achieve remarkable growth and at the same time have an income distribution that is still in a tolerable range, while other coun- tries have a distribution that is not? Let me answer this question with Bell: "Indeed, a strategy which seeks to remove poverty may prove empty or impossible to imple- ment, unless the poor develop sufficient consciousness and organization to provide support from below to maintain the momentum of the poverty alleviation program over a period of many years" (Chenery and others 1974, page 65). And he goes on to say: "The key factor is the emergence of a coalition of interest able to grasp power which sees some advantage in implementing a redistributive strategy, despite the fact that some sections of it stand to lose relatively thereby" (page 72). I mention this because of its political flavor; and I hope to read some of that in your WDR, too. Equity is vital for European societies as well; and it must be considered, especially in the process of enlargement. There is the tremendous challenge to avoid running into extreme disparities in the new member states of the EU, knowing that in the process toward enlargement, disparities have already increased. That is, unfortunately, true for many of the world's "transition economies." For instance, in China during the course of rapid growth, disparities have increased in terms of regional income and assets. It is good to know that the authors of the 2006 WDR will also consider empow- erment and gender inequality, highlight regional inequalities, and touch on the cru- cial issue of global inequality. All of this is most welcome. Generally, we in government and political circles have high expectations for your report. Even though in intellectual cycles research issues repeat themselves, this is justified when we have lost balance. One must not reinvent the wheel of development but reinvent equity and strengthen this theme, because it has somehow been forgotten in recent years. Let me refer to David S. Landes, one of my intellectual heroes and author of The Wealth and Poverty of Nations: Why Some Are So Rich and Some So Poor. Landes states that, "If we learn anything from history of economic development, it is that OPENING ADDRESS | 9 culture makes all the difference" (Landes 1998, page 516). I hope you will not shy away from exploring different cultural endowments and what they imply for addressing inequality. For example, the upper classes in Latin America have a rela- tively weak commitment to their respective countries compared to the strong com- mitment of elites in Asia. Does culture drive development in a different manner in different regions and thus determine equity and inequity accordingly? What does equity mean in Africa, with its colonial past of extreme inequality and a history of extroversion--the tendency Jean François Bayard characterizes as typical for African elites to externalize challenges and expect solutions from the outside? It is fundamental to keep in mind, as Landes states, that ". . . advanced and backward, rich and poor, do not seem to be growing closer" (page 518). However, we have to achieve the opposite; we must find the ways and means needed to close the gaps in the longer run. Even though your former colleague William Easterly denounces the "cartel of good intentions," we have to remain optimistic in order to achieve a better world for all. David Landes encourages us: "In this world, the optimists have it, not because they are always right, but because they are positive . . . The one lesson that emerges is the need to keep trying . . . We must cultivate a skeptical faith, avoid dogma, lis- ten and watch well, try to clarify and define ends, the better to choose means" (page 524). Means are essential, because in the published WDR you must present rec- ommendations. My hope is that you come up not only with general recommenda- tions, because the policies and institutions needed to change the fabric of societies have to be country-specific. You need to find a balance between general recommen- dations and best practices that are, to a large degree, country-specific. In recent years, the World Bank has become more pluralistic in a very positive way. Knowing that one size does not fit all, it is seeking solutions that are tailor- made. Carry on with pluralistic approaches; focus, and be courageous with political messages. Development and equity warrant more than technocratic fine-tuning. Structural change with equity implies political interest and the political will to serve people in need. Welcome to Berlin, and thank you for your courage in addressing the pertinent issue of equity in development. References Chenery, Hollis, Montek S. Ahluwalia, C.L.G. Bell, John H. Duloy, and Richard Jolly. 1974. Redistribution with Growth. Oxford, UK: Oxford University Press for the World Bank. Landes, David S. 1998. The Wealth and Poverty of Nations: Why Some Are So Rich and Some So Poor. New York: W.W. Norton. Opening Address Inequality of Opportunity and Economic Development FRANCISCO H. G. FERREIRA AND MICHAEL WALTON What is the role of equity in the process of development? Our proposed answer to this question, which is discussed at greater length in the World Development Report 2006: Equity and Development (World Bank 2005), draws heavily on the discus- sions we had at the InWEnt workshop on this topic held in Berlin, Germany, in Sep- tember 2004. Many of the links between normative and positive perspectives on the distribution of opportunities that are highlighted here were noted during those debates. While our views have continued to evolve over the past year, we owe a great deal to the contributions made by the extraordinary group of Berlin workshop par- ticipants, in the early stages of this project. Much of their work is summarized in this volume. One of the ideas that forcefully emerged from the InWEnt Berlin workshop on equity and development was that two relatively recent developments in thinking about distribution in economics, which have remained largely unrelated so far, ought to be much more closely connected. The first is the acknowledgement that distribution--in particular the distribution of wealth--may affect aggregate out- comes, such as the overall level of output, or its rate of growth. This was, of course, a theme of classical economists, who intuitively understood the importance of distri- bution in political economy. It has also been recognized more recently, as in Nicholas Kaldor's view that the poor and the rich have different savings rates. Kaldor hypoth- esized that increases in income inequality today could lead to greater prosperity tomorrow, by increasing the average savings rate from a given amount of output. This paper was commissioned for this volume of presentations from the Berlin workshop hosted by InWEnt (Capacity Building International, Germany) in September 2004, in preparation for the World Development Report 2006: Equity and Development. Francisco Ferreira is at the World Bank and Michael Walton is with the Kennedy School of Govern- ment at Harvard University. As the lead authors, they are grateful to their colleagues on the team that prepared WDR 2006, especially Abhijit Banerjee, Peter Lanjouw, Tamar Manuelyan-Atinc, Marta Menendez, Berk Özler, Giovanna Prennushi, Vijayendra Rao, Jim Robinson, and Michael Woolcock, on whose work this paper draws extensively. They also thank François Bourguignon for many helpful discussions, and Martin Ravallion for comments on an earlier ver- sion of this paper. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 11 12 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON But distributional considerations had been peripheral to mainstream neoclassical economics until the early 1990s, when a series of important papers suggested that, if credit and insurance markets were imperfect, the distribution of wealth might mat- ter for the level and composition of aggregate investment, and hence to total output levels (see, for example, Galor and Zeira 1993). Different initial wealth distributions could also affect occupational choice and, through its impact on the relative supply of and demand for labor, determine wage trajectories and aggregate development paths (Banerjee and Newman 1993). A variety of other mechanisms were later pro- posed, through which unequal wealth distributions could reduce economic efficiency when capital markets are imperfect.1 The result was, as Atkinson (1997) put it, to "bring income distribution in from the cold" (p.297). Separately, it was also suggested that politics could be another channel through which distribution affected outcomes. If governments were not benevolent dictators, but instead represented the (possibly conflicting) interests of different groups in soci- ety, then the expected distributional outcomes of different policies (such as tax rates, or public expenditure decisions) would feature in the public decisions about them. The implication was that the policies actually chosen and implemented need not be optimal from a social point of view. They might instead be optimal from the private point of view of the pivotal voter, dominant group, or government agent that makes the decision. To the extent that wealth (or income) affects either the individual's pref- erence for different policy alternatives, or his power to influence the ultimate govern- ment choice (or both), the distribution of wealth may affect the choice of policies, and hence the degree of the resulting inefficiency. Early models of these policy decisions in a median-voter framework included Alesina and Rodrik (1994) and Persson and Tabellini (1994). Later, the interaction between political economy mechanisms and capital market imperfections allowed for an even richer set of possible outcomes, including one in which unequal wealth leads to inequality in political power and, consequently, to inefficiently low levels of redis- tribution. Plausible models exist in which such reinforcement between economic and political inequalities might lead to multiple equilibria, with some featuring higher inequality and lower output levels than others.2 The second development in thinking about distribution preceded these models of distribution and aggregate outcomes, and took place in the areas of public choice, welfare economics, and theories of social justice--along the frontier between eco- nomics and philosophy. It consisted of a move away from ex post realizations--such as incomes and utilities--and toward ex ante potentials as the appropriate metrics for social welfare, or as the appropriate spaces in which to judge the fairness of a given allocation or system. John Rawls (1971) may have been the pioneer in this essentially normative (and highly influential) literature, but he was soon joined by others, such as Ronald Dworkin (1981), Amartya Sen (1985), G. A. Cohen (1989), and John Roemer (1998). As the titles of some of their most important contributions indicate, these authors were concerned with the space in which one should seek to measure, understand, and influence distribution.3 INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 13 Although each author was different in important respects, the thrust of their efforts begins, with the passage of time, to seem similar in essence. Rawls's "Differ- ence Principle" sought to maximize the availability of primary goods to the least privileged group; Sen wrote about capabilities; Dworkin spoke of equality of resources; and Roemer emphasized equality of opportunities.4 While a number of worthy treatises have been written on the subtle distinctions between these different normative approaches, a broad common tendency can be identified in this evolution in the theory of social justice over the past three decades or so. And that is the move- ment away from actual ex post outcomes (such as incomes) and their effects on the well-being of the individual (such as utilities), toward sets of potential outcomes, ex ante (such as capabilities or opportunities). We argue that these two separate developments in thinking, in two apparently remote areas of economics, should not remain unconnected. The reason is that inequality in opportunity--in addition to being an arguably superior concept on which to anchor the normative evaluation of alternative social states--may well turn out to be precisely the right concept for the empirical testing of theoretical hypothe- ses about how distribution affects aggregate efficiency and growth. Most models that propose links between distribution and aggregate levels of out- put do not actually refer to income distributions. The key concept is usually the dis- tribution of wealth (as in Galor and Zeira 1993, and Banerjee and Newman 1993) and, crucially, the extent to which, under imperfect capital markets, wealth levels may affect the set of feasible investment opportunities (or occupational choices). If education is a lumpy investment process with fixed costs, then those who are "too poor" may not have the opportunity to invest, despite the fact that returns may be high and that the investment would have been undertaken if the credit market were perfect. Entrepreneurship may be a preferable occupation to being a wage laborer but again, if there are imperfect or missing credit markets, then the poor may not have that option, regardless of ability. If wealth and ability both determine the allo- cation of students to the best schools or colleges, then it will not be the ablest stu- dents who attend the best schools (Fernández and Gali 1999). The models are, therefore, fundamentally about the distribution of opportunities. Inefficiencies arise because the people who seize the opportunities (for education, investment, or entrepreneurship) are neither as many nor the same individuals as would have been the case if markets worked perfectly. Aggregate output is lower because capital ends up being invested at lower marginal returns by some richer investors, rather than at higher returns by credit-constrained ones. Or because less gifted children from rich families have the chance to attend good schools, while clever children from poor families do not--and then go to schools with bad teachers or simply drop out. Think a little bit more broadly about opportunities, and the same logic begins to apply to the political economy models. That class of ideas revolves around power. The same distribution of wealth can generate very different economic outcomes-- some efficient, some not--under different assumptions mapping wealth to political 14 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON power (Ferreira 2001). If one is prepared to think of a person's--or a group's-- ability to influence political decisions in their community as part of their opportunity set, then the political economy channel between inequality and efficiency (or growth) is also fundamentally about the inequality of opportunities. The empirical implica- tions should be clear: If institutional set-ups (including the relative freedom of the press, the independence of the judicial system, and the transparency of the campaign finance system) differ across societies, the same degree of wealth inequality should lead to different aggregate efficiency outcomes. It all depends on the mapping between economic wealth and political power. There may, therefore, be a coincidence between the normative concept toward which philosophers have been gravitating as a defining feature of the just society, and the positive concept of which greater equality may imply greater efficiency. It was the possibility of this remarkable convergence that motivated this focus on equity and development in the WDR 2006. In what follows, we sketch out some of the arguments that the WDR 2006 pres- ents in much greater detail. We first present some evidence that morally irrelevant predetermined circumstances--that is, factors over which individuals have no con- trol, and which society deems to be irrelevant in terms of their deserts--do, in fact, powerfully affect outcomes. Following Roemer (1998), we interpret these effects as prima facie evidence of the existence of inequality in opportunities. We then argue that the existence of these inequalities in opportunity constrains development-- through many of the mechanisms suggested in the theoretical literature, and in other ways as well. We conclude by considering some implications of the analysis, both for policy and further research. Predetermined Circumstances Shape Lives Opportunity sets begin taking form for individuals while in utero. Who one's parents are, how rich they are, and what country they live in make a great deal of difference for a person's opportunities. The opportunity to life itself turns out to depend on such predetermined circumstances as the education and wealth of one's parents, their access to clean water and sanitation, and the availability of medical treatment. Con- sider figure 1, which plots group-specific infant mortality rates across countries, from roughly 2000­4. Each vertical line in the figure corresponds to one country, and within each country, the highest point in the line indicates infant mortality (per 1,000 live births) among children whose mothers have no education; the lowest point gives the corresponding figure for those whose mothers have completed secondary school- ing or higher. The differences are striking, not only across countries, but perhaps even more so within them. In El Salvador, for instance, babies born to mothers with no schooling are four times as likely to die before their first birthday as their coun- terparts with better educated mothers. Parental education is not, unfortunately, the only predetermined circumstance that affects the basic opportunity for life. There is some evidence that gender does INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 15 FIGURE 1. Infant Mortality Varies across Countries, but Also by Mothers' Education within Countries 160 140 births live 120 No education 1,000 100 per 80 rate 60 Secondary or higher 40 mortality 20 Infant 0 il y ia h n YemebodiZambia n a ar e dan e dpr ea Beninasc ublicineaFasoChadIvoir Gu Malawi opia ndaMaliNiger ColombiaS Jorri LankaParaguayTh Vietnam pinasailandwanaAfricaPeru onesia EgyptnesiaGhanaroccoKenya EritrSudan Nepal ades ooTo go rosHaiti Uganda Rwa ElTurkm Salva Pakistan Ethi Philip Bots BrazTurk IndBolivia Nicaragua Tu SenegalNigeria enistan Guatemala Ind Como e d' South ZimbabweMo BanglCamer Cam Madagn Rep ca Burkina Cot Mozambique Afri ral Cent Source: World Bank (2005), from Demographic Health Survey (DHS) data. Note: The continuous dark line represents the mean infant mortality rate in each country, while the endpoints of the vertical whiskers indicate the infant mortality rates by different levels of the mothers' education. too, at least in parts of Asia, where juvenile sex ratios are unusually high. A juvenile sex ratio simply measures the number of 0 to 4 year-old boys in a population, rela- tive to the number of 0 to 4 year-old girls. Because slightly more boys than girls are born in a typical population at any given time, that ratio oscillates between 1.00 and 1.05 in most countries. Remarkably, in the Indian states of Punjab and Haryana, the 2001 ratio was above 1.20. In China, it reached 1.17 in 2000. While some recent work suggests that the patterns of incidence of hepatitis B--and the fact that it leads to more male births--may account for some of these differences (see Oster 2005), the dominant view is that this unusual discrepancy is casually related to son-preference in these societies, implemented through selective abortion and postnatal care (see Sen 1990 and Klasen and Wink 2003). Opportunities continue to depend on morally irrelevant, predetermined circum- stances even if one survives the first year of life. Access to basic health care, such as immunization services, is strongly correlated with parental wealth (figure 2). Even a child's cognitive skills seem to develop at different rates depending on family back- ground, for children as young as 3 to 5 years old. Figure 3, drawn from Paxson and Schady (2004), shows the evolution of vocabulary recognition test scores (TVIP, see note at figure 3) for two groups of children from Ecuador: those whose parents have 0 to 5 years of schooling, and those whose parents have 12 or more years of schooling. 16 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON FIGURE 2. Access to Childhood Immunization Services Depends on Parents' Economic Status 70 60 de 50 cover 40 Poorest not 30 centage Per 20 10 Wealthiest 0 (*) a (*) ia (*) i occGhanaBeni o n os ia y go ea r Egyptdan mbi Faso RwandaPeruAfricaKenya MalawiBrazil Turkeytemala an Boliv tan (*) men Ye oon ndaIndia aniaHaitTo Nige Chad Uga EthiRe opiapublicgascar GuineaMali odiakistanEritr Jor Colo a Tanzanonesisia Mor Comor Paragua Pa akhs Camb South ZambiaVietnamGu Ind Philippinas Bangladesh Maurit Mada Turkmenit Kaz BurkinaCamer African Mozambique tral Cen Source: World Bank (2005), from Demographic Health Survey (DHS) data. Note: The continuous dark line represents the percentage of children without access to a basic immunization package in each country, while the endpoints of the vertical whiskers indicate the percentages for the top and the bottom quintile of the asset ownership distribution. * indicates that the poorest quintile have higher access to childhood immunization services than the wealthiest quintile. By the time these children enter primary school, at age 6, they have markedly differ- ent learning abilities, shaped in large part by differential family backgrounds. These statistical associations do not establish causality, of course. While endo- geneity should not be a concern in the associations presented here--because rates for access to immunization or child mortality today cannot cause parental education decades earlier--omitted variables clearly exist. The descriptions presented earlier are essentially bivariate correlations. They do not establish the effect of, say, parental education on vocabulary recognition, or immunization. Parental education is obvi- ously correlated with wealth, housing quality, access to water, distance to and qual- ity of school, and possibly even with genetic endowments of ability that can be trans- mitted across generations. There is an important literature that seeks to identify each of these individual effects, and it is clear that the simple patterns described here do not do so.5 What they do is suggest that the collection of these predetermined circum- stance variables (parental education, wealth, location, access to services, and so forth), which cannot be controlled by the infant or young child, do powerfully shape the choices--or opportunity set--of the young. INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 17 FIGURE 3. Child Cognitive Skill Development by Maternal Education 110 100 e scor 90 TVIP 80 Median 70 60 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 age in months 0-5 years 12 or more years Source: Paxson and Schady (2004). Note: TVIP (Test de Vocabulario en Imagenes Peabody) is the Spanish version of the Peabody Picture Vocabulary test. Unequal Opportunities Deter Development in a Number of Ways The fact that predetermined, morally irrelevant circumstances influence opportunities (and therefore final outcomes) was all that mattered to the second strand of thinking mentioned at the beginning of this paper. The normative consensus--or at least some degree of convergence--that was emerging judged such inequality of opportunities to be ethically undesirable from the point of view of social justice. In this section, we argue that it is exactly this sort of inequality--in the predetermined circumstances that shape opportunities--that leads to aggregate inefficiency, in the spirit of the first strand of thinking summarized earlier. Consider three telling examples. The first example is from an agricultural setting in Ghana, where land is allocated by custom, and security of property rights is therefore often linked to the local power structure. Goldstein and Udry (2002) find that individuals are less likely to leave their land fallow (an investment in long-run productivity of the land) if they do not hold a position of power within either the hierarchy of the village or the hierarchy of the lineage. The problem is that land is taken away from those who allow it to lie fal- low. Because women rarely hold these positions of power, their land is not left fal- low often enough and so is much less productive than land controlled by men. This land becomes degraded, because women do not have the social status needed to hold on to it during the fallow periods. The key point for our argument is that the result- ing decline in land productivity is a pure loss for society. The fact that other people 18 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON do have status and can fallow their land as needed does not, in any way, compensate for the loss of productivity on the land of the powerless. A separate study by the same authors (also in Ghana) provides a second example of how unequal opportunities that arise from the interaction between poverty and imperfect or missing markets leads directly to inefficiency. In the forest-savannah in Southern Ghana, cocoa cultivation, receding for many years because of the swollen shoot disease, has been replaced by a cassava-maize intercrop. Recently, however, pineapple cultivation for export to Europe has offered a new opportunity for farmers in this area. In this 1997 and 1998 study, more than 200 households cultivating 1,070 plots in four clusters in this area were surveyed every six weeks for about two years. The survey results reveal that the profitability of pineapple production dominates that of the traditional intercrop (figure 4).6 Despite the fact that the average returns asso- ciated with switching from the traditional maize and cassava intercrops to pineapple is estimated to be in excess of 1,200 percent, only 190 out of 1,070 plots were used for pineapple. When the authors asked farmers why they were not farming pineapple, the virtually unanimous response was: "I don't have the money."7 While it is true that some heterogeneity in ability between those who have switched to pineapple and those who have not cannot be entirely ruled out, the authors conclude that the fixed costs involved in switching crops, along with imperfections in credit markets, prevent a large number of farmers from making a very profitable investment. Output and income levels in these areas are correspondingly below potential. A final example comes from the impact of belonging to a low caste on individual performance. To examine the effect of stereotypes on the ability of individuals to FIGURE 4. Average Returns in Ghana for Switching to Pineapples as an Intercrop Can Exceed 1,200 Percent 20,000 Cedis) 15,000 (1,000 10,000 ofits Pr e 5,000 0 -Hectar Per -5,000 0 20 40 60 80 100 cumulative percent of plots pineapple profits non-pineapple profits Source: Goldstein and Udry (1999). Note: The cedi is the official currency of Ghana. INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 19 respond to economic incentives, Hoff and Pandey (2004) undertook experiments with low- and high-caste children in rural north India. The caste system in India can be described as a highly stratified social hierarchy in which groups of individuals are invested with different social status and social meaning. In the first experiment, groups composed of three low-caste ("dalits") and three high-caste junior high school students were asked to solve mazes and were paid based on the number of mazes they solved. In one condition, no personal information about the participants was announced. In a second condition, caste was announced along with each participant's name and village. In a third condition, participants were segregated by caste; and then the name, village, and caste were announced for each participant in the six-person group. When caste was not announced, there was no caste gap in performance (figure 5). But increasing the salience of caste led to a significant decline in the average perform- ance of the low caste, regardless of whether the payment scheme was piece-rate (that is, participants were paid one rupee per maze solved) or tournament (that is, the par- ticipant who solved the most mazes was paid six rupees per maze solved; the other participants received nothing). When caste was announced, the low-caste children solved 25 percent fewer mazes on average in the piece-rate treatments, compared with the performance of subjects when caste was not announced. When caste was announced and groups were composed of six children drawn from only the low caste (a pattern of segregation that for the low caste implicitly evokes their traditional out- cast status), the decline in low-caste performance was even greater. While one cannot be sure from these data what the children were thinking, some combination of loss of self-confidence and expectation of prejudicial treatment likely explains the result. The expectation by the low-caste subjects of prejudicial treatment may be rational given the discrimination in their villages. But the discrimination itself is unlikely to be FIGURE 5. Children's Performance in India Differs when Their Caste Is Made Public Average number of mazes solved, by caste in five experimental treatments 8 High caste 7 Low caste 6 5 4 3 2 1 0 Piece-rate, Piece-rate, Tournament, Tournament, Tournament, caste not caste caste not caste caste announced announced announced announced announced and segregated Source: Hoff and Pandey (2004). Note: A vertical line in the figure illustrates the statistically significant caste gaps. 20 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON fully rational. Cognitive limitations may prevent others from judging stigmatized individuals fairly. The fact that people are bounded in their ability to process infor- mation creates broad scope for belief systems--in which some social groups are viewed as innately inferior to others--to influence economic behavior. If such beliefs persist, it will generally be rational for those discriminated against to under-invest (with respect to others) in the accumulation of skills for which the return is likely to be lower for them. These three examples from Ghana and India illustrate a growing body of micro- economic evidence of the inefficiency of inequality.8 One case--that of Ghanaian farmers unable to switch from cassava and maize to pineapple--exemplifies the clas- sic interaction between fixed costs, poverty, and a missing credit market. The farm- ers were too poor to pay the fixed costs required for making the switch. If a perfect credit market had existed, they would have been able to borrow against the large expected returns of switching crops, in order to finance the investment. Market imperfections and a mass of poor people at the bottom of the distribution lead to missed opportunities and X-inefficiency. But the other two examples are different. The Indian children who solve fewer mazes when they are explicitly reminded of their inferior social status do not require any markets that might be missing or imperfect. Yet, in a convincing experimental setting, their productivity is reduced by the mere existence of the social hierarchy. If similar declines in productivity occur in real work situations, the private and social losses would be no less important. The Ghanaian women farmers who cannot ade- quately fallow their land, leading to losses in its productivity, similarly do not suffer from poverty combined with missing markets. The channel here is an inequality in power, when effective property rights are power-dependent. What all three situations have in common is that differences in wealth, power, or status generate unequal opportunities for productive investment. In all cases, these inequalities cause society to remain shy of the Pareto frontier. A concern with these mechanisms is particularly justified because of the evidence that unequal productive opportunities persist across generations, over long periods of time. The World Development Report 2006 highlights two broad mechanisms through which inequalities are reproduced--leading to what it calls inequality traps. One is the simple fact that many of an adult's outcomes (such as education and wealth levels, or where one lives) will be his children's predetermined circumstances. If the children's outcomes are affected by the circumstances, the ingredients for inter- generational persistence are present. In fact, a growing literature on intergenerational mobility (or the lack thereof) has documented the impact of parental background on both achievement and the degree of transmission of status across generations. In the United States, Mazumder (2005) finds an intergenerational earnings elasticity of 0.6, which implies that a family currently earning half the national average income can expect to take five generations to reach the average.9 Estimates for developing coun- tries are few and far between, but can be even higher. Dunn (2003) estimated an elas- ticity of 0.69 for Brazil. The second mechanism for the persistence of inequalities is institutional endogene- ity. Since Douglas North and Oliver Williamson, modern economists have under- INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 21 stood that the manner in which individuals and firms interact in markets is condi- tioned by the nature of nonmarket institutions--formal and informal rules and norms of behavior, and the agencies that enforce them. Among the most important roles of these nonmarket institutions are the definition and enforcement of property rights and contracts. People will not invest if property rights are not well-defined and enforced, or if they believe that the contracts they write will not be honored. The state must also provide a whole set of other inputs apart from social order and fair contract enforcement. These include various types of public services and regulations. Lying behind well-functioning markets are legal systems, judges, policemen, and, ultimately, social groups and politicians. But institutions, like policies, are not designed by a benevolent dictator. They evolve over time in response to the actions of individuals and groups who seek to protect their own interests. It follows that institutions--again just like policies--need not be optimal from a social viewpoint. It is perfectly possible that the institutions that are best-suited to the short-term interests of a particular group in a particular generation are not those most conducive to broad-based economic growth and devel- opment. If that particular group happens to be very powerful, however, it is also pos- sible that those institutions end up prevailing, despite not being the best ones for growth and development. The World Development Report 2006 discusses a number of examples, both his- torical and contemporary, of different institutional developments that appear to have been driven by different degrees of political and economic inequality. One revealing comparison--which draws on work by Acemoglu, Johnson, and Robinson (2001), and Engerman and Sokoloff (1997)--is that between European colonies in North and South America. Those colonies, such as present-day Brazil, Mexico, or Peru, where initial factor endowments enabled the colonizers to establish extractive insti- tutions based on highly concentrated property and control structures (such as the mita system of forced labor used in the Andean silver mines, or the capitanias hered- itárias in northeastern Brazil), tended to do less well in the long run than colonial backwaters, such as present-day Canada and northern United States, where condi- tions were not ripe for producing any of the colonial era's most desirable commodi- ties, such as gold, silver, or sugar. In these places, instead of imposing concentrated patterns of land ownership and indentured or slave labor institutions, colonists were left to their own devices. The absence of large native populations (who could be exploited and dispossessed) or slaves (who were not imported, since soils and climates were not suitable for the crops that would justify the investment) meant that free populations of European descendants were soon in the majority. Rather than making rules designed to prevent the exploited native (or enslaved) masses from sharing in prosperity, these colonists soon demanded greater autonomy in decision making. Because population density was low and there was no way to extract resources from indigenous peoples, early commercial developments in Canada and the United States had to import British labor. And, relative to much of the colonial world, the disease environment was benign, stimulating settlement. Indeed, the Pilgrim fathers decided to migrate to the United States rather than Guyana because of the high mortality rates in Guyana.10 22 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON Limited supplies of labor gave workers a greater bargaining power, forcing elites to extend political rights and create equal access to land and the law. The World Development Report 2006, following Acemoglu, Johnson, and Robin- son (2001) and Engerman and Sokoloff (1997), argues that these initial institutional differences between North and South American colonies have persisted for centuries, and led to important differences in economic outcomes between the two sets of coun- tries. The argument is that institutions that rely on a high concentration of property and control over resources, and offer limited opportunities for investment and inno- vation for large segments of the population, are less efficient. Because talent and ideas are widely distributed in the population, economies in which the property of all people is secure and in which there is equality before the law for all (rather than just for some) tend to do better. Similarly, political systems that provide access to services and public goods for all are associated with superior long-term economic performance. Policy and Research Implications The thesis of this paper is that inequality of opportunity, which has gained promi- nence in modern normative thinking about social justice, is also a highly relevant concept for understanding the positive links between distribution and efficiency. Wealth inequalities (combined with market imperfections), inequalities in power, and status differences have all been shown to lead to inefficiency--in a number of contexts. It has also been argued, prominently and plausibly, that large inequalities in power and wealth can lead to institutional characteristics that are associated with lower subsequent growth. The broad implication for public action is that the effects of any policy on the dis- tribution of opportunities will in general have an impact on aggregate efficiency, which needs to be taken into account in any assessment or evaluation of the policy. In some cases, the impact may be direct, and relatively simple to measure. Building a road that connects a poor and geographically isolated area to markets may increase profit margins on sales of its produce and lower the costs of consumption goods in (possibly) measurable ways. In many other cases, however, measuring the efficiency gain from redistributing opportunities is much more difficult. How does one quan- tify the savings from ethnic conflicts that are avoided by a (hypothetically) success- ful integration or affirmative action program? How long must one wait until the full benefits of educating girls today shows up in the opportunities available to their chil- dren in the future? We will return to these measurement and long-term evaluation challenges. Con- ceptually, however, the general point is that if highly unequal opportunities generate inefficiencies, then reductions in these inequalities may well be efficient. Ignoring the full long-run benefits (or costs) of any reductions (or increases) in inequality of opportunity will, therefore, generally result in an under-provision of efficient redistribution.11 INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 23 The World Development Report 2006 discusses implications of this general point in considerable detail, and for a variety of policy areas. The lasting impact of early childhood nutrition and mental development on subsequent opportunities implies that the payoffs to investment in early human development programs are likely to be large. The imperfections of insurance markets imply that the existence of appropri- ate social security systems can enable efficient (but riskier) investment decisions to be made. Complementarity between infrastructure and private capital often argues for an expansion of access to the poorest groups. Market rules and institutions are often skewed toward more powerful groups, sometimes generating substantial inefficiency in resource allocation. Though mentioned only briefly here, the four chapters dedicated to these themes in the WDR 2006 insist that appropriate policy recommendations can only be made with an adequate understanding of the local context. The question of whether the highest social return to a marginal dollar--even once the full benefits of equity are taken into account--accrues to improving the rural road network or expanding a conditional cash transfer scheme clearly cannot be answered in the abstract. The answer to any such question will surely depend on specific conditions in the country. In some cases, a priority for both efficiency and equity reasons will lie in reforming a captured and corrupt financial system. In others, it may be that the marginal dol- lar should be returned to the taxpayer, in the form of lower taxes. In others yet, the expected long-term returns on a publicly funded expansion in basic health care may be so high that taxes may need to rise. Such context specificity, while rather fashionable these days, has serious implica- tions of its own for applied research. If the concept of opportunity sets and the dis- tribution of opportunities actually turn out to matter for development policy, consid- erable progress will be needed in their measurement. We therefore close this paper with a suggestion for three areas where further applied research would be helpful, if policymakers are to be presented with evidence on the basis of which better-informed decisions can be made. Measurement of inequality of opportunities is surely one area where improvement can be made. If opportunities are perfectly correlated with incomes--or wealth-- then, clearly, one needs only to measure those variables accurately. The hypothesis that is often proposed, however, is that the determinants of opportunity are many, so that the partial correlation with any one of them is imperfect. Wealth may very well affect opportunities but so do, the argument goes, race, disability, gender, caste, place of birth, and so forth. If that is so, can one compute acceptable summary indi- cators of this ex ante set of potential outcomes? Initial attempts have been made, but the area is in its infancy, and more work is needed.12 Since inequality of opportunity is closer to the concept of distribution that is rel- evant to the theoretical literature on inequality and growth than income inequality, a related question is whether turning to such an indicator might shed light on the inconclusive empirical cross-country literature on the subject. While most cross- section regressions of growth on initial income inequality (with controls) have returned negative and significant coefficients on inequality, most panel regressions of 24 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON growth on lagged, time-varying income inequality (with controls) have returned pos- itive and significant coefficients.13 If we were able to measure inequality of opportu- nity directly for a number of countries, so that one no longer needed to rely on income inequality as a proxy, would the cross-country result shed light on the "macro- economic" effect that corresponds to the microeconomic impacts identified through studies such as those discussed in the previous section? The question remains open. Long-term evaluation of projects and policies aimed at greater equality of oppor- tunities is a second policy-relevant area where the current stock of knowledge is insufficient. The WDR 2006 suggests that, in a number of instances, returns on some investments in the opportunity-poor may suffer from a considerable delay. The impact of better nutrition for expecting mothers may not show up until today's fetus becomes tomorrow's adult. Gains from reduced ethnic conflict that may follow from investments in poor ethnic minorities today may accrue decades into the future. And so on. However difficult properly evaluating impacts over extended periods of time may be, attempts would have to be made, unless policymakers are expected to take such benefits on pure faith.14 A related question is that of quantifying costs and benefits. It has long been under- stood that policy choices depend not only on a careful identification of their impact, but also on a relative quantification of the costs and benefits of individual projects, relative to alternatives. The difficulties are many, but it is somewhat unfortunate that the basic insights from the literature on cost-benefit analysis--from Little and Mirr- lees (1969) and Drèze and Stern (1987), for instance--are so unfashionable these days. However unglamorous such estimations may appear to contemporary journal editors, there is likely to be substantial policy payoff from combining the much improved techniques for measuring impact (from experimental and matching meth- ods) with an understanding of the need to value their costs and benefits. The link between equal opportunities (or equity more broadly) and institutional quality is the third area where current knowledge is insufficient. While the historical evidence presented by Acemoglu, Johnson, and Robinson (2001) and Engerman and Sokoloff (1997) is persuasive, it falls short of identifying the precise mechanisms through which equity affects institutions. The arguments are plausible, but more evidence--and different categories of evidence--is needed. A promising avenue in experimental economics is the use of laboratory experiments in which subjects play variants of a basic game, with controlled changes in rules (institutions). One interest- ing example is the Repeated Public Good Game, in which individuals must decide whether to keep their money (at zero returns), or invest it in a common project with high returns. Rates are set such that the highest returns to all players are attained when all invest 100 percent in the common project, but the dominant strategy for each individual player is to free-ride on others: or, not to invest one's own resources, but seek to benefit from the high returns in the common pot. Fehr and Gachter (2000) find that actual behavior in the game differs substantially depending on whether or not punishment (expending real resources to punish noncooperative behavior) is allowed. When punishment is permitted, even a small number of altru- istic players can sustain a cooperative (and Pareto-superior) equilibrium. While in INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 25 this example institutions are exogenous, and one investigates their effects on out- comes, it might be possible to design experiments where rules are endogenous, and where the distribution of endowments changes.15 As research on these different fronts evolves, we may learn more about the nature and extent of redistributive activity that governments should seek to pursue, even if they are concerned exclusively with dynamic efficiency. Normative considerations will always remain important, and they are additional to these insights. Such an agenda would complement the substantial evidence presented in the WDR 2006 about the extent of the inequality of opportunity that exists today, both within and between countries, and about the impact of that inequality on investment and institutions. Notes 1. See, for example, Aghion and Bolton (1997), Piketty (1997), and Aghion, Caroli, and Garcia-Penalosa (1999). 2. See, for example, Bénabou (2000) and Ferreira (2001). 3. A number of papers echoed Amartya Sen's (1980) Tanner Lecture on Human Values, which was entitled "Equality of What?" 4. Sen defines a person's capabilities as the set of all possible functionings--actions and states of being--from which this person can choose. 5. See Card (2001) for a survey. 6. From Goldstein and Udry (1998); see figure 4. 7. From Goldstein and Udry (1998), p. 38. 8. To paraphrase the title of Glaeser, Scheinkman, and Shleifer (2003): "The Injustice of Inequality." 9. Mazumder's estimate is considerably higher than previous estimates (of about 0.4). See, for example, Solon (1992). The difference arises mainly from the author's use of a long- term social security earnings history of fathers and children, which allows him to reduce the variance of the transitory component of incomes inherent in previous measures of father's economic status, usually drawn from the Panel Study of Income Dynamics (PSID) or from the National Longitudinal Surveys (NLS). 10. See Crosby (1986), pp. 143­44. 11. The term "efficient redistribution" has been used before. In one prominent discussion, Bowles and Gintis (1996) argued that a number of "asset-based" redistributions--including that of property rights over firms to workers; over houses to tenants; of school vouchers to parents; and of parental income streams to children--would increase efficiency, by transfer- ring residual rights of control over assets to those whose actions more directly affect asset use or maintenance. While there are substantial differences between their detailed proposals and those put forth in the WDR 2006, there is also one basic broad similarity: both empha- size the potential for asset redistributions that enhance, rather than attenuate, productive incentives. Both approaches suggest that equity can and should be pursued in a market- friendly manner, with potentially large efficiency gains. 12. See Bourguignon, Ferreira, and Menéndez (2003), Cogneau, and Gignoux (2005), and O'Neill, Sweetman, and Van de Gaer (2000). 26 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON 13. See Bénabou (1996) for a survey of the cross-section results, Forbes (2000) and Li and Zou (1998) for two influential panel studies, and Banerjee and Duflo (2003) for a possi- ble interpretation of the differences. 14. The difficulties in carrying out even well-designed experimental evaluations over a long period are illustrated by a recent medium-term evaluation of student achievement follow- ing the PROGRESA (now Oportunidades) program (see Behrman, Parker, and Todd 2005). 15. Other research strategies are evidently also possible, including detailed case studies of how particular institutions develop, or looking for governments (national or local) that are willing to conduct "institutional experiments." References Acemoglu, Daron, Simon Johnson, and James Robinson. 2001. "The Colonial Origins of Comparative Development: An Empirical Investigation." American Economic Review 91 (5):1369-401. Aghion, Philippe, and Patrick Bolton. 1997. "A Trickle-Down Theory of Growth and Devel- opment with Debt Overhang." Review of Economic Studies 64 (2):151-72. Aghion, Philippe, Eve Caroli, and Cecilia Garcia-Penalosa. 1999. "Inequality and Economic Growth: The Perspective of New Growth Theory." Journal of Economic Literature 37 (4):1615-60. Alesina, Alberto, and Dani Rodrik. 1994. "Distributive Politics and Economic Growth." Quarterly Journal of Economics 109 (2):465-90. Atkinson, Anthony B. 1997. "Bringing Income Distribution in from the Cold." Economic Journal 107 (441): 297-321. Banerjee, Abhijit, and Esther Duflo. 2003. "Inequality and Growth: What Can the Data Say?" Journal of Economic Growth 8 (3): 267-99. Banerjee, Abhijit, and Andrew Newman. 1993. "Occupational Choice and the Process of Development." Journal of Political Economy 101 (2):274-98. Behrman, Jere, Susan Parker, and Petra Todd. 2005. "The Longer-Term Impacts of Mexico's Oportunidades Program on Educational Attainment, Cognitive Achievement, and Work." Department of Economics, University of Pennsylvania. Unpublished. Bénabou, Roland. 1996. "Inequality and Growth." In Ben Bernanke and Julio J. Rotemberg, eds. National Bureau of Economic Research Macroeconomics Annual 1996. Cambridge, MA: MIT Press. ------. 2000. "Unequal Societies: Income Distribution and the Social Contract." American Economic Review 90 (1):96-129. Bourguignon, François, Francisco Ferreira, and Marta Menéndez. 2003. "Inequality of Out- comes and Inequality of Opportunities in Brazil." World Bank Policy Research Working Paper 3174, World Bank, Washington, DC. Bowles, Samuel, and Herbert Gintis. 1996. "Efficient Redistribution: New Rules for Markets, States, and Communities." Politics & Society 24 (4): 307-42. Card, David. 2001. "Estimating the Returns to Schooling: Progress on Some Persistent Econo- metric Problems." Econometrica 69 (5): 1127-60. Cogneau D., and J. Gignoux. 2005. "Earnings Inequalities and Educational Mobility in Brazil over Two Decades." DIAL Working Paper 2005/03, Développement, Institutions et INEQUALITY OF OPPORTUNITY AND ECONOMIC DEVELOPMENT | 27 Analyses de Long terme, Paris. http://www.dial.prd.fr/dial_publications/PDF/ Doc_travail/2005-03.pdf. Cohen, G.A. 1989. "On the Currency of Egalitarian Justice." Ethics 99: 906-44. Crosby, Alfred. 1986. Ecological Imperialism: The Biological Expansion of Europe 900-1900. New York: Cambridge University Press. Drèze, Jean, and Nicholas Stern. 1987. "The Theory of Cost-Benefit Analysis." In A. Auer- bach and M. Feldstein, eds. Handbook of Public Economics. Amsterdam: North- Holland/Elsevier Science Publishers. Dunn, Christopher. 2003. "Intergenerational Earnings Mobility in Brazil and Its Determi- nants." Department of Economics, University of Michigan. Unpublished. Dworkin, Ronald. 1981. "What is Equality? Part 2: Equality of Resources." Philosophy and Public Affairs 10 (3):185-246. Engerman, Stanley, and Kenneth Sokoloff. 1997. "Factor Endowments, Institutions, and Dif- ferential Paths of Growth among New World Economies: A View from Economic Histo- rians of the United States." In Stephen Haber, ed. How Latin America Fell Behind: Essays on the Economic Histories of Brazil and Mexico, 1800-1914. Palo Alto, CA: Stanford Uni- versity Press. Fehr, Ernst, and Simon Gachter. 2000. "Cooperation and Punishment in Public Goods Exper- iments." American Economic Review 90: 980-94. Fernández, Raquel and Jordi Gali. 1999. "To Each According to . . . ? Markets, Tourna- ments, and the Matching Problem with Borrowing Constraints." Review of Economic Studies 66 (4):799-824. Ferreira, Francisco H. G. 2001. "Education for the Masses? The Interaction between Wealth, Educational, and Political Inequalities." Economics of Transition 9 (2): 533-52. Forbes, Kristin J. 2000. "A Reassessment of the Relationship between Inequality and Growth." American Economic Review 90 (4): 869-87. Galor, Oded, and Joseph Zeira. 1993. "Income Distribution and Macroeconomics." Review of Economic Studies 60: 35-52. Glaeser, Edward, Jose Scheinkman, and Andrei Shleifer. 2003. "The Injustice of Inequality." Journal of Monetary Economics 50: 199-222 Goldstein, Markus, and Christopher Udry. 1999. "Agricultural Innovation and Resource Management in Ghana." Department of Economics, Yale University. New Haven, CT. Unpublished. ------. 2002. "Gender, Land Rights, and Agriculture in Ghana." Department of Economics, Yale University. New Haven, CT. Unpublished. Hoff, Karla, and Priyanka Pandey. 2004. "Belief Systems and Durable Inequalities: An Exper- imental Investigation of Indian Caste." World Bank Policy Research Working Paper 2875, World Bank, Washington, DC. Klasen, Stephan, and Claudia Wink. 2003. "Missing Women: Revisiting the Debate." Femi- nist Economics 9: 263-99. Li, Hongyi, and Heng-fu Zou. 1998. "Income Inequality Is Not Harmful for Growth." Review of Development Economics 2 (3):318-34. Little, Ian, and James Mirrlees. 1969. Manual of Industrial Project Analysis in Developing Countries: Social Cost Benefit Analysis. Paris: Organisation for Economic Co-operation and Development. Mazumder, Bhashkar. 2005. "The Apple Falls Even Closer to the Tree Than We Thought." Chapter 2 in S. Bowles, H. Gintis, and M. Groves, eds. Unequal Chances: Family Back- ground and Economic Success, Princeton, NJ: Princeton University Press. 28 | FRANCISCO H. G. FERREIRA AND MICHAEL WALTON O'Neill, Donal, Olive Sweetman and Dirk Van de Gaer. 2000. "Equality of Opportunity and Kernel Density Estimation: An Application to Intergenerational Mobility." Department of Economics, NUI Maynooth, Maynooth, Ireland. http://www.nuim.ie/academic/economics/ pdf/N950999.pdf. Oster, Emily. 2005. "Hepatitis B and the Case of the Missing Women." Harvard University, Department of Economics, Cambridge, MA. Unpublished. Paxson, Cristina H., and Norbert Schady. 2004. "Cognitive Development among Young Chil- dren in Ecuador: The Roles of Wealth, Health, and Parenting." World Bank. Washington, DC. Unpublished. Persson, Torsten, and Guido Tabellini. 1994. "Is Inequality Harmful for Growth?" American Economic Review 84 (3):600-21. Piketty, Thomas. 1997. "The Dynamics of the Wealth Distribution and the Interest Rate with Credit Rationing." Review of Economic Studies 64: 173-89. Rawls, John. 1971. A Theory of Justice. Cambridge, MA.: Harvard University Press. Roemer, John E. 1998. Equality of Opportunity. Cambridge, MA.: Harvard University Press. Sen, Amartya. 1980. "Equality of What?" In S. McMurrin, ed. Tanner Lectures on Human Values. Cambridge, UK: Cambridge University Press. ------. 1985. Commodities and Capabilities. Amsterdam: North-Holland. ------. 1990. "More than 100 Million Women are Missing." The New York Review of Books 37 (20) (December 20). Solon, Gary. 1992. "Intergenerational Income Mobility in the United States." American Eco- nomic Review 82 (3): 393-408. World Bank. 2005. World Development Report 2006: Equity and Development. Washington, DC: World Bank. Keynote Address Distribution, Equity, and Development FRANÇOIS BOURGUIGNON At the outset, I wish to thank our host and organizers of the Berlin Workshop for giving us this wonderful opportunity to discuss our ideas and offer suggestions regarding the 2006 World Development Report (WDR). I also extend my thanks to all the presenters for their feedback on the topic of "Equity and Development," the central theme of this WDR. As noted yesterday, there is a definite European flavor to this topic--and I appreciate the expectations that have been raised on this side of the Atlantic with respect to this publication. However, the selection of this topic should not be considered as a means for Euro- pean influence on a major World Bank report. Instead, it reflects a true global con- cern for equity and development. There is increasing evidence of such a concern all around us, starting perhaps with my own appointment as Chief Economist of the World Bank. We hope that many questions are raised and answered on the subject of equity: over the course of these meetings, those that will follow in other parts of the world among different participants, and certainly in the forthcoming WDR. The organizers of this seminar are to be commended for opening up several different and fundamen- tal directions for this debate. I draw from my recent lecture delivered to the Congress of the European Eco- nomic Association and the Econometric Society in Madrid, a fantastic quote from Victor Hugo. In his novel Les Misérables he makes the following recommendation to societies and their rulers: "Limiter la pauvreté sans limiter la richesse;" that is, "curb poverty without containing affluence." It appears that this quote is remarkably well- aligned with today's economists' views on the subjects of equity and development. I was amazed to see such relevant ideas on economic equity presented by Hugo more than a century ago. I believe that this idea should be the backbone of the WDR in preparation. François Bourguignon is Chief Economist and Senior Vice President, Development Economics at the World Bank. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 29 30 | FRANÇOIS BOURGUIGNON Poverty and Equity Moving on to the key topic of this conference: What is equity? What is the role of governments in the promotion of equity? Do governments of developed and devel- oping nations have differing roles in this regard? Equity is a concept that has much to do with moral and political philosophy. At this juncture, however, I prefer a more analytical view of this issue and will elabo- rate on the ways in which we at the World Bank may think about equity. I will also discuss the relationship of this concept both to poverty and to the poverty agenda of this institution. Over the past few weeks, I have been asked a lot of interesting ques- tions on the topic of this WDR: Are inequality, inequity, and poverty the same thing? If so, what would be the difference between the 2006 WDR and the 2001 WDR, which was on poverty? Would you bring anything new to the debate on poverty? If inequity is different from poverty, then is the Bank moving away from its previous course? All of these questions are equally important and contain different elements of truth. The broad answer to these questions is that there is continuity as well as inno- vation. Let me now briefly expand on these two ideas of continuity and innovation. The continuity in this WDR is in the emphasis placed on poverty. If we use an inter- national definition of equity--that is, global equity as will be the case during the ses- sions later today--then it seems that we interpret equity as simply the problem of the divide, or the gap, between income and the standard of living across countries. Thus, increasing equity in the world is reducing the gap between the poor countries and the rich countries universally. Dealing with equity and development at the global level is similar to dealing with poverty in the world. In particular, this implies promoting growth and improving welfare in poor countries. In that sense, dealing with global equity is achieving full continuity in development objectives. While focusing on equity today, we carry with us the goal of poverty eradication throughout the world. The novelty of this WDR is that equity must also be considered at the national level. At this level, there are differences between poverty and equity--just as there are differences between poverty and inequality. When referring to poverty, we talk about only one part of the whole distribution of income and standards of living in a coun- try. To focus on poverty is, to some extent, the equivalent of ignoring what is going on at the top of the distribution and considering only those people whose incomes are below some poverty line, defined either in absolute or relative terms. The biggest difference that the concept of equity introduces with respect to poverty is that one must look at the spread in its entirety. Specifically, saying something about what is going on at the bottom requires saying something about what is going on at the top of the distribution. In this way, the originality aspect of the 2006 WDR will be to no longer ignore any one part of the distribution. Equity versus Equality Some other questions frequently asked are: What is the difference between equity and equality? Why not call the 2006 WDR "Inequality and Development" instead of "Equity and Development"? KEYNOTE ADDRESS: DISTRIBUTION, EQUITY, AND DEVELOPMENT | 31 The distinction is crucial and is highlighted in the following way. Traditional analyses of inequality focus on the distribution of income, consumption per capita, or some meaningful welfare equivalent, concluding that some countries are more or less "equal" than others, or that policy may more or less reduce the degree of "inequality" of income or consumption. A critical problem here is the implicit assumption of an optimum state of the world where inequality would be eliminated. This egalitarian view of the world is not the approach that we plan to take in this WDR. We do not wish to defend or to promote egalitarianism per se. When we ana- lyze policies to "level the playing field," the objective is to equalize opportunities in order to generate income or consumption rather than income or consumption per se. When examining equity, we are more interested in the opportunity sets, or the fea- sible sets of actions open to all agents--their "capabilities" according to Amartya Sen, not necessarily (or exclusively) in the outcomes of their activity in terms of income or welfare. Equity is about the equality of opportunity sets and the possibil- ity for everybody to generate outcomes above some arbitrary deprivation threshold. This focus is not independent of income inequality or income redistribution. For example, if we want to achieve equality of opportunities in education, then we need to invest in educating the poor. But, in order to raise that money, we need to increase taxes, which may achieve some redistribution in income. The ultimate objective is, however, to equalize opportunities; taxes and investment in education are only a means to ensure that everybody starts more or less at the same level. Is equity an abstract concept with fewer or lesser implications for policy than inequality? The second session of this seminar discusses policies that can impact equity. History has much to offer in this regard. Major changes in the degree of inequality of opportunities have been observed in certain parts of the world. In Europe, such changes were most notable after the Great Depression in 1929, and during the periods preceding and following World War II. When we look at figures for these years, we observe not only a major change in the degree of income inequal- ity, irrespective of the index or indicator used, we also observe very substantial progress in the field of social protection, the universal supply of public goods like education or health care, or in the equality of treatment of ethnic and gender groups. This is extremely important, as it shows that it is actually possible to change the degree of equity in a society. Notwithstanding the fact that these outcomes were achieved with considerable difficulty, what matters is that the transformation was accomplished. There are many similarities when we compare the degree and level of development of the provision of social services and of equitable economic and political processes in Europe back in the 1930s or 1940s, with the development of many middle-income countries today. The achievements of Europe are not impossible to recreate in these middle- income countries. This evolution should be considered when looking at the history-based evidence on social transformations in Europe. One very important dimension to the recent transition in Eastern Europe and Central Europe is the move toward greater democ- racy. This really touches upon a key dimension of equity that is difficult to address when one has a pure economic view of development issues. We at the World Bank 32 | FRANÇOIS BOURGUIGNON are convinced that it is important to leave room in the WDR for discussion of the noneconomic dimensions of equity: human rights in particular, and democracy, as well as the examples of transition countries. Social Demand and State Welfare Reform The third session of the seminar is about building efficient welfare states and the les- sons that have been learned in this regard. A lot has been written on welfare states, especially those in Europe, over the past couple of decades. In particular, a vast amount of literature addresses the need to reform the perceived inefficiency-generating fea- tures of the welfare state. But at the same time, it is interesting to observe that today, despite this strong emphasis on limiting the welfare states, very little change is tak- ing place. When we look at the share of public spending corresponding to the wel- fare state--the redistribution that is taking place via income tax, social transfers, the whole social insurance system--we find that its share in GDP spending has not changed among the OECD countries. This means that social demand is very strong, even though there is concern about reforming that system to make it as consistent as possible with economic efficiency. Several middle-income countries are in a situation today that more or less corre- sponds to the situation many European countries were in at the time that the welfare system was being constructed. When I travel to Latin America, for example, we have discussions on pension systems, health insurance systems, labor taxes to finance those expenditures, and the problem that only a small part of all workers are in the formal sector. The situation is not unlike that in Europe 60 or 70 years ago: only a fraction of the population was covered by social insurance, only a fraction of the population was covered by redistribution, with a large part of the population sim- ply escaping those systems. What happened was simply that over time, the propor- tion of people who could not be covered declined significantly. Thus, there is much to be learned about building welfare states and more equitable societies in develop- ing countries. When we look at the evolution of the welfare state in Europe in particular, it is very difficult to imagine that this system would be what it is today without the phenome- nal growth performance of the 20 to 30 years that followed World War II. In those days, there was a huge social demand for redistribution and welfare policy. It was pos- sible to meet the social demand with this strong growth. The problems we are facing today in many developing countries are the opposite: growth has a tendency to slow down, and there are demographic imbalances emerging over time. As a result, we are in a situation where some parts of the welfare system may not be sustainable if imported without suitable adaptations in developing countries. The challenge then is to identify these modifications and the ways in which they might contribute to more equity and more efficiency of the economy in developing countries. So, the issue then arises of whether it is possible to do much in developing coun- tries in terms of development of the welfare system if there is no growth. Economic KEYNOTE ADDRESS: DISTRIBUTION, EQUITY, AND DEVELOPMENT | 33 theory indicates that, given the imperfection of actual economies, more social protec- tion and more equity in general may enhance economic efficiency and growth. This question thus provides us with an interesting perspective. We need satisfactory growth in order to develop and meet the social demand for more welfare policies and more equity. We need more equity and social protection to foster economic efficiency and growth. These two dimensions of the question will be addressed in the 2006 WDR. The title of this section referred to the welfare state, whereas the argument was about social protection and equity in general. There is, of course, a considerable overlap between social protection and equity defined above as equality of opportu- nities. But equity means more than social protection and social insurance. It has implications both for social expenditures, particularly in the fields of education and health, and also for equity in the way individuals are dealt with both in the economic and political spheres. Examples in the economic sphere involve the absence of dis- crimination in the markets for labor or goods and services. Examples of equitable treatment in the political sphere include the need to give all people equal voice in public decisionmaking at the local or at the central level. While refering to progress of the welfare state in European countries over the past decades, we actually have in mind this broader equity framework. We are aware that results in several dimensions of that framework are still limited--as with the issue of social exclusion that arose in several countries over the past couple of decades. Global Inequalities and Global Equity The fourth session in this seminar deals with the reduction of global inequalities in income and standards of living. There are diametrically opposing viewpoints on this subject, with one section contending that inequality is increasing and the other arguing that the degree of global inequality along with global poverty is actually on the decline. The World Bank has recently published figures tracking the evolution of poverty over the past 20 years, showing that, without any doubt, poverty reduced between 1980 and 2002 for those persons living on less than one dollar per day--from around 40 percent to 20 percent of the world population.1 This is a major achievement. It is important, however, to understand why this evidence seems to be ignored by some of the participants of the global inequality debate and why there are diverging points of view. First, some people measure inequality across countries by weighing their population size, whereas others do not. Obviously, if we take into account that big countries such as Bangladesh, China, and India did well over the past 20 years, it is clear that because of the size of those giant countries, the whole world did well and there is less global inequality and poverty. It is also true that if we compare the poor countries within Africa to the richest countries in the world and examine the difference in terms of average income per capita between them, the gap has undoubt- edly increased. If we focus on the first aspect, we conclude that inequality and poverty have gone down; but if we focus on the second aspect, we have to conclude that inequality and poverty are on the rise. 34 | FRANÇOIS BOURGUIGNON One explanation, which is consistent with the idea that "equity" and "inequality" are different concepts, is that people who insist on the growing gap between coun- tries take into account not only what the distribution looks like today--the current number of poor people in the world--but also the evolution of people over time.2 In other words, they consider the dynamics of the whole process of changes in incomes and standards of living rather than comparative statics. It is true to some extent that the world is faring better because many formerly very poor Chinese people are now far less poor. But there are now additional poor Africans that have suffered a con- siderable loss in their incomes. For some, the fact that more people emerged from poverty as compared to the numbers still facing poverty is seen as a favorable evolu- tion. For others, however, the fact that some Chinese improved their living standards is negated by the fact that others in sub-Saharan Africa suffered a loss, even though there may be more people in the former situation than in the latter. Supporting one view or the other requires specific value judgments. What we know for sure is that there was no Pareto improvement in the world; that is, not everybody in the world is faring better today compared to what could be observed 10 or 20 years ago. As a result, then, both viewpoints need to co-exist. That a con- siderable number of people in the world were able to improve their living standard is unambiguously a good thing. That some poor people became poorer is unambigu- ously a bad thing. Yet, weighing one aspect against the other is bound to lead to ambiguous conclusions based on particular views at global welfare. This argument is important, because we often tend to have an anonymous view of the world. We do not care about the identity of the poor; we simply care about the number of poor people, or the proportion of people below some threshold. But in my earlier discussion about equity, the view is not an anonymous one. We do in fact care about the identity of the people, their level of income, and their opportunity sets. Hence, when we talk about equity, this dimension must be taken into account. In the Berlin workshop program booklet, I liked the description of the session on global inequalities. It reads: "If the world were a country . . ." If this were true, we would certainly care about the fact that some of our fellow citizens have seen their oppor- tunities shrinking over time. This idea of a world being a country is a dream, but it presents an extremely inter- esting perspective on the problems the world is facing today. Due to information flows in all directions, the world is converging more and more into one single coun- try. Since we now have up-to-date information on our neighbors and their future eco- nomic prospects, one might ask whether there is a maximum level of inequality that the world can bear. This is the most important question that we would like to answer at the global level. Conclusions I will conclude with the fifth topic, the "New Europe"--and the fact that we are witnessing an ongoing integration within the European region. What is striking from the standpoint of the previous expansion of the European Union (EU) was KEYNOTE ADDRESS: DISTRIBUTION, EQUITY, AND DEVELOPMENT | 35 its absolute success. Greece, Portugal, and Spain are cited as extraordinary examples of growth and progress in international equity, owing in no small measure to their inclusion into the EU. This model will most probably be replicated with the latest round of expansions. It might not happen as quickly--as suggested by some recent public consultations among EU members--or it may occur under different condi- tions, but there is at least the expectation that something corresponding to the cre- ation of a bigger and better community is currently under way. In response to the question "Is there a world community?" we could say that the European integration is an example of the ways in which solid international communi- ties can be built. This is a good example, because we know that only so much can be achieved through international flows like aid, trade, or (constrained) migration. Much more can be achieved through the progressive integration of countries. Clearly, the global equity agenda is of utmost importance. Promoting more equity at the global level requires faster growth in poor and slow-growing middle-income countries, which in turn may require more equity within those countries. We need freer flows of goods, services, capital, people, and knowledge between rich and poor countries. Ultimately, we also need faster progress toward economic integration of countries at different levels of development. Along with the international develop- ment community, we welcome your collaboration on this forward-looking agenda of the 2006 WDR. Notes 1. See World Bank (2004). 2. See Bourguignon, Levin, and Rosenblatt (2004). References Bourguignon, F., V. Levin, and D. Rosenblatt. 2004. "Declining International Inequality and Economic Divergence: Reviewing the Evidence through Different Lenses." Economie Internationale 100: 13-25. World Bank. 2004. World Development Indicators. Washington, DC: World Bank. What Is Equity? What Is the Role for Governments in the Promotion of Equity? Does the Role Differ between Developed and Developing Countries? Promoting Equity through Social Cohesion: The Council of Europe Strategy GABRIELLA BATTAINI-DRAGONI Equity in Principle: The COE Strategy for Social Cohesion The Council of Europe (COE) believes that the best response to the risks of the grow- ing income inequality, poverty, and social exclusion that threaten our societies is a comprehensive strategy for social cohesion. As a vehicle for international cooperation on social cohesion, the COE offers its unique situation as a pan-European forum, bringing together 46 states united by their commitment to a Europe based on pluralist democracy, respect for human rights, and the rule of law.1 Through 50 years of intergovernmental cooperation, the COE has established European norms for social rights and built up a wealth of accumulated experience through the supervisory mechanisms of its legal instruments, as well as in its recom- mendations and reports. Its committees and working parties make the council a forum for the exchange of ideas and experience across the whole European conti- nent. Increasingly, moreover, through its cooperation programs, the council assists its member states in putting the organization's standards into practice in their spe- cific national situations. Its many and varied means of action give the council the potential to exert a considerable influence on social development across the continent. The Human Dignity and Social Exclusion Project (1994­8) The Human Dignity and Social Exclusion Project (HDSE Project), a pan-European initiative launched by the COE in 1994 and completed in 1998, conducted analysis on the status of poverty and social exclusion in Europe and submitted proposals for action that were validated by the final conference of the project, which convened in Helsinki in May 1998. The project proposed action on five major themes: health, Gabriella Battaini-Dragoni is Director General of Education, Culture and Heritage, Youth and Sport for the Council of Europe (COE), headquartered in Strasbourg, France. She was formerly Director General of Social Cohesion. This paper was co-authored by Stefano Dominioni, who is an administrator of the COE. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 39 40 | GABRIELLA BATTAINI-DRAGONI housing, employment, social protection, and education. It focused primarily on pro- ducing "alarm indicators" that record reductions in freedoms and rights such as increases in unemployment, violence, and conflicts; the worsening of social and health care services, and so forth--mapping out the presence of social exclusion. The goal was to present a series of corrective intervention measures in these areas, aimed at integrating those who are excluded from society into the labor market and society as a whole. However, through this work the COE realized that its member states must not only find ways of dealing with the problems of those who find themselves excluded from society. They must also, more ambitiously, find ways to build more cohesive societies in which the risks of social exclusion are minimized and access to a reason- able or good quality of life are made possible. The Second Summit of Heads of State and Governments: Committing to a Strategy for Social Cohesion (1998­2002) The COE decided therefore to review its approach and, in turn, develop a strategy for creating social cohesion rather than a strategy for combating social exclusion. Heads of state and representa- tives of the governments of the COE member states who met in October 1997 for the council's second summit identified social cohesion as "one of the foremost needs of the wider Europe and . . . an essential complement to the promotion of human rights and dignity" (Final Declaration). They instructed the Committee of Ministers "to define a social strategy to respond to the challenges in society and to carry out the appropriate structural reforms within the Council of Europe." The first step taken by the ministers was to set up a new intergovernmental steering committee, the European Committee for Social Cohesion (CDCS), bringing together several for- merly separate areas of work. The committee's terms of reference state that "the first task of the Committee will be to prepare a strategy for the development of social cohesion activities within the Council of Europe for consideration by the Committee of Ministers;" it is further indicated that this strategy should contain "a programme of work for the medium term." The Revised Strategy for Social Cohesion The Council of Europe Strategy for Social Cohesion was adopted by the Committee of Ministers on July 1, 2000 and reconfirmed, in its revised version, on March 31, 2004.2 A Functional Definition of Social Cohesion In existing formulations, the def- inition of social cohesion is generally based on a necessary condition, such as shared values and objectives, the sense of belonging to the same community or interpersonal commitment, or the expected outcomes of social cohesion (including the ability to work together; stable, cooperative, and lasting communities; or social interaction). Similarly, etymological definitions emphasizing links only focus on a particular aspect of social cohesion. The COE has adopted the functional definition shown PROMOTING EQUITY THROUGH SOCIAL COHESION | 41 here, rather than a descriptive one, highlighting four aspects of individual and collec- tive well-being that are fostered by social cohesion: equity, individual and collective dignity, autonomy, and participation. Social cohesion is the capacity of a society to ensure the welfare of all its members, min- imizing disparities and avoiding polarization. A cohesive society is a mutually support- ive community of free individuals pursuing these common goals by democratic means. Adopting the COE definition for social cohesion means adopting the promotion of a society where human dignity, equity, pluralism, sustainability, and responsibility are the fundamental determinants of social life. In this framework, the council believes that both the state and the economic sectors have fundamental roles to play in ensur- ing the social cohesion of society, as well as civil society, and the community as a whole. The Essential Role of the State as Guarantor of Human Rights At a time when the proper role of the state is the subject of much debate, the COE strategy on social cohesion clarifies and reaffirms the essential responsibilities of government in the area of social policy. First, the state acts as guarantor of human rights (including social rights) and par- ticipatory democracy. Rights provide the firmest foundation for social policy; and rights put all members of society on an equal footing. With a basis in human rights, the action of the state in the social policy field is no longer a matter of charity or wel- fare directed at the less fortunate members of society; it is a question of guarantee- ing rights that are the same for all--through equitable access and nondiscriminatory practices. Second, the state restores a sufficient degree of equity in the distribution of wealth, through mechanisms of solidarity such as redistributive taxation and social security. Social security systems are, indeed, one of the most powerful institutional expressions of social solidarity. A primary goal of the strategy for social cohesion is therefore to strengthen sustainable social security systems, especially at a time when many questions are posed about the future development and financing of those systems. Third, the state acts to protect vulnerable groups that are at risk of social exclu- sion. Social protection systems thus provide not only social security--which is designed to protect all members of society against life's risks--but also social assis- tance and social services for those in particular need. Although the COE strategy acknowledges that these responsibilities of the state remain essential, the manner in which the state fulfills them is changing. In many countries there is a tendency to bring social services closer to the people by devolv- ing responsibility to regional and local authorities. In addition to this, govern- ments increasingly find that the best way of facilitating access to social rights is by working in partnership with civil society in order to achieve shared goals of social cohesion. 42 | GABRIELLA BATTAINI-DRAGONI Equity through Integrating the Social Dimension into Economic Life Since the 2002 World Summit on Sustainable Development in Johannesburg, it has been increasingly recognized that sustainable economic development depends on both sus- tainable social development and a sustainable environment. For these reasons, eco- nomic policy and social policy need to be brought into much closer alignment than has been customary. The right to work is a fundamental right and a key element of citizenship. Access to decent and adequately remunerated employment is probably the most important single factor in combating poverty and exclusion--and one of the most powerful in creating more equitable societies. This is why social protection systems now try to assist as many people as possible in moving from a situation of passive welfare dependence to active participation in the economy. The social responsibility of economic actors is emerging as a crucial question for postindustrial societies. There is growing interest in the question of the social respon- sibility of business, and numerous instruments are being developed to put this responsibility into practice. Private actors, businesses, and corporations now have the economic weight and capacity to mobilize resources and transfer technologies that largely outshine those of international institutions; and they can have an enormous impact on development processes. It is crucial that the actions of these groups be embedded in a system of rules and with a greater ethos of responsibility for the con- sequences of their undertakings, thereby contributing to the well-being of society as a whole--not just that of a privileged minority. Today there is also concern on the part of individuals who want to develop new forms of economic action that enable them to contribute to more equitable societies. People are using the means of action now available to them as individuals, such as their power as consumers and the ways in which they use their savings. In all this, the challenge is to find ways of ensuring that the market economy contributes to social cohesion and does not exclude those who are considered the least attractive as consumers. The Family, the Community, and Civil Society as the Fabric of Social Cohe- sion and Promoters of Equity Human beings find their fulfillment, above all, in relationships with others. A social cohesion strategy, while fully respecting the auton- omy of the private sphere and of civil society, must therefore seek to support fami- lies and other networks and relationships that bring individuals together into wider circles of solidarity. Families are the place where social cohesion is first experienced and learned, and they play a crucial role in preparing children for life in society. Families are often called upon to support elderly people as they become more dependent--and are the place of last-resort for social protection in time of need. Changes in values and behavior have resulted in a greater variety of family patterns than in the past, but this does not in any way detract from the social role of families, which remains as funda- mental as ever. In a complex and changing society, it is necessary to support families PROMOTING EQUITY THROUGH SOCIAL COHESION | 43 in fulfilling their functions. Parents in particular need help in counteracting harmful social and market pressures, in reconciling the demands of work and family life, and in adapting to situations in which children are seen as bearers of their own rights as individuals. Society has a special responsibility toward those who, for one reason or another, cannot count on the support and protection of a family or other social network. The rights and dignity of children or elderly people in need of care must be the subject of special attention. Traditional care institutions play a diminishing role, so it is impor- tant to develop alternative means of care wherever possible. Individuals participate in many other networks and institutions that help to knit society together. A variety of civil society bodies, including churches, political par- ties, and trade unions, are also important factors of social cohesion--although in most countries each of these has seen some loss in membership and influence, per- haps a reflection of a society that values individual freedom more than collective belonging. However, nongovernmental organizations (NGOs) and voluntary bodies of all kinds are flourishing and becoming indispensable partners of government in building social cohesion. Charitable, sports, and cultural associations, together with organizations for children and young people, play a particularly important part in building social cohesion. It is bodies of this type that often make suitable partners for government-sponsored programs to build social cohesion and equitable societies. Promoting Values of Diversity, Pluralism, and Solidarity A strategy for social cohesion must first identify the factors of division within society and then design or facilitate mechanisms, processes, and institutions that prevent those divisions from becoming so acute as to endanger social order and peace. Diversity is not in itself divi- sive. European societies have been learning, albeit hesitantly, to see ethnic, religious, cultural, and ideological pluralism not as an obstacle to social cohesion but as a source of wealth, dynamism, adaptability, and strength. The question is then how to manage diversity so that it becomes a source of mutual enrichment rather than a source of divi- sion and conflict. A cohesive society is one that has developed satisfactory ways of coping with these and other strains in an open and democratic manner. To manage diversity in a way that releases its potential for good is, however, a challenge. It involves fighting racial, ethnic, religious, gender, and other forms of dis- crimination. And it requires active policies to integrate migrants and minorities and groups with particular needs into mainstream society on the basis of respect for their differences and full recognition of their rights. An important part of social cohesion is to find an adequate balance between the responsibility of the individual and the solidarity of society. For without solidarity, the individual will not be ready to adapt to the structural changes brought about by a rapidly changing economy and social structure. The challenge is to develop a greater sense of interdependence, of mutual responsibility, and belonging--between diverse individuals and groups--so that they can effectively contribute as members of equitable societies. 44 | GABRIELLA BATTAINI-DRAGONI Equity in Practice: The Normative Work of the COE The COE's social cohesion strategy to prevent inequality and social exclusion has at its core a strategy for guaranteeing access to social rights for all. Social Rights and Minimum Standards for All In support of its long-standing commitment to social rights, the COE has always considered access to be based on the following principles: · equality of rights for all, without discrimination; · availability of quality services for all; · special attention to the needs of the vulnerable members of society; · avoidance of stigmatization for those with special needs; · maintenance of equitable and sustainable social protection policies; and · participation of users in available services and institutions. For these reasons, the council advocates international legal mechanisms as an essen- tial foundation for equitable societies. The European Social Charter (1961) and the Revised European Social Charter (1996) protect a broad range of social and eco- nomic rights encompassing labor rights, including the right to work and the right to just (or fair) working conditions, as well as social security rights. In addition to these, the charter provides for the right to social and medical assistance, the right to bene- fit from social services, and the protection of certain disadvantaged groups. The revised social charter reinforces this by introducing several new rights. In this con- text, article 30 ensures the right to protection against poverty and social exclusion, by ". . . promoting the effective access of persons who live or risk living in a situa- tion of social exclusion or poverty, as well as their families, to, in particular, employ- ment, housing, training, education, culture and social and medical assistance." The charter's supervisory machinery plays a decisive role in the implementation of these rights. Based on the examination of national reports, it allows a regular and systematic legal appraisal of the observance of activities in the various states. By now, nearly all of the 46 COE member countries have either signed or ratified the Euro- pean Social Charter or the Revised European Social Charter.3 This means that both the relatively prosperous countries and the countries in difficult economic situations adhere to these common standards. Why are social rights essential in promoting equity and in fighting inequality? · Instead of being considered a social cost, social rights are the best investment strat- egy for a community. Social rights empower individuals, especially the most vul- nerable, by giving them physical, economic, educational, and health resources to promote their human capital. Social rights enable each individual to contribute to their own social development, by giving them access to educational and employ- ment opportunities, thereby providing access to stable income. PROMOTING EQUITY THROUGH SOCIAL COHESION | 45 · The cost of social rights is often advanced in order to justify reductions in social provision or social programs. However, the economic, social, and political costs that result from the implementation of these restrictive measures (or from the fail- ure to invest in social rights) are rarely mentioned. What would be the social, eco- nomic, and political consequences of abandoning or undermining policies based on social rights? What would be the human cost? For these reasons, the COE advocates not only access to social protection, but also minimum standards of social protection for all (through various forms of health care, old-age pension schemes, unemployment benefits, sickness and work injury schemes, maternity insurance, child benefits, and so forth)--as the foundation of a society free from extreme poverty and inequality. The European Code of Social Security and its Protocol (1964) and the Revised Code of Social Security (1990) guarantee a mini- mum level of protection in the nine traditional branches of social security, by speci- fying the risks and the benefits to be provided.4 The control procedures are also based on national reports. Why are minimum standards essential in the fight against poverty and inequality? Because: · Minimum standards promote social stability, ensuring a less fragmented society, with lower levels of social conflict and smaller social disparities. These instru- ments help the vulnerable sectors of society to manage or mitigate social risks derived from economic instability, avoiding huge income gaps between social groups. · Minimum social standards improve productivity. When people are well-protected, they do not have to live in fear and insecurity--and their long-term productivity and job continuity greatly benefit from their situation, providing them with steady income. Improving Access to Social Rights: Removing the Barriers In the current environment of rapid economic and social change, the need for protection with regard to employment, health, housing, income, welfare, and education are para- mount. The arguments for social rights are as strong, if not stronger, than ever. But there is compelling evidence, from the work on access to social rights undertaken under the auspices of the council that formal entitlement to a social right is no guar- antee that a right will or can be realized. Bringing together the results of research and investigations carried out in the fields of access to employment, social protection, housing, health care, and education, the COE report on access to social rights in Europe, published in 2002, analyzes the numerous obstacles in the process of access to social rights. The report recommends cross-sectoral policy guidelines, which, if implemented, should help facilitate access to these rights, especially for people in vulnerable situations.5 Among the barriers preventing access to social rights, the most important are: · The legal framework. The structure and form of the provisions that legally codify and organize social rights are crucial. The kinds of obstacles that arise in this 46 | GABRIELLA BATTAINI-DRAGONI regard not only create difficulties for the realization of rights but also exclude some people. · Inadequate resources. The resources that are necessary for service providers in the realization of social rights include financial resources, human resources, and per- sonal capacities. · Poor information and communication. People continue to be under-informed about their entitlements, and that lack of information may lead to the delay or loss of benefits. In addition, the views of rights holders do not always sufficiently inform service providers. · Psychological and sociocultural obstacles. Among service providers, barriers are caused by factors such as negative attitudes toward (or stigmatization of) the beneficiaries--and the failure on the part of officials and others to appreciate that applicants or potential applicants may subscribe to a particular subculture that is different from the mainstream culture. · Failure to give due attention to the particular situation and needs of vulnerable groups and regions. Vulnerable groups within the population often lack the resources to effectively claim their rights; and some regions or localities suffer from multiple deprivations and inadequate investment. Despite these obstacles, most of the architecture necessary for the achievement of social rights is already in place in Europe. While there are some gaps, the problems that exist derive in many cases from shortcomings in the operation or functioning of existing provisions or the specification of rights. This is why the COE proposes the following set of guidelines for eliminating barriers to social rights. Some key recom- mendations and policies include: · The personalized approach. Each employment and social services client should be treated as an individual in need of an integrated package of benefits, training, and other forms of assistance that are tailor-made to their particular needs. · The "activation" approach. People should be given the help and encouragement they need in order to build new futures for themselves, in which they will become active subjects of their own lives rather than passive recipients of benefits and services. · A revolution in communication. Public agencies must find more innovative ways of informing people about their rights, by listening to their ideas and proposals and helping them to navigate the labyrinths of bureaucracy. This should be seen as a fundamental duty of the public services in a democracy. · The partnership approach. Government bodies cannot achieve their objectives by working in isolation. If they are to be effective in fighting poverty and social exclu- sion, the various public agencies must work with one another--and with social partners, NGOs, and private sector providers. Social Rights Are the "Gateway" to Equitable, Inclusive Societies The expansion of social rights that has taken place in Europe under the auspices of both PROMOTING EQUITY THROUGH SOCIAL COHESION | 47 the COE and the European Union (EU) can be read as evidence of the continuing sig- nificance of a rights-based approach to equity. Economies cannot function in a social vacuum, and the strengthening of social rights serves, among others, to reduce the risk of future social and political disruption. Viewed in this light, social rights serve to reduce social tensions and contribute to economic development. They also work because sustainable societies and sustainable environments are a necessary condition for sustainable economic development--and for sustainable democracies. In these and other ways, social rights have added social value. Social Rights and Social Responsibility The capacity of a society to ensure the welfare of all must be seen as a responsibility for all sectors of society.6 During the 20th century it was often considered that, apart from the role of the family and other traditional social relationships, welfare was the responsibility of the state (the "wel- fare state"), while business was responsible for economic development. The eco- nomic problems of recent decades and the globalization of the economy have shown that this model is no longer adequate and that the welfare of all in a society must become a goal shared by all social actors, the state, businesses, and individuals (a "welfare society"). This new situation does not imply a withdrawal of the state but rather a redefinition and clarification of its role. A society is cohesive when people accept responsibility for one another. Individual rights are best protected in societies where people feel a shared responsibility for the rights and welfare of all. Equity in Monitoring: The COE Methodological Guide on Social Indicators The COE is developing operational tools intended to monitor progress in social cohe- sion. In preparing a methodological guide to the use of social cohesion indicators published in 2005 (COE 2005), work was carried out to test this methodology in selected countries and regions. The social cohesion indicators differ greatly from the indicators employed by other organizations such as the EU and the Organisation for Economic Co-operation and Development (OECD). The OECD utilizes a more "neg- ative definition" of social cohesion that focuses on the concept of social exclusion and poverty, and therefore measures "alert indicators" of the dysfunction of society. The "alarm" indicators in the COE guide, aimed at identifying the symptoms of the "pathologies of society," also serve as a common frame of reference for an even more heterogeneous group of countries. Several indicators proposed by the OECD to pro- vide a comparative overview of developed societies are of this type.7, 8 Based on the "positive and functional definition" of social cohesion presented ear- lier, the council has identified 20 main indicators that focus on four main aspects of social life: equity, dignity, participation, and autonomy as well as on four main actors (public authorities, enterprises, public, families) and two additional basic com- ponents of social cohesion (social bonds, level of violence). So a set of indicators is proposed specifically to monitor equity in areas such as income, health, social secu- rity, housing, and employment (see table 1 for this list of indicators). TABLE 1. Twenty Key Indicators for Assessing the Components of Social Cohesion 48 Processes to be measured Main indicators Other possible indicators 1. Situations Equity in the 1. Equity in income Inequality of income distribution enjoyment of Long-term unemployment rate rights 2. Equity of access to employment 3. Equity in health Life expectancy at birth Average non-reimbursed proportion of the cost of consulting a general practitioner Proportion of persons not covered by social security 4. Equity in housing Western Europe: Eastern and Central Proportion of Europe: Population homeless in the without access to population quality housing Dignity/ 5. Gender: equal opportunities Assumption of senior responsibilities Involuntary part-time work recognition for women 6. Cultural and ethnic origin; Ethnic or religious ghettos equal opportunities 7. Age: dignity of elderly people Elderly people who receive a minimum Elderly people without a contributory old age allowance pension Gap between the minimum amount of social assistance and the poverty threshold Autonomy/ 8. Income sufficiency Western Europe: Eastern and Central Percentage of the population who occupational, Proportion of Europe: Proportion receive the minimum guaranteed family, and overindebted of households income personal households below the poverty development threshold in spite of both parents working 9. Educational sufficiency Dropout rate at the minimum school- Children who work before the leaving age without qualifications statutory school-leaving age 10. Social mobility Ability of children from disadvantaged social backgrounds to succeed at school Participation/ 11. Participation in elections Participation in elections by 18-34 year- commitment olds 12. Commitment of local authorities Proportion of the budget reserved for social issues 2. Action 13. Commitment of the corporate Workers with disabilities in the public and Contribution of companies to the (shared sector private sector financing of training responsi- bility) Fixed-term/permanent employment 14. Citizen commitment Proportion of jobs in the voluntary sector 15. Family commitment Elderly people living with their families 16. Confidence Confidence in public institutions Proportion of abandoned children Corruption index 17. Loss of social bonds Suicide rate 3. Basic 18. Shared knowledge Awareness of human rights and of the compo- right to justice nents of 19. Perception/satisfaction Subjective perception of health life 20. Tolerance and respect Murder rate Proportion of convicted persons or of prisoners per 1,000 inhabitants Feeling of security Source: COE (2005). 49 50 | GABRIELLA BATTAINI-DRAGONI Each aspect includes several dozen sub-indicators that can be used to analyze in greater detail the situation of a particular country. These 20 indicators are currently being used and tested with a view to demonstrating the value of social cohesion indi- cators as a tool for policymaking at all levels of government. It has also been sug- gested that it might be possible to develop a social cohesion index to measure the degree of social cohesion within the organization's 46 countries and adapt the strat- egy to their specific needs. Any such index would have to take into account the dif- ferent levels of socioeconomic development of the 46 member states of the COE, and so would focus on evolutions and trends. Conclusions At a time when the future of the European social model is under widespread discus- sion, the COE's social cohesion strategy represents a convincing attempt, endorsed by the governments of all European countries, to show how social equity and eco- nomic development can be combined, both in theory and in practice. Notes 1. There are currently 46 member states of the Council of Europe (COE): Albania, Andorra, Armenia, Austria, Azerbaijan, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, the Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Macedonia (former Yugoslav Republic of), Moldova, Monaco, the Netherlands, Norway, Poland, Portugal, Romania, the Russian Federation, San Marino, Serbia and Montenegro, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, and the United Kingdom. 2. Excerpted from the Revised Strategy for Social Cohesion (see CDCS, COE 2004). 3. The Revised Social Charter has been signed or ratified by the following countries (see appendix 2 of the charter): Albania, Andorra, Armenia, Austria, Azerbaijan, Belgium, Bosnia and Herzegovina, Bulgaria, Cyprus, the Czech Republic, Denmark, Estonia, Fin- land, France, Georgia, Greece, Hungary, Iceland, Ireland, Italy, Lithuania, Luxembourg, Malta, Moldova, Monaco, the Netherlands, Norway, Portugal, Romania, Russia, San Marino, Serbia and Montenegro, the Republic of Slovakia, Slovenia, Spain, Sweden, Turkey, Ukraine, and the United Kingdom. 4. The European Code of Social Security has been signed or ratified by the following COE member states (see appendix 2 of the charter): Austria, Belgium, Cyprus, the Czech Republic, Denmark, Estonia, France, Germany, Greece, Ireland, Italy, Latvia, Luxem- bourg, Moldova, the Netherlands, Norway, Portugal, Romania, Slovenia, Spain, Sweden, Switzerland, Turkey, and the United Kingdom. 5. See COE (2002). 6. From paragraph 17 of the Revised Strategy for Social Cohesion (CDCS 2004). 7. These are the European Union's social cohesion indicators: 1. low income rate after trans- fers, with low-income threshold set at 60 percent of median income (with breakdowns by gender, age, most frequent activity status, household type, and tenure status as examples PROMOTING EQUITY THROUGH SOCIAL COHESION | 51 of the values for typical households); 2. distribution of income (income/ quintile ratio); 3. persistence of low income; 4. median low-income gap; 5. regional cohesion; 6. long- term unemployment rate; 7. people living in jobless households; 8. early school leavers not in further education or training; 9. life expectancy at birth; 10. self-perceived health sta- tus (European Union Social Protection Committee 2001). 8. These are the social cohesion indicators proposed by the Organisation for Economic Co- operation and Development (OECD): 1. divorce rate; 2. fertility rate; 3. percentage of lone-parent families; 4. membership of local associations; 5. rate of participation in elec- tions; 6. population born abroad; 7. mixed marriages; 8. asylum-seekers; 9. suicide rate; 10. age of women at birth of first child; 11. working women with children; 12. crime rate; 13. mortality rate due to drug use; 14. work stoppages; 15. persons detained at penal establishments; 16. acquisition of nationality (see OECD, DEELSA/ELSA 1999). References COE (Council of Europe). 1998. Opportunity and Risk: Trends of Social Exclusion in Europe. Strasbourg, France: Council of Europe Publishing. ------. 1999. "Social Protection in the European Social Charter." Human Rights Social Char- ter monographs, number 7. Strasbourg, France: Council of Europe Publishing. ------. 1999. European Convention on Social Security and Supplementary Agreement for the Application of the European Convention on Social Security (Strasbourg 14.12.1972), European Treaty Series, number 78. Strasbourg, France: Council of Europe Publishing. ------. 2000. European Code of Social Security and Protocol to the European Social Security (Strasbourg, 16.4.1964), European Treaty Series, number 48. Strasbourg, France: Council of Europe Publishing. ------. 2000. European Social Charter and Protocol. Strasbourg, France: Council of Europe Publishing. ------. 2000. European Committee for Social Cohesion Strategy for Social Cohesion, CDCS, number 43. Strasbourg, France: Council of Europe Publishing. ------. 2002. Access to Social Rights in Europe. Strasbourg, France: Council of Europe Publishing. ------.2004. "Revised Strategy for Social Cohesion." COE CDCS. ------. 2005. Concerted Development of Social Cohesion Indicators: Methodological Guide to the Development of Social Indicators. Strasbourg, France: Council of Europe Publishing. European Union. 2001. "Report on Indicators in the Field of Poverty and Social Exclusion." Social Protection Committee. European Union, Brussels, Belgium. OECD (Organisation for Economic Co-operation and Development). 1999. "Social Indica- tors: A Proposed Framework and Structure." DEELSA/ELSA (99)11. Paris, France. Equality of Opportunity and Other Equity Principles in the Context of Developing Countries DENIS COGNEAU A personal reading of some theories of social justice seems appropriate at a time when the issue of equality (or equity) appears to be back on the "development agenda," so that is a focus of this paper. An overview of empirical and theoretical knowledge about the persistence of within-country inequality over time is presented, which underlines the central role played by the intergenerational transmission of resources and behaviors. The acknowledgment of this role is part of the reason why today, the term equity tends to be most often associated with the normative princi- ple of equality of opportunity. A brief summary of the equality of opportunity stand- point is followed by a review of the two main criticisms of the approach as addressed in the economic literature: the right-wing meritocratic criticism on the one hand, and the left-wing egalitarian criticism on the other. In turn, these internal criticisms are supplemented with a sociological (or anthropological) point of view that advocates for a more pluralist definition of justice and acknowledges the importance of the competition between elites for legitimacy. This paper argues that despite its indubitable potency (even for issues such as international inequalities between countries), the equality of opportunity principle is incomplete--and that some meritocratic principles and some equalization of out- comes should come into play when addressing social justice in a given society. More- over, a socially relevant conception of justice should take into account cultural vari- ations in the definition of fairness: A universalist definition of justice is better preserved when the issue of tyranny and separation of powers is considered at both the social and political levels. In guise of conclusion are comments on the benefits of extending the equality of opportunity principle to the issues of international equity and of foreign aid allocation. Denis Cogneau is a researcher for the Institut de recherche pour le développement (IRD) at Développement, Institutions et Analyses de Long terme (DIAL) in Paris. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 53 54 | DENIS COGNEAU Persistence of Inequality and Intergenerational Mobility Available international datasets reveal not only tremendous average income differ- ences but also large differences in income inequality levels between countries of the world, with Gini indexes ranging from as high as 0.65 to as low as 0.25.1 The coun- tries of Latin America and sub-Saharan Africa appear to be the most unequal in the world in that respect. Contrary to the wisdom of Simon Kuznets, the poorest coun- tries are not less unequal, at least now (if they have ever been). These datasets also show that within-country inequality levels are fairly stationary over time, with some exceptions linked to historical changes in economic regimes: In the recent period, the liberalization of former socialist economies (in Eastern Europe and China) and the Reagan and Thatcher "shocks" in the United States and the U.K. have been exten- sively documented. Among former colonies, no "legal system" is more unequal than others,2 but a higher pre-colonial level of development, as reflected for instance by population density, seems to have an equalizing effect (Cogneau and Guénard 2003). From this fact, one may conjecture that the antiquity of state institutions explains a great deal of the intercontinental inequality differentials between Latin America and Africa on one side, and Asia on the other side. For developed countries, stability also prevails for intergenerational occupational mobility (see the 1991 Erikson and Goldthorpe study, with its evocative title "The Constant Flux").3 Few indicators for some of the developing countries in Latin America (Behrman, Gavaria, and Székely 2001) and Africa (Bossuroy et al., 2005) do not reject stability over time as a first approximation, except in the case of strong shocks like civil wars and forced popu- lation displacements. The literature on poverty traps and microeconomic income convergence between the rich and the poor brings mixed and rather inconclusive evi- dence (Fields and others 2002), due to measurement errors and econometric identi- fication problems, and because of the brief panel periods under review. Much hope has been placed in the expansion of education as a long-run means of alleviating poverty and inequality. In contrast with income inequality, available evi- dence shows a large worldwide reduction of within-country and between-country educational inequalities--as measured, for instance, by the number of years in school or by basic skills like literacy. As years of schooling are bounded, any large expan- sion of schooling generates an equalization that has been observed since the 1950s in almost every country. However, schooling may be more and more qualitatively dif- ferentiated, whether in terms of transmitted skills or of signalling properties. Even if there is some evidence that this more equal distribution of schooling contributed to a reduction of inequalities within countries, it is also the case that returns to educa- tion have sometimes increased--or that other factors have counteracted this effect, such as an increased selection on unobserved skills or absolute advantages (see, for example, Bourguignon, Fournier, and Gurgand 2001 on Taiwan; and Lam 1999 on Brazil and South Africa). Furthermore, the intergenerational transmission of educa- tion has remained high and, in most cases, shown a great deal of stability once struc- tural moves are discounted: The ratio of the probability to reach a given education level between a pupil with uneducated parents and a pupil with educated parents has not changed.4 EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 55 Most theories of the long-term persistence of economic inequalities rely on the intergenerational transmission of economic and social status through the imperfec- tion of human and physical capital markets, social capital, or spatial segregation. These either prevent individuals in disadvantaged groups from climbing the social ladder, or prevent children in advantaged groups from falling down (Piketty 2000 and other works; Loury 1981). Poverty reduction itself is increasingly conceived of as a dynamic process centered on the "acceleration of individual exits from poverty," which involves the separation of transient short-term components and permanent intergenerational issues (see, for example, Cogneau 2003 on this point). Since it is acknowledged that only a small proportion of intergenerational correlation in income may be explained by the direct transmission of genetic or acquired skill endowments, new theories also stress the transmission of preferences such as school and work habits, self-confidence, competitive aggressiveness (Bowles and Gintis 2002 and other works), and adaptative preferences. Beyond parental preferences and endowments, the reference group average behavior and characteristics seem to carry a great deal of weight in the explanation, so that policies encouraging social mixing and fighting against spatial or social discrimination and/or segregation are increas- ingly taken into consideration (see, for example, Durlauf 1996). Holistic changes in values, as in the cases of feminist, anti-racist, or anti-caste movements, no doubt also play an important role when confronting such phenomena as self-reinforcing statis- tical discrimination. Thus, historical evidence shows that it is not easy to improve intergenerational mobility. Is the equalization of opportunities between groups of different social ori- gin out of reach because of difficult yet unclear difficulties that are deeply rooted in the basement of societies? Or is it because of the lack of measurement tools needed to identify small, significant (although slow) changes as opposed to larger or more revolutionary changes? Or, is it simply that not enough political effort has been ded- icated to the task? Perhaps all three arguments are part of the answer. Conceptualizing and Measuring Intergenerational Equity Starting from John Rawls's impressive book (1971), the sometimes called "post- welfarist" theories of distributive justice have criticized both the utilitarist/welfarist theories and the raw egalitarianism from the standpoint of freedom and personal responsibility. In order to assess the fairness of social arrangements, these theories suggest the need to focus on the processes through which free individuals might achieve their own conception of good. Individuals were first described as transform- ing resources or "primary goods" into utility or welfare, and to be held responsible for their use of resources and their goals (Rawls 1971; Dworkin 1981). The problem of equity was described as a problem of equalization of primary resources. It was then acknowledged, in a second step, that individuals might also differ in their capac- ity to transform resources into utility, and that they should only be held responsible for seizing opportunities: the problem of equity then shifted to the equalization of capabilities (Sen 1992) or opportunities (Roemer 1998). However, despite the 56 | DENIS COGNEAU divergences between "resource egalitarians" and "opportunity egalitarians," new egalitarians agree on the equalization of a "midfare" (Cohen 1993), which lies between initial opportunity sets and final achievements. They also agree on making a distinction between illegitimate inequalities, most often traced back to the social origin of individuals (gender, parental wealth and education, status, or ethnic group of origin), and morally irrelevant inequalities that result from the free play of indi- vidual responsibility. In the field of responsibility-sensitive justice, equality of capa- bilities or of opportunity seems to have become more prominent than equality of resources, at least among economists. Outside of the academic world, the word "equity" also seems to be most often used in that sense, especially in the language of the Western "Third Way" Social Democrats, whose political programs have put for- ward equality of opportunity principles while promoting a transition from status- based systems of redistribution to more individualized systems. John Roemer has constructed a rather simple formal framework for defining equality of opportunity policies (abbreviated as EOp in his works) and has proposed a way of measuring the extent of EOp achieved in some "playing field" of a given society. Unlike some welfarist constructs, Roemer's framework is only partial; it does not jointly address multiple dimensions of inequality and leaves aside most efficiency issues. It does, however, highlight and make clear the important ethical and political choices that are at stake if one tries to design policies that equalize opportunities. There is first the choice of what Roemer calls the "advantage variable," which can also be called outcome or achievement, and which may be utility, income, education, risk of death, and so forth. The problems raised here are not new in that they are fun- damentally the same as those in the previous utilitarist/welfarist frameworks-- mainly those related to reduction of multidimensionality within a utility index and interpersonal comparisons of utilities (see Bourguignon 2002). In contrast, the sec- ond choice lies at the heart of the equality of opportunity approach. It requires the definition of what Roemer calls "circumstances," which can also be called handi- caps. These include all the morally irrelevant variables that are admittedly outside the reach of individual responsibility. Individuals are aggregated by "types of circum- stances" between which opportunities have to be equalized. All other factors of the advantage variable are assumed to pertain to the sphere of responsibility, and hence are not to be compensated. Roemer called these factors "effort," rather improperly, in fact as they stand, more prosaically, for all factors that are uncorrelated with the circumstances that influence an individual's rank or quantile in the outcome distri- bution for his/her "type." Roemer proposes that for each quantile of "effort," the achievement of the worse-off type (that is, the minimum across all types) be consid- ered, and to then measure equality of opportunity by a sum of all quantiles minima (see box 1). Hence noncircumstantial factors (or "effort") are only implicitly mea- sured as a residual in the association between the outcome variable and the circum- stance variable.5 This is what makes the choice of the circumstance variables so important. Roemer claims that this choice must be made by society through some kind of ethical and political process. The right-wing set of morally irrelevant circum- stances will typically be narrower than the left-wing set. Social origin variables like EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 57 gender, race, parental education or wealth, region of birth, and ascribed caste or eth- nic group often appear as the most consensual choice--even if there is disagreement about the "true" relative weight of circumstances and "effort" in the production of achievements.6 This kind of choice establishes a direct link between intergenerational transmission and equality of opportunity. BOX 1. Equality of Opportunity Indexes The Roemer index reads as follows: ROE = 1/Q. qmint {yt,q}, Equation (1) where t = 1, . . . T is the type of circumstance variable, and yt,q is the qth decile level of the "advantage variable" within type t (q = 1, . . . ,Q). Roemer (1998) proposes giving equal weights to each quantile minima in order to main- tain the "effort neutrality" of equality of opportunity principles (see below for two critiques: about efficiency and incentives to effort on the one hand, and about the nondeserving poor on the other hand). Following Roemer's seminal work, other contributions have proposed alter- native indexes within the same theoretical framework; for instance, inequalities between the expected (average) outcomes of various "types" (Van de Gaer, Schokkaert, and Martinez 2001), instead of the average of inequalities between quantile outcomes for "types," which may lead to distinct orderings when the mobility matrixes are not monotone in the Shorrocks sense. The Van de Gaer index reads as follows: VdG = mint {yt}, Equation (2) where yt is the mean level of the advantage variable for circumstances' type t. Note that these maximin indexes are not normalized by the mean of the advantage variable. Therefore, they are more welfare indexes than pure inequal- ity indexes: If the advantage variable grows at the same rate for each individual, the indexes in equations (1) and (2) also grow at this rate (homogeneity of degree one). The choice of a maximin criterion can be replaced by any inequality index or Lorenz dominance: ROE = 1/Q. q t I {Yt,q} Equation (3) or: VdG = It {yt}, Equation (4) where It is any between-types inequality index with good properties (such as the Gini index, for example). In contrast with the former maximin indexes (1) and (2), these indexes are pure inequality (of opportunity) indexes (homogeneity of degree zero). 58 | DENIS COGNEAU Some Examples of Inequality of Opportunity between Social Origins in Brazil and Africa Inequalities in the labor earnings of Brazilian males from 40 to 49 years of age remained remarkably stable at a high level during the 20-year period 1976­96 (see table 1). The inequalities even increased a little. EOp indexes based on social origins also remained fairly stable and followed the same evolution of total inequalities. According to Roemer's maximin index, the earnings of the worse-off types even decreased. In the five countries of Africa examined in table 2, inequalities of consumption expenditures between households again reach a fairly high level, with Gini indexes between 0.4 and 0.5. Here the measurement of social origin is less precise, as only 18 types of social origins of the household head have been distinguished (against 128 in the case of Brazil). In four countries, the Gini EOp indexes lie between 0.1 and 0.15. Madagascar is the exception, with the EOp index reaching 0.3. Migra- tions and class mobility are, indeed, much lower in Madagascar than in the other African countries. TABLE 1. Inequalities of Labor Earnings Between Social Origins in Brazil 1976 1988 1996 Roemer's maximin criterion* 1.297 1.048 1.223 Average of Gini index within deciles** 0.341 0.375 0.343 Gini of types' average earnings*** 0.362 0.392 0.352 Overall Gini index of hourly earnings 0.570 0.623 0.599 Source: PNAD (Pesquisa Nacional por Amostra de Domicilios) surveys; see Cogneau and Gignoux (2005). Coverage: 40-49 year-old males. Method: 128 types of social origin constructed from race (2); region of birth (4); father's education (4) and father's occupation (4); decile regressions estimates; hourly earnings in 2002 reais (Brazilian currency). Notes: * see equation (1) in box 1; ** see equation (3) in box 1; *** in keeping with Van de Gaer; see equation (4) in box 1. TABLE 2. Inequalities of Household Income Per Capita Between Social Origins in Africa Côte d'Ivoire Ghana Guinea Madagascar Uganda (1985-8) (1987-8) (1993-4) (1993-4) (1992-3) Gini of types' average earnings* 0.120 0.147 0.109 0.297 0.135 Overall Gini index 0.409 0.452 0.457 0.503 0.448 Source: LSMS (Living Standards Measurement Study) surveys, various years; author's computations. Coverage: Households where the household head is 40-49 years old and is born in the country. Method: 18 types of social origin of the household head constructed from region of birth (2); father's education (3) and father's occupation (3); decile regressions estimates; consumption expenditures per capita Notes: * in keeping with Van de Gaer; see equation (4) in box 1. EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 59 The Meritocratic Critique of Equality of Opportunity: Efficiency Concerns Even from individualistic and meritocratic standpoints, equality of opportunity has been criticized on the basis of the efficiency losses it generates for the whole society. Opportunity equalization indeed addresses the effect of all causal factors of the out- come variable which are correlated, even if not caused by the circumstance variables. For example, genetically transmitted innate abilities that would partly contribute to the achievements of both parents and their children are to be compensated when intergenerational equality of opportunity is targeted. Likewise, for example, the propensity to work hard in school that educated parents would be more able to transmit to their children is considered in this framework, where effort is only mea- sured "relatively" (by quantiles, not by grades), as morally irrelevant. When talent and effort are considered in their absolute meaning, less "talented" and/or less "deserving" pupils will then benefit more than pupils from richer backgrounds who show the same "relative effort." With equalization of opportunities, there may then be fewer "talented" and/or less "deserving" (again absolutely speaking) in the higher grades than before. Moreover, knowing that, advantaged children may be discour- aged, or disadvantaged children may expend less effort (whether help is a positive or negative incentive of effort). Roemer makes a strong point against effort fetishism, as a reminder that effort itself is not the target and that average effort is not to be max- imized. However, he proposes to delimit the extent of equalization of opportunity by weighing it against average outcome (not effort) through a social welfare function in line with the welfarist tradition. And he argues, moreover, that his equality of oppor- tunity principle is better suited to the initial building of resources like education and health. After an opportunity-equitable distribution of resources is met, then a meri- tocratic efficiency-maximizing principle may enter into play in the allocation of job positions and in the determination of earnings schedules. This is meant to avoid the frequently alleged perverse effects of some forms of affirmative actions. (In the exam- ple that Roemer presents, the access to grades in surgery should be ruled by equality of opportunity principles based on social origins, while the selection of the best sur- geons should follow a meritocratic rule based on both grades and talent.) The Egalitarian Critique of Equality of Opportunity: A Return to Rawls A purist equality of opportunity approach would allow large inequalities within types of circumstances, as they are not considered as morally irrelevant. Among egal- itarians, this property is a matter of worry.7 Roemer's framework already mixes a pure equality of opportunity approach with a basic need approach--through the maximin criterion, which denotes a care for the absolute welfare level of the worse- off, even of those who exert the lowest level of effort and might otherwise be seen as the "non-deserving" poor. With a very large number of types of circumstances, Roe- mer's rule in fact translates into a pure maximin Rawls's rule, as within-type inequal- ity reaches zero. With such a rule, Roemer's analysis tends to promote a large 60 | DENIS COGNEAU amount of transfers in favor of the most disadvantaged types.8 However, it may most benefit the better-off among the most disadvantaged type, because of the sum of min- ima over "effort" quantiles.9 The "non-deserving poor" may be left on the side of the road. Because of this worry, some equalization of achievements or outcomes can- not be definitively excluded from egalitarian thinking. The notion of individual responsibility indeed loses its meaning when poverty is extreme. Taking the famous example of Sen, when the poor no longer have the choice to fast but only to starve, only a tyrannical society can treat them as "nondeserving." Intertemporal issues also reintroduce equalization of outcomes. First, the intergenerational transmission process is such that parental achievements determine the initial resources and oppor- tunities available to their children. Second, it is defendable that some "forgiveness" (Fleurbaey 1998 and other works) should be given to people who have failed to seize opportunities, in the form of a second chance principle and, for example, through socialized insurance. Starting from these ethical considerations and other more tech- nical problems of feasibility and consistency raised by the equality of opportunity approach, Fleurbaey (1995 and 1998) ends up proposing a return to Rawls. "Soci- ety should decide on a bundle of functionings . . . it considers as important enough to be taken in charge collectively," like good health, "and try to equalize (maximin) an index . . . [of functionings] across individuals" (Fleurbaey 1998, p. 229). Func- tionings that are not selected are not under societal supervision and, even if not under individual control, fall only under private responsibility (or "accountability" in that sense). In contrast with Rawls, Fleurbaey recommends replacing "primary goods" by Sen's "functionings" in order to avoid the problem of unjust differential capacities in transforming resources in outcomes.10 Such an approach of equity puts the stance on the political selection (and aggregation) of socially relevant functionings, in place of the selection (and aggregation) of the morally irrelevant set of circumstances. "Responsibility over factors" is given up--and replaced by accountability over out- comes that are outside of the list of primary functionings to be equalized. The con- sideration of social origins (as morally irrelevant factors) is no longer central to nor- mative equity measurement, even if it remains important in the positive analysis of observed inequality in primary functionings.11 Cultural Variations in Fairness: Equity as Pluralism Michael Walzer (1981) puts forward a universalistic theory of justice as pluralism, as an explicit alternative to that of John Rawls on justice as fairness. His argument is based on historical and anthropological examples rather than on a procedural demonstration. While Rawls's theory is inspired by Rousseau's social contract, Walzer is instead inspired by Pascal's definition of tyranny. His criticism of Rawls starts from the observation of the anthropological multiplicity and diversity of Rawls's primary goods, a problem also addressed by Sen in his theory of capabilities. In Walzer's view, these primary goods cannot be properly aggregated for the basic reason that they relate to distinct, socially determined, distribution principles. Dis- EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 61 tributive criteria for a good follows from its social meaning, the functions it serves in society. Walzer calls the association of a primary good and its specific principle a "sphere of justice." Justice as pluralism consists of ensuring that one sphere does not predominate over others, with legitimate possession of one primary good ensuring illegitimate access to the others. American society, for example, could be said to be characterized by the tyranny of money and its principle of distribution, which is com- petition, while Indian society could be said to be characterized by the tyranny of caste and its principle of distribution, which is purity. One might also say that French soci- ety is characterized by a tyranny of education and merit in school. The latter exam- ples reveal the advantage of a theory that is reflective and which allows for the incor- poration of a certain form of cultural relativism, without necessarily forgoing universalism. Walzer's proposals can be expressed as follows: 1. There exist several spheres of justice; that is, several types of primary goods which have different legitimate principles of distribution.12 2. The primary form of injustice consists in one sphere encroaching on all others, which borders on tyranny. Tyranny does not recognize the plurality of spheres of justice; all primary goods are subsumed under a primary imperative (be it noble birth, wealth, education, party membership, or geographical origin), while, if necessary, tyranny is justified as the defense of a "common good" (the welfare and security of citizens, the purity of race, the power of the nation, the glory of science, or the carrying out of God's work). In a society that functions tyrannically, there is a particular possibility that certain cat- egories of people--the poor, the scheduled castes, or the uneducated--might be denied access to even the lowest levels of welfare and security because they lack a pre- dominant good. This conception of equity as pluralism still agrees upon the fact that achievement in some sphere should only be minimally influenced by affluence in some others. There should be some equality of opportunity in each sphere, where morally irrele- vant circumstances are here the level of achievements in other spheres. In the case of education, ability and desire to learn are considered as relevant grounds for differen- tial achievements, with parental income or party membership, caste or status consid- ered irrelevant factors. The selection of morally relevant and irrelevant factors must be drawn from a socially and culturally relevant mapping of spheres of justice. Walzer's conception rejoins some approaches of the "political economy and sociol- ogy" of redistribution and of progresses towards equity. For instance, Bourdieu (1989) and Elias (1939) both emphasize the role of conflicts of interests between elites that are characterized by different structures of resources (the highly educated, the capital owners, the landlords, and so forth). According to Bourdieu, when this conflict translates in a fight for political legitimacy, the voice and the claims of dom- inated classes can be heard, and "true universalism" can achieve some progress (see also Dezalay and Garth 2002, on the Latin American case). As Walzer argues, social (not only legal) separation of powers is a prerequisite of equity as pluralism, in order to bar tyranny. 62 | DENIS COGNEAU Apart from this general reflection there is, unfortunately, limited knowledge on the cultural variation in the meanings of justice apart from some experimental evi- dence: On outcomes of equity-sensitive simple games, see Heinrich and others (2001); or for survey evidence, see Schokkaert and Devooght (1999), and Diener and Suh (1999), which indeed show large intercultural differences but are difficult to interpret. Walzer's relativism serves as a reminder that the conception of freedom and responsibility which inspire the post-welfarist theory of justice are better suited to philosophies of freedom and to "societies of individuals" (Elias 1939), where it is considered that all human beings should be allowed to "choose" as freely as possi- ble their destiny, or to build and follow their own conception of welfare, without being ascribed to a social position or function according to their gender, race, or social origin. In that sense, the progress of individualism may be, at least for a while, contradictory to some notion of social cohesion in a society where roles and statuses are strictly ascribed. The egalitarian critique described earlier also claims that equal- izing opportunities is no panacea. Anomy, suicide, new mental illnesses, stress, and crime may result from a more individualistic and "competitive" society, where per- sons have more autonomy but are also made more "accountable" for their "failures" and errors as much as for their "successes." In an individualistic society, equality of opportunity as an abstract principle enforced by the welfare state comes to counter- balance the hardness of meritocratic competition. But it does not compensate for the loss of the protection once offered by communities of origin. It does not alleviate the sufferings of the unsuccessful--and may even add to the disadvantage of a subordi- nate position, the moral torture of guilt. Three Dimensions of Equity From what precedes, one may risk arguing that equity in a real society probably requires a complementary mix of equality of opportunity, of meritocratic-efficient allocation, and of equalization of primary functionings or basic achievements. It was shown that none of those three dimensions of equity was exempt from criticism when taken in isolation, but that each of them may be better suited to some domain of social life (see figure 1). Equality of opportunity principles are probably better suited to the domain of resource building, particularly during childhood and adolescence when persons are still dependent and their individual identities are largely undetermined. Meritocratic allocation of tastes and skills between jobs, social tasks, and power positions may then enter into play; for example, through market-based or other efficient mechanisms, with enough antidiscrimination and antinepotism legal safeguards, in order to achieve a high level of aggregate outputs and a high quality of services (as in Roemer's "sur- geons" example mentioned earlier). And third, some equalization of primary func- tionings should intervene with respect to achievements as fundamental as health, edu- cation, wealth, social inclusion, and voice--in order to achieve a society that would guarantee all individuals equal respect and the equal ability to provide for their dependents, and guarantee forgiveness for opportunities not seized. The three mecha- EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 63 FIGURE 1. Three Dimensions of Equity Resource Building (Child rearing, nutrition, education . . .) EQUALITY OF OPPORTUNITY (Targeted transfers, aff. action . . .) Allocation of skills and tastes (Jobs, social tasks, power positions . . .) MERITOCRATIC RULES (Market or other mechanisms . . .) Primary functionings (health, education, inclusion, voice . . .) EQUALIZATION OF OUTCOMES (mutual care, redistribution . . .) nisms should, therefore, reinforce each other: Equality of opportunity should make more acceptable the meritocratic selection, which, in turn, should help in providing a high quality of services in resource building (child rearing, education, health, and so forth). A high level of aggregate social outcome should facilitate the provision of mutual care and the equalization of primary functionings. Redistribution of wealth and status among adults is, in turn, the unavoidable ingredient of equality of oppor- tunity among the next generation. All in all, the association of those three dimensions of equity, in specific proportions for each "sphere of justice," is more consistent with a pluralist society where resource-specific distributive rules are preserved, and where neither wealth, or education, or status would play a tyrannical role, in Walzer's sense. Reciprocally, some social separation of powers between elites is certainly a precondi- tion for equity to progress at the political level. This kind of separation of powers may 64 | DENIS COGNEAU be the principal limit to the progress of equity in many developing societies whose institutions have been marked by a colonial tyrannical rule. Conclusions: Extending Equality of Opportunity to International Equity Development differences between countries may be seen as inequality of prospects between individuals born in different countries, where the country of birth is a morally irrelevant variable that determines inequality of outcomes. In the interna- tional context, attacking inequality from the standpoint of inequality of opportuni- ties may be suited to the building of a fair community of nations. Persisting develop- ment differences between countries are indeed rooted in inequality of opportunities: Being born in an underdeveloped country means, on average, a large deprivation of opportunities; and this individual deprivation can be traced back to country-level morally irrelevant factors such as geographic handicaps and historical illegitimate interventions like colonialism or neo-colonialism. Even within-country inequalities have something to do with natural resource curses or fates and with colonial history, through institutional linkages (Engerman and Sokoloff 2002). The provision and allocation of international aid should therefore be inspired not only by efficiency considerations, but also by equity principles.13 Giving aid to countries where both institutions and growth are good enough indeed maximizes the efficiency of aid in reducing the number of poor people in the world (Collier and Dollar 2001); but it can also prove highly inequitable (Cogneau and Naudet 2004). Notes 1. These data are drawn from the well-known World Bank (Deininger and Squire) and United Nations (WIDER) databases. See http://www.worldbank.org/research/growth/dddeisqu.htm and http://www.wider.unu.edu/wiid/wiid.htm. 2. When measured by the national identity of the colonial power, as in the La Porta, Lopez- de-Silanes, Shleifer, and Vishny series of papers (see La Porta and others 1999). 3. In terms of levels but not dynamics, lower income inequalities do not seem to come with lower intergenerational mobility. For a comparison between Sweden and the United States, see Björklund and Jäntti 2001. 4. Health inequalities are even more limited by biological constraints (regarding, for instance, the duration of life). In that dimension, many developing countries have indeed been "catching up" quickly with the levels reached by developed countries over the past 50 years, so that (in contrast with income and wealth inequalities) within-country health inequalities represent a larger share of overall worldwide health inequalities; see Pradhan, Sahn, and Younger 2003 on nutrition-based health indicators for children. 5. This approach is rather consistent with the fact that individual effort is in real life very often hard to observe without bias. EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 65 6. Piketty (2000) has proposed a model where agents draw differential beliefs on these rela- tive weights from their own dynastic experience. Voting patterns would then reflect this distribution of beliefs. 7. Fleurbaey names the purist approach "conservative egalitarianism." 8. There is usually only one of these: the most disadvantaged social group, as in Rawls's dif- ference principle, when mobility matrixes have Shorrocks's monotonicity property. How- ever, not all matrixes are monotone. 9. Of course, the more restricted is the set of morally irrelevant circumstances, the lower the level of transfers required (the "right-liberal option," in Cohen's terminology; see earlier Cohen citation). 10. In his 1995 paper, Fleurbaey proposes six functionings for western societies: respect for the private sphere, health, education and information, wealth, collective decisionmaking power, and social integration (p. 53). 11. Social origins also remain useful variables to purge the effect of measurement errors or of other irrelevant unobservable factors (such as luck or transient components) in the mea- surement of inequality. 12. Among the primary goods, Walzer singles out security and welfare, money, education, and access to public office, which are regulated respectively by the distribution principles of equality, productive competition, merit, and devotion to public good. 13. International migration could also be seen as a powerful means for equalizing opportuni- ties between people born in different places of the world. General References Björklund, A., and M. Jäntti. 2001. "Intergenerational Income Mobility in Sweden Compared to the United States." American Economic Review 87 (5): 1009-18. Boudon, R. 1974. Education, Opportunity, and Social Inequality. New York: Wiley. Bourdieu, P. 1989. La noblesse d'Etat. Paris: Editions de Minuit. Bourguignon, F. 2002. "From Income to Endowments: The Difficult Task of Expanding the Income Poverty Paradigm." DELTA Working Paper 2003­03, Paris-Jourdan Sciences Economiques, Paris. http://www.delta.ens.fr/abstracts/wp200303.pdf. Bowles, S., and H. Gintis. 2002. "The Inheritance of Inequality." Journal of Economic Per- spectives 16 (3): 3-30. Cohen, G. A. 1993. "Equality of What? On Welfare, Goods, and Capabilities." In M.C. Nuss- baum and A. Sen, eds., The Quality of Life. New York: Oxford University Press. Durlauf, S. N. 1996. "A Theory of Persistent Inequality." Journal of Economic Growth 1(1): 75-93. Dworkin, R. 1981. "What is Equality? Part II: Equality of Resources." Philosophy and Pub- lic Affairs 10: 283-345. Elias, N. 1939. (revised 1991). The Society of Individuals. Oxford, UK: Blackwell. Erikson, R., and J. Goldthorpe. 1991. The Constant Flux: A Study of Class Mobility in Indus- trial Societies. Oxford, UK: Clarendon Press. Fleurbaey, M. 1995. "Equal Opportunity or Equal Social Outcome?" Economics and Philos- ophy 11: 25-55. 66 | DENIS COGNEAU ------. 1998. "Equality among Responsible Individuals." In J. F. Laslier, M. Fleurbaey, N. Gravel, and A. Trannoy, eds. Freedom in Economics. London: Routledge. Loury, G. 1981. "Intergenerational Transfers and the Distribution of Earnings." Economet- rica 49 (4): 843-67. Piketty, Th. 2000. "Theories of Intergenerational Mobility." In A.B. Atkinson, and F. Bour- guignon, eds. Handbook of Income Distribution, volume I. Amsterdam: North-Holland. Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Roemer, J. 1996. Theories of Distributive Justice. Cambridge, MA: Harvard University Press. ------. 1998. Equality of Opportunity. Cambridge, MA: Harvard University Press. Sen, A. 1992. Inequality Reexamined. Cambridge, MA: Harvard University Press. Van de Gaer, D., E. Schokkaert, and M. Martinez. 2001. "Three Meanings of Intergenera- tional Mobility." Economica 68 (272): 519-38. Walzer, M. 1981. Spheres of Justice. New York: Basic Books. References on Developing Countries Behrman J. R., A. Gaviria, and M. Székely. 2001. "Intergenerational Mobility in Latin Amer- ica." Economia 2 (1): 1-43. Bossuroy T., D. Cogneau, P. De Vreyer, C. Guénard, V. Hiller, P. Leite, S. Mesplé-Somps, L. Pasquier-Doumer and C. Torelli, 2005, "Inequality and Equity in Africa," Working paper AFD and DIAL, 30 pp. Bourguignon, F., M. Fournier, and M. Gurgand. 2001. "Fast Development with a Stable Dis- tribution of Income: Taiwan 1979-1994." Review of Income and Wealth 57 (2): 139-63. Cogneau, D. 2003. "Poverty, Inequality of Conditions, and Inequality of Opportunities: Chances and Risks of New Poverty Reduction Strategies." In J.-P. Cling, M. Razafin- drakoto, and F. Roubaud, eds. The New International Strategies for Poverty Reduction. London: Routledge. Cogneau, D., and J. Gignoux. 2005. "Earnings Inequalities and Educational Mobility in Brazil over Two Decades." DIAL Working Paper 2005/03, Développement, Institutions et Analyses de Long terme, Paris. http://www.dial.prd.fr/dial_publications/PDF/Doc_ travail/2005-03.pdf. Cogneau, D., and Ch. Guénard. 2003. "Colonisation, Institutions, and Inequality--A Note on Some Suggestive Evidence." DIAL Working Paper 2003/05, Développement, Institutions et Analyses de Long terme, Paris. http://www.dial.prd.fr/dial_publications/PDF/Doc_ travail/2003-05.pdf. Cogneau, D., and J.-D. Naudet. 2004. "Who Deserves Aid? Equality of Opportunity, Interna- tional Aid, and Poverty Reduction." DIAL Working Paper 2004/10, Développement, Insti- tutions et Analyses de Long terme, Paris. http://www.dial.prd.fr/dial_publications/ PDF/Doc_travail/2004-10.pdf. Collier, P., and D. Dollar. 2001. "Can the World Cut Poverty in Half? How Policy Reform and Effective Aid Can Meet International Development Goals." World Development 29 (11): 1787-1802. Dezalay, Y., and B. G. Garth. 2002. The Internationalization of Palace Wars: Lawyers, Econ- omists, and the Contest to Transform Latin American States. Chicago Series in Law and Society. Chicago, IL: University of Chicago Press. EQUALITY OF OPPORTUNITY IN DEVELOPING COUNTRIES | 67 Diener, E. and E. M. Suh. 1999. "National Differences in Subjective Well-Being." In D. Kah- neman, E. Diener, and N. Schwarz, eds. Well-Being, The Foundations of Hedonic Psychol- ogy. New York: Russell Sage Foundation. Engerman, S. L., and K. L. Sokoloff. 2002. "Factor Endowments, Inequality, and Paths of Development among New World Economies." NBER Working Paper 9259, National Bureau of Economic Research, Cambridge, MA. Fields, G., P. Cichello, S. Freije, M. Menéndez, and D. Newhouse. 2002. "For Richer or for Poorer: Did Household Incomes Converge or Diverge in the 1990s?" Journal of Economic Inequality 1 (1): 67-99. Henrich J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, R. McElreath, M. Alvard, A. Barr, J. Ensminger, K. Hill, F. Gil-White, M. Gurven, F. Marlowe, J. Q. Patton, N. Smith, and D. Tracer. 2001. "Cooperation, Reciprocity and Punishment in Fifteen Small-Scale Societies." American Economic Review Papers and Proceedings 91 (2): 73-8. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny. 1999. "The Quality of Govern- ment." Journal of Law, Economics, and Organization 15 (1): 222-79. Lam, D. 1999. "Generating Extreme Inequality: Schooling, Earnings, and Intergenerational Transmission of Human Capital in South Africa and Brazil." Population Studies Center Report No. 99-439. University of Michigan, Ann Arbor. Llavador, H. G., and J. Roemer. 2001. "An Equal-Opportunity Approach to the Allocation of International Aid." Journal of Development Economics 64 (1): 147-71. Pradhan, M., D. Sahn, and S. Younger. 2003. "Decomposing World Health Inequality." Jour- nal of Health Economics 22 (2): 271-93. Schokkaert, E., and K. Devooght. 1999. "Responsibility-Sensitive Fair Compensation in Dif- ferent Cultures." Catholic University of Leuven Discussion Paper No. DARP 46, Leuven, Belgium. What Is Equity? STEPHAN KLASEN The equity issue is a timely one for the World Bank to tackle head-on in its World Development Report 2006 (WDR). In particular, it appears that an exclusive focus on poverty, the more traditional approach followed by the Bank, has missed some important issues of great intrinsic and instrumental significance for the development process. Also, issues of equity seem to play a very large role in national and interna- tional discussions about success and failure of the development enterprise as a whole. Thus it is critical to investigate whether there is something that can, and should, be done to promote greater equity within countries and between countries. At the same time, it is important to be clear about what one means by equity, how it compares to the other often-used term, equality--and what an understanding of these terms means for the analysis of development issues. These are the focus this note. Definitions: Equity versus Equality Many economists have developed their own notions of what is meant by equity and equality. It may, therefore, be useful to look at some authoritative definitions of the terms, such as those in the Oxford English Dictionary (OED), or its U.S. equivalent, the American Heritage Dictionary. According to these sources, the definitions of equity and equality are as follows: Oxford English Dictionary (2003 edition) Equity: impartial justice, fairness (equitable: uniform in treatment; fair) Equality: having one measure; the same in magnitude, quantity, degree, worth, etc. American Heritage Dictionary (1987 edition) Equity: The state, ideal, or quality of being just, impartial, and fair Equality: The state or instance of being equal; especially, the state of enjoying equal rights, as political, economic, social.1 Stephan Klasen is professor of economics at the University of Göttingen in Germany. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 69 70 | STEPHAN KLASEN While there are nuances between the definitions of these terms in the OED and American Heritage Dictionary, in both cases it is quite clear that equity has to do with fairness and impartiality in some form, while equality is more clearly a mathe- matical concept that focuses on one item being of the same magnitude as another. Among economists, the two terms have very different connotations. The term "equality" is often seen as an impossible, idealistic, or even suspicious goal, often associated with communism/socialism. The term generally has a negative connota- tion and is often seen as a cynical ploy by rulers to rally the masses or remain in power. Both strands of meaning are combined in the famous quote from George Orwell's 1946 novel Animal Farm, where "some are more equal than others." In contrast, equity is more often seen by economists in terms of equal opportunities and is thus seen in a positive light--as something that promotes fairness and possibly even efficiency.2 Interestingly enough, economists evaluate the opposite of these terms--inequality and inequity--quite differently. Both terms are considered problematic, although inequity is probably seen as a source of greater concern. Thus, policymakers often want to achieve greater equity (rather than equality) and complain about great inequality (rather than inequity).3 WDR 2006 is a typical example of the use of this terminology. Equality as a concept is hardly discussed, and the chapters that refer to inequities then proceed to talk about inequalities in particular dimensions, without clarifying the relationship between these inequalities and the problem of inequity. As the central goal, however, equity is the central focus of the report. This clarification of terminology does not go to the heart of the matter, however, which has to do with the dimension of inequality. One cannot usefully speak about equity or equality and their interrelationship without referring to the dimension (or space) in which one considers this question (see Sen 1992).4 Equity/Equality of What? Before discussing particular indicators in which equality or inequality persist (as is done comprehensively in the WDR), it is important to first consider the dimensions from a more fundamental perspective. As it will later be made clear, a choice of space as the focus of equity (or equality?) will necessarily mean tolerating inequality in another space. The suggestive matrix presented in table 1 (where the assignments to cells should be viewed as highly tentative) concentrates on two important distinctions about the dimensions of equity. The columns distinguish whether one is interested in processes or in outcomes (that is, consequences) and are thus concerned with the nature of the welfare goal. The second distinction in the rows differentiates between means to achieve these goals (processes or outcomes) and the ends, the goals themselves. For example, if one chooses to measure equity in the dimension of incomes, one can draw upon the best available database and a large body of literature that has concerned itself with the measurement and interpretation of income inequality.5 This approach also has the advantage of being linked (under very stringent assumptions) WHAT IS EQUITY? | 71 to axiomatic welfare economics (see Grün and Klasen 2003). But there are many drawbacks. At the conceptual level, income is in the cell "Outcomes" and "Means" in table 1. As the income focus is usually derived from utilitarianism, it shares with utilitarianism the feature of being purely consequentialist, ignoring procedural issues that might be of importance (for example, various forms of economic and noneco- nomic freedoms); and it also is a means to an end, not an end in itself (as utility would be but is impossible to measure). Given the heterogeneity among people in their ability to transform incomes into valuable achievements (and the problems of externalities and public goods), equity in incomes provides only a fairly rough guide to equity in well-being. In addition, when considering inequality in incomes, one must also confront the question of absolute versus relative inequality. Most people focus on the latter, although a case can be made to consider the former as well (Atkinson and Brandolini 2004). In addition to these conceptual problems, focusing on incomes leads to practical problems in the measurement of equity, particularly for comparisons between households of different size and composition, and even more so for intra-household equity, where the (household) income metric is not particu- larly useful, despite some claims to the contrary (see Klasen 2004). Instead of incomes, one could focus on equity in rights and liberties in the sense proposed by Nozick (1974). Here the focus is on impartiality of social arrangements and in procedural justice, so rights and liberty clearly belong in the process column, which explicitly disregards consequences. Also, since the procedural liberties are the goal, they are the end to be monitored. Clearly, a focus on equity in procedural liberties will certainly entail inequalities in incomes and in functionings, as equal freedoms will not generate equal outcomes (Sen 1992). It is also unclear how such a focus deals with inequalities due to birth or fortune, which are important in affecting one's life prospects within a country and even more so globally. For example, is it sufficient to give everyone an equal right to schooling and disregard the fact that some are born rich or poor, and some in rich or poor countries, and thus have vastly different abilities to make use of that right? It would be difficult to argue that focusing only on equity in procedures is sufficient given the vast inequities in outcomes that would then have to be tolerated. Sen's concept of functionings ("doings" and "beings" such as being healthy and educated, well-fed and housed, and so forth) clearly falls into the outcome/ends cell. This cell is focused on consequences; but, in contrast to income, it does not measure a means but a direct outcome--and thus avoids the problem of income being an imperfect proxy. (For empirical applications, see Klasen 2000). It also has the TABLE 1. Classifying Different Notions of Equality Process Outcome (consequence) Means Opportunities Income, primary goods Ends Procedural liberties, capabilities Functionings (capabilities) Source: Author's compilation. 72 | STEPHAN KLASEN advantage that it can be measured at the individual level and thus largely avoids the problems of equivalence scales and intra-household distribution. There are, however, a host of practical problems when measuring inequality in functionings, having to do with the choice of functionings, the preferred indicator, or possible weights and aggregations (see Klasen 2000 and Kuklys 2004). Also, there are conceptual prob- lems, such as its total disregard for processes.6 Sen's concept of capabilities is more process-oriented and thus finds itself here in the process/ends cell. But since he calls not just for the freedom to choose but also for the ability to choose among different ways to achieve "substantively valuable" functionings, however, it is not a purely process-oriented approach. It is instead con- cerned about the valuable outcomes that are on offer. Quite clearly, achieving equity in capabilities sets will mean inequality in incomes and many other dimensions, as there are people who need a lot more income (and other things) to achieve the same capabilities due to handicaps of birth or fortune. It has the same advantages as func- tionings but also considers procedural issues, and thus addresses one of the short- comings of the functioning approach. While this concept of capabilities is arguably the most attractive among the lot, it is quite difficult to operationalize since policy- makers do not observe people's capability sets and thus need to rely on observing their actual achievements, that is, the functionings (see Kuklys 2004). So where does the World Bank come down on these questions in the 2006 WDR, and where should it? The outline defines equity to mean that `individuals should have equal opportunities to pursue a life of their choosing and be spared from extreme deprivations in outcomes' (World Bank, 2005: 2). This formulation is most closely related to Roemer's work on `equality of opportunities' which calls for an equaliza- tion of advantages among groups (or, conversely, a removal of disadvantages arising from membership in a certain group (e.g. a certain, race, gender, ethnic group, etc., Roemer, 1998). It adds on, as an institution concerned about poverty and noting an influence from Rawls (1971), an additional concern for absolute deprivation in terms of outcomes. In the way the equality of opportunities is used in report, it would largely belong in table 1 in the process-ends cell. It is concerned with processes and sees these opportunities as means to give people (and their societies) more desirable outcomes.7 It quite explicitly tolerates inequalities in outcomes, including in health, education, and consumption, that are due to `differences in preferences, talents, efforts, and luck' World Bank 2005: 3). Thus the approach taken here leans more in the direc- tion of Nozick than Sen and stops well short of calling for equality in capability sets. In particular, for people disadvantaged in the distribution of talents or luck, equal capability sets would mean that those with these disadvantages would need, for example, higher incomes or greater support in education than others in order to equalize capability sets. This is quite the opposite of saying that one would allow those with advantages in talents or luck to get ahead of the others. One could easily extend the `equal opportunity' approach to take on board Sen's concern if one sim- ply recognized that the distribution of talent and luck are just as much beyond the control of an individual as that individual's race, ethnic origin, or parental back- WHAT IS EQUITY? | 73 ground. This would be a considerably more radical approach and given the new ground the World Bank as an institution has already broken with this report, it prob- ably would have been too much to expect them to take the approach this far. A second reason for the reluctance to move more in the direction of Sen's capabil- ity approach might have been related to measurement issues. It is hard enough, as the report shows, to clearly identify inequality of opportunities in the way defined here. It would have been even harder to measure inequalities in capability sets. Despite the validity of this concern, I would suggest that it would be useful to separate concep- tual and measurement issues. Although one must take short-cuts when it comes to measuring inequalities and equity, the terms should be carefully defined at the con- ceptual level. My personal preference would be to argue for "equality in capabilities" as the goal and then measure inequalities in functionings, procedures, and means (such as incomes) as useful proxies to get at the concept. It would certainly also be useful to think a bit more about operationalizing the capability approach, for exam- ple, through survey-based methods (where the World Bank would have a compara- tive advantage). It is one thing to choose the space in which to consider equity, another to then argue for equity in that space, or why equity is an important concept. While the out- line of the WDR was quite vague on these issues, the report itself has a whole chap- ter (chapter 4) devoted to it which links the issue of equity to a human concern with fairness, an issue I turn to now. Equity as Fairness While fairness is never clearly defined in the report, chapter 4 now points to a human concern with fairness which is deeply rooted in many religious traditions and is enshrined in basic legal principles of impartiality and basic human rights. Second, it reports on inequality aversion as an apparent empirical fact. Third, it draws on a large body of experimental evidence about that many (or in some settings even most) people, even when given the opportunity to act in a selfish manner, appear to act in an egalitarian manner out of concern of fairness. Lastly, it makes a strong instrumen- tal case that a concern for equity as it important for development outcomes and this is the central case made in the report, which is well-argued and supported by much evidence that the World Bank itself has generated in the past few years. The intrinsic arguments need to be examined more closely. There is, indeed, some empirical and experimental evidence about inequality aversion, or a disutility associ- ated with higher inequality. But the range of estimates for inequality aversion from the literature is vast, and the methods used in this literature are not beyond dispute.8 In short, this seems to be a fairly weak leg to stand on as an intrinsic case for equity. Also, it begs the question of why people are inequality averse, which goes back to the question of fairness. I think it is important to have a brief discussion on fairness. Here, too, there is a growing empirical and experimental literature on fairness as an important concern 74 | STEPHAN KLASEN that appears to shape behavior in many contexts which is now discussed in some detail in the report.9 In addition, it is useful to have a more conceptual discussion about the importance of fairness. A useful starting point here is philosopher John Rawls's Theory of Jus- tice, which defines justice as fairness (Rawls 1971). The critical element of his approach is that he treats the assignment of people to talents, abilities, locations, and situations as a random process implying that no one deserves to be born poor or dis- abled, to illiterate parents, into a difficult environment, and so forth. In the so-called original position, Rawls therefore asks us to abstract from this random assignment process and choose a social contract. And he convincingly argues that one would choose a "fair" society in the sense that the disadvantages to the losers in the assign- ment process should be as small as possible. That is, those disfavored by birth or for- tune would have access to redress from everyone else, as no one deserves their luck to be born rich, healthy, and educated in Sweden for example, just as no one deserves to be born poor, illiterate, and starving in the Darfur region of Sudan. Sen would largely concur here, and his call for equal capabilities would translate into specific concern to those disfavored by birth or fortune.10 If one takes this rather radical line, then equality of opportunities in the conven- tional sense will not do--unless it also calls for redress of inequalities of birth or for- tune. Thus, tackling the question of what is meant by fairness helps clarify matters a great deal, and this is where the report could have taken a clearer position. National versus International Inequality Much of the literature, both conceptual and empirical, has focused on equity within countries. Can those studies be translated one-for-one to international inequality? This is an issue that is not discussed in detail in the report. The origin of Rawls's position on justice as fairness was explicitly developed for countries; that is, it did not address the issue of international inequality. For interna- tional inequality issues, he developed a different argument in The Law of Peoples (Rawls 1999). Amartya Sen is also not fully explicit about the international dimen- sion of his call for basic capability equality. But unless one can come up with a defendable argument for why justice as fairness is inherently bound up with the nation-state and stops at the border, there is no reason why the arguments made by Rawls and Sen for a country cannot be applied internationally. Why can one not think of the mental experiment of an original position, where all of humanity has to decide on a social contract without knowing who will then be born rich and healthy in the Netherlands, or poor and starving in Eastern Congo? I strongly believe that Rawls's approach to equity at the national level is, in principle, translatable to the global level.11 Thus, a separate motivation is not needed for global equity.12 Sen makes less of a distinction between national and international considerations here, and one would not do him great injustice by arguing that his call for capability equality applies globally, not just in particular countries. With Sen, however, another potential difficulty arises. He explicitly calls for location-specific standards for meet- WHAT IS EQUITY? | 75 ing such capability equality. Quoting Adam Smith and his famous "shoe" example (for Smith to appear in public without shame in the late 18th century required leather shoes in Scotland, but required no shoes at all in France), Sen argues that to achieve certain capabilities one allegedly needs different resources. For example, the capabil- ity to be integrated with the community might require a car in rural areas of the United States but just a bicycle in rural Bangladesh. Should one, therefore, be con- tent with the inequalities between the United States and Bangladesh that allow Amer- icans to afford cars and Bangladeshi only bicycles? While this is a conceptual problem to sort out, in practice it may not be so vex- ing. First, global inequalities in functionings (and presumably capabilities) are so large that we are far from capability equality as defined here. Also, if we come closer to reaching it, we may confront the problem of a shifting goalpost. In rapidly grow- ing China, being integrated with one's community might soon also require a car-- even if a bicycle used to be just fine. To sum up the discussion so far, it is useful to start with a clear definition of equity as fairness, have a clear discussion of why fairness is something intrinsically valuable, define the dimension in which to measure it, and address the national versus interna- tional issue. My preference would be to define equity as fairness in the Rawlsian sense, focus on equality of capability sets as the dimension in which to conceive it, and to not make a distinction between national and international equity issues (at the conceptual level). If this approach were taken, then there would still be plenty of work to be done to move from concept to measurement. And here, the World Bank would be in an excellent position to break new ground given its expertise in mea- surement issues, particularly at the micro level. Principles versus Policy Issues It is easy to attack the argument so far as hopelessly idealistic. Given the size of national inequalities and the political economy defending the status quo, how can one realistically call for capability equality within a country? If tackling inequality at the national level is already difficult enough, does doing something about the much larger international inequalities appear totally hopeless? It is impossible to conceive of a system of policies or transfers (and the associated institutional set-up) that would be able to achieve capability equality at the global level within a reasonable time frame. Also, such systems and structures would have all sorts of implementation problems, might generate adverse incentives, be inefficient and thus shrink the global economic pie to be distributed, and so forth. Moreover, given today's political econ- omy in industrialized countries and the absence of institutions of global governance, it appears utterly hopeless to advance such arguments. While clearly the discussion so far is about setting conceptually defensible goals, it is also clear that these goals would not lead to short-term policies that are easily implemented. But if one tackles an issue such as equity in development, it is important to get the principles right; and the ones mentioned here are conceptually defensible. Once this is achieved, one can then move to strategic and implementation 76 | STEPHAN KLASEN issues. And then, additional considerations clearly must come into play. But these considerations should not be allowed to alter the principle--just the strategy for get- ting there. Moreover, the willingness to think about global capability equality is more wide- spread than meets the eye. After all, the Millennium Development Goals are a mod- est (and nonbinding) attempt to set goals that go into the direction of basic capabil- ity equality. And they have certainly succeeded in redirecting development efforts to focus on achieving the goals. Conclusions and Policy Questions This note has argued that it is important to tackle head-on the conceptual issues sur- rounding equity in development. I have suggested ways to do that, and encourage the World Bank in follow-up work to consider these strategies--and particularly focus its efforts on trying to operationalize them via the capability approach. This note is not about policy issues. But the menu of issues the World Bank con- siders in the 2006 WDR seems just right. In terms of national inequalities, there is a clear assessment of the potential and limitations of asset redistributions, of pro-poor investments in assets, of redistribution via the tax and transfer systems, and of using trade and exchange rate policies for greater equity. It is also commendable that a par- ticular focus was placed on finding ways to reduce the inter-generational transmis- sion of inequalities, and various ways of affirmative action policies for marginalized groups are discussed. There is much room to disagree on particular policy recom- mendations, but the general direction seems the right one. Regarding international inequalities, a direct focus on inequalities in capabilities at the global level and ways to overcome them would have been particularly useful. A number of important policy issues are taken up, including the role of trade and aid in reducing global inequalities. A greater focus on inequalities in power, which pre- vent many of the recommended policies from being implemented would have been a nice addition. Notes The WDR 2006 outline is posted on the Internet at http://siteresources.worldbank.org/ INTWDR2006/Resources/WDR_on_Equity_FinalOutline_July_public.pdf. 1. Both "equity" and "equality" originate from the Latin aequalis (equal) and entered Mid- dle English via Old French. 2. This critical view of equality and the positive view of equity is related to the Two Funda- mental Theorems of Welfare Economics. While equality would often mean heavy inter- vention in the market mechanism that might hamper efficiency (see the First Theorem), equity can be interpreted as ensuring an acceptable initial distribution of assets which, according to the Second Theorem, can ensure Pareto Optimal outcomes. WHAT IS EQUITY? | 77 3. There is a certain incongruence to this, as all measures of inequality implicitly have a notion of equality as a reference standard and thus seem to implicitly call for "equality" as a goal, which is rarely stated or wanted. In practice, the problem is less severe, since the demonstration of inequality is often just used to argue for some reduction in such inequal- ity (without calling for equality) that might be consistent with a concern for equity. 4. Sen (1992), in fact, argues that all major ethical theories demand equality in some dimen- sion, for reasons of logical plausibility. 5. Once again, however, equity is being equated with absence of inequality. What "fairness" would mean in the income space is not entirely obvious, but this will be considered later in this note. 6. See, for instance, the slavery example from Sen's 1998 book Development and Freedom. 7. The concern about avoiding extreme deprivation would be in the outcomes, ends cell but this is largely seen as a secondary concern throughout the report with equality of oppor- tunities taking precedence. 8. For a discussion of the literature, see Grün and Klasen (2003). 9. For a detailed discussion, see Fehr and Schmidt (2003). 10. They differ on technicalities, however. See Sen (1990). 11. Thus, I differ from his treatment of international issues in Rawls (1999). 12. This poses an additional problem for focusing on inequality aversion as the argument for greater equity. There is much less literature, and even less consensus, on inequality aver- sion at the global level. References Atkinson, Anthony B., and Andrea Brandolini. "Global Inequality and Poverty: Absolute, Rel- ative, or Intermediate?" Paper presented at 2004 IARIW (International Association for Research in Income and Wealth) Conference in Cork, Ireland, August 22­27. Fehr, Ernst, and Klaus Schmidt. 2003. "Theories of Fairness and Reciprocity: Evidence and Economic Applications." In Mathias Dewatripont, Lars Peter Hansen, and Stephen J. Turnovsky, eds. Advances in Economics and Econometrics, Eighth World Congress of the Econometric Society, Vol. 1. Cambridge, U.K.: Cambridge University Press, 208­57. Grün, Carola, and Stephan Klasen. 2003. "Growth, Inequality, and Well-Being: Intertempo- ral and Global Comparisons." CESifo Economic Studies 49: 617­59. Klasen, Stephan. 2000. "Measuring Poverty and Deprivation in South Africa." Review of Income and Wealth 46 (1): 33­58. ------. 2004. "Gender-Related Indicators of Well-Being." WIDER Discussion Paper 2004/05, World Institute for Development Research, United Nations University, Helsinki, Finland. Kuklys, Wiebke. 2004. Amartya Sen's Capability Approach: Theoretical Insights and Empir- ical Applications. Berlin: Springer. Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Harper and Row. Rawls, John. 1971. A Theory of Justice. Oxford, UK: Clarendon Press. ------. 1999. The Law of Peoples. Cambridge, MA: Harvard University Press. Roemer, John. 1998. Equality of Opportunity. Cambridge, MA: Harvard University Press. 78 | STEPHAN KLASEN Sen, Amartya. 1990. Justice: "Means versus Freedoms." Philosophy and Public Affairs 19: 111­21 ------. 1992. Inequality Re-Examined. Cambridge, MA: Harvard University Press. ------. 1998. Development as Freedom. New York: Knopf. World Bank. 2005. World Development Report 2006: Equity and Development. New York: Oxford University Press. Should Poor People Care about Inequality? MARTIN RAVALLION The fact that inequality changes little on average in growing economies, and that absolute poverty tends to fall over time, has led many observers to argue that inequality is unimportant--that all one needs to worry about in fighting absolute poverty is the rate of growth. This paper argues against that view. There are a num- ber of ways in which the extent of inequality in a society (and how it evolves over time) exerts a powerful influence on the extent of poverty today and the prospects for rapid poverty reduction in the future. Efforts to fight poverty need to focus on policies to help assure that poor people are better able to participate in, and con- tribute, to future economic growth. Inequality and Growth A series of papers in the literature have found that changes in inequality at the coun- try level--most often measured by the Gini index--have virtually zero correlation with rates of economic growth; see, for example, Ravallion and Chen (1997), Raval- lion (2001), and Dollar and Kraay (2002).1 Among growing economies, inequality tends to fall about as often as it rises--and similarly among contracting economies. In other words, growth tends to be distribution-neutral on average. If all levels of income grow at roughly the same rate, then absolute poverty must fall. Thus it is not surprising that the literature has also found that absolute poverty measures tend to fall with growth.2 There are a number of reasons to be cautious in drawing implications for policy. First, finding that there is no change in overall inequality can be consistent with con- siderable "churning" under the surface, with gainers and losers at all standards of living. This cannot be seen in the cross-sectional surveys used in almost all empirical Martin Ravallion is a senior adviser in the Development Research Group (DECRG) at the World Bank in Washington, D.C. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 79 80 | MARTIN RAVALLION work on this issue. The more limited panel data sets that are available point to churn- ing.3 Simulations of the impacts of specific policy changes intended to promote eco- nomic growth also point to such "horizontal inequality" in the impacts of reform.4 Second, an unchanging Gini index with growth can accompany large increases in absolute income disparities between "rich" and the "poor" (Ravallion 2003). And the absolute differences may well be more obvious, and of greater concern, to many people in a growing economy than the proportionate differences.5 It may well be that the debate about what is happening to inequality in the world is largely a debate about the meaning of "inequality."6 Third, it must be acknowledged that there are likely to be measurement errors in widely used data. These include sampling errors (probably a minor concern for the surveys used here), errors arising from selective compliance (whereby certain types of households participate in surveys with lower probability than others), under- reporting of incomes, and comparability problems between surveys (arising from dif- ferences in questionnaires, interviewing procedures, or processing methods). These errors can greatly weaken the power of tests using cross-country and inter-temporal comparisons for detecting the true relationships of interest.7 Even putting these measurement problems to one side, there are deeper concerns about the policy interpretation often ascribed to the apparent lack of correlation between changes in inequality and growth. In particular, this lack of correlation does not imply that policymakers who wish to fight poverty in any given country can safely focus on growth alone. All this empirical finding indicates is that, on average, there is little effective redistribution in favor of the poor. It does not indicate that redistribution rarely happens or that distribution is unimportant to the outcomes for poor people from economic growth. The rest of this paper takes up these issues. Inequality and Poverty Reduction While it may be readily agreed that economic growth tends to lead to lower measures of absolute poverty, there is nonetheless a wide variation in the impact of a given rate of growth on poverty. Ravallion (2001) estimates that the 95 percent confidence interval implies that a 2 percent annual growth rate in average household income will bring anything from a modest drop in the poverty rate of 1 percent to a more dra- matic 7 percent annual decline. So, for a country with a headcount index of 40 per- cent, there is 95 percent confidence that the index will fall by somewhere between 0.4 percentage points and 2.8 points in the first year. Why does the same rate of growth bring such different rates of poverty reduction? In answering this question it is convenient to start with the identity that the propor- tionate rate of poverty reduction is the product of the "growth elasticity of poverty reduction" and the rate of growth. Note that this is not the same as the elasticity of a poverty measure with respect to the mean holding distribution constant (as found in Kakwani 1993). The latter can be thought of as the "partial elasticity," as distinct from the total elasticity given by the proportionate rate of poverty reduction divided SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 81 by the rate of growth. Of course, if growth is distribution-neutral on average, then the two elasticities will be similar on average, although they may differ greatly in spe- cific countries and time periods. Two sets of factors can be identified as the main proximate causes of the differing total elasticities of poverty reduction found in practice: the initial level of inequality and how inequality changes over time. Initial Inequality While the evidence suggests that one cannot expect absolute poverty to fall without positive growth, it is intuitive that the higher the initial inequality in a country (even if it does not change over time), the less the gains from growth tend to be shared by the poor. In other words, one expects that a smaller ini- tial share tends to mean a smaller subsequent share of the gains from aggregate eco- nomic expansion. While this intuition is compelling, the theory on how differences in initial inequal- ity will affect the growth elasticity of poverty reduction is ambiguous (Ravallion 2005b). Some special cases do yield unambiguous results, which are essentially achieved by collapsing the potential differences in initial distribution into just one parameter. Analytic predictions obtained under the assumption that household income or consumption is log-normally distributed predict that the growth elasticity of poverty reduction will fall (in absolute value) as inequality rises (Bourguignon 2003). In another single-parameter model, Son and Kakwani (2004) assume that the Lorenz curves across countries only differ in a special way, namely that the entire curve shifts by a constant proportion of the difference between the actual value on the Lorenz curve and the line of equality. They also assume that the growth process is distribution-neutral and that the poverty line is less than the mean. Under these assumptions, Son and Kakwani show that the growth elasticity of poverty reduction (for a broad class of poverty measures) is strictly decreasing in the initial value of the Gini index, which essentially becomes the sole parameter locating the Lorenz curve. These theoretical results are instructive and intuitively consistent. In practice, however, distributions vary by more than one parameter; and growth processes are only (roughly) distribution-neutral on average. Growth in specific countries and time periods is rarely distribution-neutral, so the neutrality assumption can be quite deceptive in predicting outcomes of specific growth episodes. For example, consider the growth process in Brazil in the 1980s. Datt and Ravallion (1992) show that if one had assumed at the outset of the decade that growth would be distribution-neutral, then one would have predicted a 4.5 percentage point decline in the headcount index of poverty, driven by economic growth. In fact, there was no change over the decade, with Brazil's headcount index staying at 26.5 percent. Distributional shifts working against the poor exactly offset the gains from growth. What does the empirical evidence suggest about the relationship between initial inequality and the growth elasticity of poverty reduction? Support for the intuition that higher inequality countries tend to have lower (absolute) growth elasticities of poverty reduction was first presented in Ravallion (1997) and subsequently verified by Ravallion (2001) and Kraay (2005). In modeling the relationship between poverty 82 | MARTIN RAVALLION reduction and growth, Ravallion (1997) postulated that the rate of poverty reduction (measured as the difference in the log of the measure of poverty) is directly propor- tional to the "distribution corrected rate of growth," where the latter is given by the ordinary rate of growth in the mean times a distributional correction given by one minus the initial Gini index. This implies that at very high levels of inequality, the poor will gain little or nothing from economic growth. And at the extreme, in which inequality is at its maximum (an inequality index of one), the richest person will have all of the income, and so all of the gains from growth will go to that person. The elas- ticity will, of course, be zero. This model can be improved (in terms of fit with data on actual spells of changes in poverty matched with growth) by using instead an adjustment for nonlinearity in the relationship between the growth elasticity of poverty and the initial inequality (Ravallion 2005b). Higher levels of initial inequal- ity tend to have progressively smaller impacts on the elasticity as inequality rises. To help interpret the implications of this model, consider the rate of poverty reduction with a 2 percent rate of growth in per capita income (roughly the mean rate for the developing world 1980­2000), with no change in distribution and a headcount index of 40 percent (the mean poverty rate for the developing world around 1980). In a low-inequality country, with a Gini index of 0.30, for example, the headcount index will fall by 6.4 percent per year, or 2.6 percentage points in the first year. The headcount index will be halved in 11 years. By contrast, in a high- inequality country, with a Gini index of 0.60 growing at the same rate and with the same initial headcount index, the latter will fall at an annual rate of 1.2 percent, rep- resenting a decline of only 0.7 percentage points in the first year. It will then take 57 years to halve the initial poverty rate. Poverty responds slowly to growth in high inequality countries; or (to state the same point differently) high-inequality countries will need unusually high growth rates to achieve rapid poverty reduction. The argument works in reverse too; high inequality will help protect the poor from the adverse impact of aggregate economic contraction (Ravallion 1997). Low inequality can thus be a mixed blessing for poor people living in an unstable macro- economic environment: It helps them share in the benefits of growth, but it also exposes them to the costs of contraction. There is evidence that this also happens at local levels during an economy-wide crisis; high inequality districts of Indonesia experienced less dramatic rates of increase in poverty during the 1998 financial cri- sis than did low inequality districts (Ravallion and Lokshin 2004). None of this is inconsistent with the findings in the literature indicating that a large share of the variance in rates of poverty reduction can be attributed to differ- ences in ordinary rates of growth (Ravallion 1995; Ravallion and Chen 1997; Fields 2001; Kraay 2005). In a recent contribution, Kraay (2005) presents Datt-Ravallion decompositions of changes in "$1 per day" poverty measures into growth and redis- tribution components for as many countries as possible. The growth component is the product of the growth rate and the partial elasticity.8 Kraay finds that the vari- ance in the growth component is largely attributable to the growth rate rather than the partial elasticity or its covariance with growth. For example, he attributes 81 per- cent of the variance in the log absolute value of the growth component of changes in the headcount index to the variance in the log absolute growth rate. SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 83 All of this is perfectly consistent with finding that poverty is relatively unrespon- sive to growth in specific countries. Kraay's results are based on averages formed from cross-country comparisons. (A variance is an average too, namely the mean of the squared deviations from the ordinary mean.) For a developing country with aver- age inequality and for which inequality does increase with growth, Kraay's results offer some support for his policy conclusion that, for reducing poverty, the main thing to worry about is achieving a higher rate of growth. However, that does not mean that growth is sufficient, even when inequality is low. If growth in a low-inequality country comes with a sufficient increase in inequality, then it will bypass the poor. And, as noted earlier, the empirical finding that growth is roughly distribution-neutral on average is consistent with the fact that it increases roughly half the time during spells of growth. Policy efforts to keep inequality low may then be crucial to pro-poor growth in many low-inequality countries. Furthermore, as shown earlier, for high inequality countries, growth will be a blunt instrument against poverty--unless that growth comes with falling inequality. The heterogeneity in country circumstances is key here. Averages formed across countries can be quite uninformative about how best to achieve pro-poor growth in specific countries. While initial inequality is an important proximate determinant of differing rates of poverty reduction at a given rate of growth, to help inform policy one needs to probe more deeply into the relevant sources of inequality. There are inequalities in a number of dimensions that are likely to matter, including access to both private (human and physical) assets and public goods. Inequalities in access to infrastructure and social services naturally make it harder for poor people to take up the opportu- nities afforded by aggregate economic growth. Further research is needed on the specific factors that influence the growth elastic- ity of poverty reduction. Some clues as to the role played by initial conditions have been found by comparing rates of poverty reduction across states of India, for which one can compile a long series of reasonably comparable survey data back to about 1960. The analyses of these data confirm that economic growth has tended to reduce poverty in India. Higher average farm yields, higher public spending on development, higher (urban and rural) nonfarm output and lower inflation were all poverty reduc- ing (Ravallion and Datt 2002). However, the response of poverty to nonfarm output growth in India has varied significantly between states; and the differences reflect sys- tematic differences in initial conditions. Low farm productivity, low rural living stan- dards relative to urban areas, and poor basic education all inhibited the prospects of the poor for participating in growth of the nonfarm sector. Rural and human resource development appear to be strongly synergistic with poverty reduction through an expanding nonfarm economy. Changing Income Distribution A second factor influencing the rate of poverty reduction at a given rate of growth is changing income distribution. Finding that growth tends to be distribution-neutral on average does not, of course, mean that dis- tribution is unchanging. Whether or not inequality is rising can make a big difference 84 | MARTIN RAVALLION to the rate of poverty reduction. Among growing economies, the median rate of decline in the "$1 per day" headcount index is 10 percent per year among countries that combined growth with falling inequality, while it is only 1 percent per year for those countries where growth came with rising inequality (Ravallion 2001). Either way, poverty tends to fall, but at very different rates. And similarly among contract- ing economies; poverty rises on average--but much more rapidly when inequality is rising than when falling. As one would expect, changes in distribution matter even more for higher-order poverty measures that can respond quite elastically to even small changes in overall inequality. What happens to inequality in growing economies can make a big difference to the rate of poverty reduction. But what underlies the changes in distribution? There are a great many country-specific idiosyncratic factors (such as shocks to agricultural incomes, changes in trade regimes, tax reforms, welfare-policy reforms, and changes in demographics). Generalizations across country experience are never easy, but one factor that is likely to matter in many developing countries is the geographic and sec- toral pattern of growth. The marked concentrations of poor people in specific regions and/or sectors that occur in many countries point to the importance of the pattern of growth to overall poverty reduction. The extent to which growth favors the rural sec- tor is often key to its impact on aggregate poverty. The geographic incidence of both rural and urban economic growth is often important as well. However, the extent to which the pattern of growth (not simply the overall rate of growth) matters to the rate of poverty reduction is likely to vary from country-to-country, depending on (inter alia) how unbalanced the growth process has been in the past and (hence) how much difference one currently finds between sectors or regions in levels of poverty. Higher growth in a number of developing countries has come with widening regional disparities and often little or no growth in lagging poor areas. For example, nonfarm economic growth in India has not occurred in the states where it would have the most impact on poverty nationally (Datt and Ravallion 2002). And non- agricultural growth has not been concentrated in the states where it would have had the greatest impact on poverty nationally. As a result, the overall nonfarm growth process in India has tended to become less pro-poor over time (Ravallion 2005b). Nor has the geographic pattern of agricultural growth in India been particularly pro- poor. The states with higher growth in agricultural yields were not the key states with higher share's of India's poverty; indeed, there is a mild negative correlation, although not statistically significant (Datt and Ravallion 2002). Agricultural growth as a whole has also lagged relative to India's (primarily urban) nonfarm economy. A Growth-Equity Trade Off? Making growth more pro-poor requires a combination of more growth, a more pro- poor pattern of growth, and success in reducing the antecedent inequalities that limit the prospects for poor people to share in the opportunities unleashed in a growing economy. The ideal combination will naturally vary with country circumstances. In some countries, attention can safely focus on the overall rate of growth to assure SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 85 rapid poverty reduction; elsewhere, a broader approach will be called for. This begs the question as to whether there might be a trade-off between interventions to make growth more pro-poor and the rate of growth. While poverty is more often seen as a consequence of low average income, there are reasons for thinking that there is a feedback effect whereby high inequality also impedes future growth.9 In many developing countries, a plausible way this can hap- pen stems from credit market failures, which mean that some people are unable to exploit growth-promoting opportunities for investment. And it will tend to be the poor for whom these constraints are most likely to be binding. With declining mar- ginal products of capital, the output loss from the market failure will be greater for the poor. So the higher the proportion of poor people there are in the economy, the lower the rate of growth. Then poverty is self-perpetuating. There are other ways in which initial distribution matters to growth prospects. In the presence of capital market failures due to moral hazard, high inequality can dull incentives for wealth accumulation. It has also been argued that high inequality can foster macroeconomic instability and impede efficiency-promoting reforms that require cooperation and trust.10 There is supportive evidence for the view that inequality is bad for growth from cross-country comparisons of growth rates, suggesting that countries with higher ini- tial inequality experienced lower rates of growth (controlling for other factors such as initial average income, openness to trade, and the rate of inflation).11 At the same time, there are also a number of concerns about the data and methods underlying these findings based on cross-country comparative analysis (Ravallion 2001). The hope is that future research will throw more light on the magnitude of the efficiency costs of inequality. China, since around 1980, has often been cited as a country and period in which letting inequality rise allowed rapid growth and (hence) rapid poverty reduction. There can be no doubt that absolute poverty in China has fallen greatly since the 1980s. However, one of the most surprising aspects of China's success against poverty to emerge from recent research is that there is very little evidence of an aggre- gate growth-equity trade off (Ravallion and Chen 2004). Inequality in China has clearly shown a tendency to rise over time. But this correlation could well be spuri- ous, since both inequality and mean income have strong trends, though possibly associated with different causative factors. A better test is to compare the growth rates with changes in inequality over time. Then it becomes far less clear that higher inequality has been the price of China's growth. The periods of more rapid growth did not bring more rapid increases in inequality; indeed, the periods of falling inequality (1981­5 and 1995­8) had the highest growth in average household income. Also, the sub-periods of highest growth in the primary sector (1983­4, 1987­8 and 1994­6) did not accompany lower growth in other sectors (Ravallion and Chen 2004). Nor does one find that the provinces with more rapid rural income growth experienced a steeper increase in inequality; if anything it was the opposite. This lack of any evident aggregate trade-off in China's experience since around 1980 has important implications. On the one hand, it means that growth will tend 86 | MARTIN RAVALLION to reduce absolute poverty. Naturally, with the same growth rate and no rise in inequality, the number of poor in China would be lower; indeed, it would be less than one-quarter of its actual value according to the calculation in Ravallion and Chen (2004). This calculation would be deceptive if inequality rises with economic growth, as the "price" of that growth. However, the evidence does not support that view. On the other hand, the absence of such a trade-off also means that rising inequality put a serious brake on China's pace of poverty reduction. That is also borne out by the finding of Ravallion and Chen (2004) that the provinces with a more rapid rise in rural inequality saw less progress against poverty, not more. As China's policymakers now realize, it will be harder for China to maintain its past rate of progress against poverty without addressing the problem of rising inequality. To the extent that recent history is any guide to the future, one can expect that the historically high levels of inequality found in many provinces today will inhibit future prospects of poverty reduction--just as one finds that the provinces that started the reform period with relatively high inequality faced a double handi- cap in future poverty reduction: They had lower subsequent growth and the poor shared less in the gains from that growth (Ravallion and Chen 2004). Other factors point to the same conclusion. It appears that aggregate economic growth increasingly comes from sources that bring more limited gains to the poor- est. The low-lying fruit of efficiency-enhancing pro-poor reforms are perhaps becom- ing scarce. Inequality is continuing to rise and poverty is becoming more responsive to rising inequality. At the outset of China's current transition period to a market economy, levels of poverty were so high that inequality was not an important con- cern. But that has changed. It also appears that perceptions of what "poverty" means are evolving in China. It can hardly be surprising to find that the standards that defined poverty 20 years ago have lost relevance to an economy that quadrupled its mean income over that period. China could well be entering a stage of its development in which relative poverty emerges as a more important concern than in the past. Economic growth will then be a blunter instrument for fighting poverty in the future. Lessons for Development Policy and Future Research If one accepts that inequality is bad for poor people, what should policymakers do about it? First, the objective must be made clear. If it is agreed that poverty reduc- tion is a far more important overall goal for development policy than reducing inequality per se, then redistributive policies that come at the expense of lower, longer-term, living standards for poor people should not be accepted. Accepting that there is no aggregate trade-off between mean income and inequality does not mean that there are no trade-offs at the level of specific policies. Reducing inequality by adding further distortions to an economy may well have ambiguous effects on growth and poverty reduction. But neither should it be presumed that there will be such a trade-off with all redistributive policies. The potential for "win-win" policies SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 87 stems from the fact that some of the factors that impede growth also mean that the poor share less in the opportunities unleashed by growth. Economists have learned that more rapid poverty reduction requires a combina- tion of more growth, a more pro-poor pattern of growth, and success in reducing the antecedent inequalities that limit the prospects for poor people to share in the oppor- tunities unleashed by a growth economy. Even a distribution-neutral growth process--which is hardly a high standard for "equitable growth" in high-inequality countries--can leave many poor people behind. The challenge for future research is to better understand the specific factors that constrain some poor people from par- ticipating in the benefits of a growing economy, and to draw out the lessons for the types of policies that are needed for both rapid poverty reduction and economic growth. A majority of the world's poor people still live in rural areas, and this is likely to remain true for some time to come (Ravallion 2002). Agriculture and non- farm rural development should thus remain a high priority for sectoral policies. However, past interventions have had a mixed record. New approaches based on community-driven development have held promise but need careful monitoring and evaluation. Also needed is the recognition that the likely heterogeneity in perform- ance across institutional settings depends upon how successful local elites are in cap- turing the gains (Mansuri and Rao 2004; Galasso and Ravallion 2005). The continuing existence of marked regional disparities in living standards has prompted renewed interest in explicit geographic dimensions in policymaking, such as "poor-area programs" and attempts to set up "growth poles." However, many questions remain. Are infrastructure investments in poor areas (often with poor nat- ural resources) effective in reducing poverty? Does it make more sense to move jobs to people, or people to jobs? Is there a trade-off between achieving greater regional equity--such as by focusing on areas with high poverty rates but low poverty densities-- and poverty reduction in the aggregate? Recognizing that it is typically the poor rather than the rich who are locked out of profitable opportunities for self-advancement by the failures of markets and gov- ernments, interventions that make these institutions work better for poor people today can also help promote pro-poor growth in the future. Successful policies can focus on either correcting the underlying market and governmental failure or on directly intervening to redress the asset inequalities, by fostering accumulation of (physical and human) assets by poor people. One can point to the potential impor- tance of a wide range of policies, including sound public investments in rural infra- structure, better policies for delivering quality health and education services to poor people, and policies that allow key product and factor markets (for land, labor, and credit) to work better from the point of view of poor people. The combination of interventions needed will, of course, depend on country and regional circumstances. Much is still unknown about the most appropriate policy combinations in specific circumstances, although some pointers have emerged from research. Making the pro- vision of health and education services more responsive to the needs of poor people is likely to be crucial to achieving pro-poor growth in most settings (World Bank 2004a). In rural economies, security of access to land through tenancy reform and 88 | MARTIN RAVALLION titling programs is arguably no less important (World Bank 2004b). In some circum- stances, rural infrastructure development can also play a decisive role; for example, research has revealed the importance of rural roads to achieving more pro-poor growth processes in rural China (Jalan and Ravallion 2002) and that quite reason- able rates of return are possible from well-designed poor-area development programs (Ravallion and Chen 2005). Better instruments for credit and insurance can also help, both in smoothing consumption and underpinning otherwise risky growth- promoting strategies. Removing biases against the poor in taxation, spending, and regulatory (including migration) policies can also play an important role. Again tak- ing an example from China, reducing the government's taxation of farmers through foodgrain procurement quotas has been a powerful instrument against poverty (Ravallion and Chen 2004). The challenge for policy is to combine growth-promoting policies with the right policies to assure that the poor can participate fully in the opportunities unleashed, and so contribute to that growth. If a country gets the combination of policies right, then both growth and poverty reduction can be rapid. Get it wrong, and both may well be stalled. Future research can help meet this challenge by: · throwing light on the country-specific and subnational factors that influence the distributional nature of aggregate growth; · identifying to what extent those factors are amenable to policy intervention; and · quantifying the trade-offs between alternative policies for promoting pro-poor growth, embracing both redistributive social policies and alternative growth strategies. Notes 1. The Gini index takes a value between zero (perfect equality, meaning that everyone has the mean income) and unity (the richest person has all the income). The specific value of the index is given by twice the area between the Lorenz curve (giving the cumulative pro- portion of total income held by a given proportion of the population ranked by income) and the line of perfect equality. Although it is only one of many measures of inequality, the Gini index has been the most popular one in practice. For a good technical overview of alternative measures, see Cowell (2000). 2. Supportive evidence can be found in World Bank (1990, 2000), Ravallion (1995), Raval- lion and Chen (1997), Fields (2001), Dollar and Kraay (2002), and Kraay (2005). 3. A useful compilation of studies using panel data can be found in the August 2000 special issue of the Journal of Development Studies; see the introduction by Baulch and Hod- dinott (2000). 4. In the context of trade reform, see the discussion in Ravallion (2005a). 5. For further discussion on the distinction between absolute and relative inequality, see Amiel and Cowell (1999), Atkinson and Brandolini (2004), and Ravallion (2003, 2004). 6. Ravallion (2004) develops this argument further. SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 89 7. Ravallion (2005b) further examines the robustness of some of the claims made about growth and changing inequality to allowances for measurement error. 8. Recall that it is a partial elasticity because it holds distribution constant; by contrast the total elasticity lets distribution vary consistently with the data. The analytic elasticities of poverty measures discussed in Kakwani (1993) and Bourguignon (2001) are partial elasticities. 9. There is now a sizable theoretical literature on the various ways in which inequality can impede growth. Contributions include Galor and Zeira (1993), Banerjee and Newman (1993), Benabou (1996), Aghion, Caroli, and Garcia-Penalosa (1999) and Bardhan Bowles, and Gintis, (1999). 10. Aghion, Caroli, and Garcia-Penalosa (1998) and Bardhan, Bowles, and Gintis (1999) review these and related arguments as to why high inequality can reduce aggregate output. 11. See Persson and Tabellini (1994), Alesina and Rodrik (1994), Clarke (1995), Birdsall, Ross, and Sabot (1995), Perotti (1996), Deininger and Squire (1998) and Easterly (2002). References Aghion, Philippe, Eva Caroli, and Cecilia Garcia-Penalosa. 1999. "Inequality and Economic Growth: The Perspectives of the New Growth Theories." Journal of Economic Literature 37 (4): 1615­60. Alesina, Alberto, and Dani Rodrik. 1994. "Distributive Politics and Economic Growth." Quarterly Journal of Economics 109 (2): 465­90. Amiel, Yoram, and Frank Cowell. 1999. Thinking about Inequality: Personal Judgment and Income Distributions. Cambridge, U.K.: Cambridge University Press. Atkinson, Anthony B., and Andrea Brandolini. 2004. "Global World Inequality: Absolute, Relative or Intermediate?" Unpublished paper, Nuffield College, Oxford University. Banerjee, Abhijit V., and Andrew F. Newman. 1993. "Occupational Choice and the Process of Development." Journal of Political Economy 101 (2): 274­98. Bardhan, Pranab, Samuel Bowles, and Herbert Gintis. 1999. "Wealth Inequality, Wealth Con- straints, and Economic Performance." In A.B. Atkinson and F. Bourguignon, eds. Hand- book of Income Distribution, vol. I. Amsterdam: North-Holland. Baulch, Bob, and John Hoddinott. 2000. "Economic Mobility and Poverty Dynamics in Developing Countries." Journal of Development Studies 36 (6): 1­24. Benabou, Roland. 1996. "Inequality and Growth." In Ben Bernanke and Julio Rotemberg, eds. National Bureau of Economic Research Macroeconomics Annual. Cambridge, MA: MIT Press, pp.11­74. Birdsall, Nancy, Ross, D. and R. Sabot. 1995. "Inequality and Growth Reconsidered: Lessons from East Asia." World Bank Economic Review 9 (3): 477­508. Bourguignon, François. 2001. "The Pace of Economic Growth and Poverty Reduction." Paper presented at LACEA 2001 Conference in Montevideo, Uruaguay, October 18­20. ------.2003. "The Growth Elasticity of Poverty Reduction: Explaining Heterogeneity across Countries and Time-periods." In T. Eichler and S. Turnovsky, eds. Growth and Inequal- ity. Cambridge, MA: MIT Press. Chen, Shaohua, and Martin Ravallion. 2001. "How Did the World's Poorest Fare in the 1990s?" Review of Income and Wealth 47: 283­300. ------.2004. "How Have the World's Poorest Fared Since the Early 1980s?" World Bank Research Observer 19 (2): 141­70. 90 | MARTIN RAVALLION Clarke, George R. G. 1995. "More Evidence on Income Distribution and Growth." Journal of Development Economics 47: 403­28. Cowell, Frank. 2000. "Measurement of Inequality." In A.B. Atkinson and F. Bourguignon, eds. Handbook of Income Distribution. Amsterdam: North-Holland. Datt, Gaurav, and Martin Ravallion. 1992."Growth and Redistribution Components of Changes in Poverty Measures: A Decomposition with Applications to Brazil and India in the 1980s." Journal of Development Economics 38: 275­95. ------. 2002. "Has India's Post-Reform Economic Growth Left the Poor Behind?" Journal of Economic Perspectives 16 (3): 89­108. Deininger, Klaus, and Lyn Squire. 1998. "New Ways of Looking at Old Issues: Inequality and Growth." Journal of Development Economics 57(2): 259­87. Dollar, David, and Aart Kraay. 2002. "Growth Is Good for the Poor." Journal of Economic Growth 7 (3): 195­225. Easterly, William. 2002. "Inequality Does Cause Underdevelopment: New Evidence," Work- ing Paper 1, Center for Global Development, Washington DC. Fields, Gary S. 2001. Distribution and Development. New York: Russell Sage Foundation. Foster, James., J. Greer, and E. Thorbecke. 1984. "A Class of Decomposable Poverty Mea- sures." Econometrica 52: 761­65. Galasso, Emanuela, and Martin Ravallion. 2005. "Decentralized Targeting of an Anti-Poverty Program." Journal of Public Economics 85: 705­27. Galor, Oded, and Joseph Zeira. 1993. "Income Distribution and Macroeconomics." Review of Economic Studies 60 (1): 35­52. Jalan, Jyotsna, and Martin Ravallion. 2002. "Geographic Poverty Traps? A Micro Model of Consumption Growth in Rural China?" Journal of Applied Econometrics 17:329­46. Kakwani, Nanak. 1993. "Poverty and Economic Growth with Application to Côte d'Ivoire." Review of Income and Wealth 39: 121­39. Kraay, Aart. Forthcoming. "When Is Growth Pro-Poor? Evidence from a Panel of Countries." Journal of Development Economics. Mansuri, Ghazala, and Vijayendra Rao. 2004. "Community-Based and -Driven Develop- ment." World Bank Research Observer 19 (1): 1­40. Perotti, Roberto. 1996. "Growth, Income Distribution and Democracy: What the Data Say." Journal of Economic Growth 1 (2): 149­87. Persson, Torsten, and Guido Tabellini. 1994. "Is Inequality Harmful for Growth?" American Economic Review 84: 600­21. Ravallion, Martin. 1995. "Growth and Poverty: Evidence for Developing Countries in the 1980s." Economics Letters 48: 411­17. ------. 1997. "Can High Inequality Developing Countries Escape Absolute Poverty?" Eco- nomics Letters 56: 51­7. ------. 2001. "Growth, Inequality, and Poverty: Looking Beyond Averages." World Devel- opment 29 (11): 1803­15. ------. 2002. "On the Urbanization of Poverty." Journal of Development Economics 68: 435­42. ------. 2003. "The Debate on Globalization, Poverty, and Inequality: Why Measurement Matters." International Affairs 79 (4): 739­54. ------. 2004. "Competing Concepts of Inequality in the Globalization Debate." In Susan Collins and Carol Graham, eds. Brookings Trade Forum 2004. Washington, DC: Brook- ings Institution, pp. 1­38. SHOULD POOR PEOPLE CARE ABOUT INEQUALITY? | 91 ------. 2005a. "Looking Beyond Averages in the Trade and Poverty Debate." Policy Research Working Paper 3461, World Bank, Washington DC. ------. 2005b. "Inequality Is Bad for the Poor." Background Paper to the World Develop- ment Report 2006. World Bank, Washington DC. Ravallion, Martin, and Shaohua Chen. 1997. "What Can New Survey Data Tell Us about Recent Changes in Distribution and Poverty?" World Bank Economic Review 11 (2): 357­82. ------. 2004. "China's (Uneven) Progress against Poverty." Policy Research Working Paper 3408, World Bank, Washington, DC. ------. 2005. "Hidden Impact: Household Saving in Response to a Poor-Area Development Project." Journal of Public Economics 89: 2183­2204. Ravallion, Martin, and Gaurav Datt. 2002. "Why Has Economic Growth Been More Pro-Poor in Some States of India than Others?" Journal of Development Economics 68: 381­400. Ravallion, Martin, and Michael Lokshin. 2004. "Lasting Local level Impacts of a Crisis." Pol- icy Research Working Paper 3503, World Bank, Washington, DC. Son, Hyun, and Nanak Kakwani. 2004. "Economic Growth and Poverty Reduction: Initial Conditions Matter" Working Paper 2, International Poverty Center, UNDP, Geneva, Switzerland. World Bank. 1990. World Development Report: Poverty. New York: Oxford University Press. ------. 2000. World Development Report: Attacking Poverty. New York: Oxford University Press. ------. 2004a. World Development Report: Making Services Work for Poor People, New York: Oxford University Press. ------. 2004b. Land Policies for Growth and Poverty Reduction. New York: Oxford Univer- sity Press. Equity-Enhancing Social Transformation: Historical Evidence from Europe and Transition Countries Inequality and Political Regimes CARLES BOIX The aim of this paper is to discuss the effects that economic and social inequality have on two key dimensions of politics: the type of political regime (democratic or authoritarian) chosen by the citizenry, and the extent of political violence in any given country. Even though there is a long tradition in political science examining the relationship between the level of economic equality and the type of political institu- tions, that research, which includes work by Aristotle, Machiavelli, Montesquieu, and Tocqueville, is mostly nonformal in nature. The first part of this paper presents the political impact of inequality with the aid of formal reasoning.1 The second part offers systematic empirical evidence assessing the theoretical claims of the first part. Theory A democracy is defined by a procedure through which its citizens decide, by casting a vote or a sequence of votes, how to govern themselves--that is, what rules should bind their collective life, what should be the optimal distribution of assets, and so forth. This mechanism of decision implies that once a resolution is taken (a vote is con- ducted over a certain issue), a majority of the population determines the position (or welfare) of each member of the population--including the minority that has not agreed with that majority. In other words, in a democracy the minority is subjected to the will of the majority. As a result, a democratic regime is possible only if the minority accepts the electoral outcome in which it participates. More generally, a sta- ble or successful democracy (defined by the uninterrupted use of free and fair voting mechanisms to arrive at political decisions and to select public officials) will only take place if both the winners and the losers of elections (or, the majority and the minority) have an interest in complying with the outcomes of the periodic votes they employ to decide how to govern themselves.2 Carles Boix is professor of political science at the University of Chicago, in Chicago, Illinois. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 95 96 | CARLES BOIX Whether or not the electorate and its representatives comply with the electoral procedure will depend on the alternative value reported by each of the political strategies they may follow. If the gain that voters expect to obtain from an electoral victory is larger than the net benefit that they would accrue from rejecting the dem- ocratic outcome, they will accept the democratic game. Otherwise, they will contest the democratic procedure and try to curtail the space of policies in which the major- ity may govern. Stated differently, electoral losers may abide by the election, accept the defeat, and wait until a new electoral contest takes place. But if the outcome is too unsatisfactory for them to behave well--and the current benefits of the office they are forsaking are too large--those that suffer defeat will be likely to denounce the results and try to block the adoption of democratic procedures. More specifically, the extent to which all citizens accept or reject the democratic order is a function of three conditions: the distribution of income, the nature of assets in the country, and the organizational capacity of political actors.3 Income Distribution The type of political regime derives, first of all, from the underlying distribution of income. To see why, the discussion here departs from (and then extends) the optimal taxation model introduced by Romer (1975) and Meltzer and Richards (1981). In this model, individuals have different skills and assets (and thus income) and choose to work different amounts. The state sets a linear tax on income, and all tax revenues are redistributed lump sum (in an equal amount) to everyone. That transfer, which is the total tax revenue divided by the number of indi- viduals in the economy, equals the taxes paid by the voter with average income. As a result, all of the individuals who are poorer than the average income voter will favor a positive tax, since the transfer they will receive under this scenario is larger than the amount of tax they pay. Conversely, all of the individuals who are richer than the average income voter will oppose such a tax. By assumption, the distribution of skills and incomes is skewed so that the median voter (50 percent of the other voters fall above and below this voter on the income ladder) is poorer than the average income voter (who pays the average tax and receives a transfer that equals the amount paid in taxes).4 As a result of this distribu- tion, the median voter will favor a positive tax rate. In a democratic setting with a one-person one-vote rule and following the median voter theorem, the policy that is adopted is the one most preferred by the median voter--a positive tax and positive transfers. Naturally, the poorer the median voter is relative to the average income individ- ual, the larger the tax rate favored, because the difference between the tax paid by the median voter and the transfer received becomes larger. In other words, as the income distribution becomes more skewed and unequal, the tax pressure should increase according to the Meltzer-Richards model. The level of taxation will only be constrained by the distortions that come from tax rates; the median voter will never impose taxes that lead to an overall decline in total revenue.5 Economists have always treated the voting rule of this taxation model (where all individuals vote, and the median voter determines the tax policy) as a fixed or INEQUALITY AND POLITICAL REGIMES | 97 exogenous parameter. Yet this theoretical decision is wrong.6 The choice of the voting rule precedes the actual process of voting about the tax rate and the distri- bution of assets. In fact, all political actors determine who will vote (and how the vote will take place), informed by the tax outcomes they anticipate will take place under each alternative voting rule. Those individuals who are richer than the aver- age income voter will be interested in excluding the poorest voters (and thus mov- ing the median voter toward them)--in order to minimize the amount of tax to be paid. Conversely, the voters with the lowest incomes will press for their full polit- ical participation. In other words, in highly unequal economies, the demand and the political pres- sure for redistribution is so intense that the wealthy will prefer an authoritarian regime that excludes the majority of the population rather than pay high, quasi- confiscatory taxes. Naturally, since the introduction of an authoritarian regime requires bearing some costs of repression, the well-off will only be interested in excluding the poorest if those repression costs are larger than the taxes they would pay under a democratic regime.7 Correspondingly, as the distribution of income becomes more equal among individuals, redistributive pressures from the poorest social sectors on the well-off voters diminish. The relative costs of tolerating a mass democracy (that is, the taxes they would pay) decline for the holders of the most productive assets to the point that they become smaller than the costs of repression they would have to bear to exclude the majority of citizens. Democracy becomes acceptable to all parties.8 Type of Assets As pointed out earlier, the median voter sets the tax rate con- strained by its distortionary effects on the economy, and hence on total revenue. Those effects, and the extent to which this constraint binds the median voter, vary with the type of capital. As capital becomes more mobile (that is, as it can be moved abroad to obtain a return equal to the domestic return), the tax rate will decline; otherwise, the capital holder would have an incentive to transfer his assets abroad. Similarly, whenever capital can be easily hidden from the state or is of a kind that can only be used by its owner, the temptation to confiscate it also declines. As the constraints on the tax rate due to the type of assets increase and the redis- tributive pressures from noncapital holders on capitalists diminish, the likelihood of democracy will rise. By contrast, economies with a large proportion of fixed assets, such as the oil countries, will generally remain authoritarian--even if they have extremely high levels of per capita income (which modernization theory has corre- lated with the extension of democracy). The combination of the distribution of assets and income and the nature of wealth leads to the set of predictions displayed in figure 1. A stable democracy prevails at low levels of either inequality or specificity of wealth. As the cost of taxation increases (due to increases in wealth inequality and asset specificity), authoritarian- ism starts to pay off. Finally, for high levels of inequality and asset specificity, author- itarianism becomes the dominant strategy of high-income individuals. 98 | CARLES BOIX FIGURE 1. Democracy and Inequality High Area of political violence Authoritarianism Inequality Democracy High Asset Specificity Source: Author's depiction. Political Violence A final prediction of the theory is that political violence increases with inequality in fixed-assets economies. As income inequality rises, the resistance of the well-off (such as landowners or government officials in control of mining resources in rentier states) to the introduction of democratic, peaceful means to set government policy grows--the losses they would incur from majority rule would be simply too substantial. Correspondingly, resorting to violence becomes more attractive to those that are excluded from the state apparatus--the prize of vic- tory raises with inequality. Political violence becomes particularly acute in unequal economies in which assets are fixed and are not complementary to the skills of their current owners. In those cases, the potential rebels can apply violence to overturn the existing regime with the relative certainty that assets will not be moved out of the country and that the elimination of their owners will not reduce their economic value. Empirics The material presented in this section relies primarily on work presented in Boix (2003, 2004). The first two subsections summarize empirical evidence on the causes of democratization and democratic stability. The first subsection uses evidence for the past 150 years. The second subsection employs direct measures of inequality for the second half of the 20th century. The third subsection describes the social and eco- nomic roots of political violence. Political Regimes since the Mid-19th Century Although only the postwar period offers relatively consistent and direct measurements of inequality, it does not include the key democratization waves (and the periods of democratic breakdowns) that took place prior to World War II. To overcome this shortcoming and to increase INEQUALITY AND POLITICAL REGIMES | 99 TABLE 1. Estimating the Probability of Political Transition, 1850­99 DEPENDENT : 1. Probability of transition to democracy: beta coefficient VARIABLE 2. Probability of stable democracy: sum of alpha and beta coefficients INDEPENDENT Model 1 Model 2 VARIABLES Beta Alpha Beta Alpha Constant ­2.7179*** 3.2752*** ­2.7593*** 3.3053*** (0.0979) (0.2089) (0.1046) (0.2390) Percentage of 0.0007^^^ 0.0126*** 0.0002^^^ 0.0121*** family farmsa (0.0019) (0.0039) (0.0021) (0.0042) Index of Educationb 0.0113*** ­0.0024^^^ 0.0134*** ­0.0063^^^ (0.0028) (0.0066) (0.0031) (0.0069) Index of Occupational 0.0050* 0.0093^^^ 0.0082** 0.0076^^^ diversificationc (0.0028) (0.0064) (0.0040) (0.0082) Per capita income ­0.0000 0.0418 (in thousands)d (0.0000) (0.0418) Log-likelihood ­750.61 ­690.93 P>Chi-2 0.0000 0.0000 Pseudo R2 0.8878 0.8905 Number of observations 10715 9960 Source: Democratic Institutions, democracy as coded in Boix and Rosato (2001) for 1850­1949, and Przeworski and others (2000) for 1950­90. a. Area of family farms as a percentage of the total area of holdings. Source: Vanhanen (1997) and Statistical Office of Finland. b. Arithmetic mean of the percentage of literates in the adult population and the "level of students." The level of stu- dents is the number of students per 100,000 inhabitants normalized so that 1,000 students per 100,000 inhabitants corresponds to a level of 100 percent. Source: Vanhanen (1997) and Statistical Office of Finland data (www.stat.fi). c. Arithmetic mean of percentage of non-agricultural population and percentage of urban population. Urban popu- lation is defined as population living in cities of 20,000 or more inhabitants. Source: Vanhanen (1997) and Statistical Office of Finland. d. Per Capita Income. Log of per capita GDP in $ in 1985 constant prices. Source: Heston, Summers, and Aten (2002), Maddison (1995), and Bourguignon and Morrisson (2002). Estimation: Dynamic probit model. Standard errors in parenthesis. *** p<0.01; ** p<0.05; * p<0.01 ^^^ p<0.01 in joint test of interactive terms and its components; ^^ p<0.05 in joint test of interactive terms and its components; ^ p<0.01 in joint test of interactive terms and its components. the number of observations, this analysis presents estimations done for a panel of country-year observations ranging from 1850 until 1999 (about 10,000 data points). Since data on income inequality are scarce for any country before World War II, the focus here is on two indicators that predict the extent of economic inequality consid- erably well: the distribution of agricultural property (measured through the percent- age of land belonging to family exploitations) and the quality of human capital. In addition, both the index of human capital and an index of occupational diversifica- tion (the average of urbanization and industrialization) proxy for the level of asset specificity once per capita income is controlled for, since it can be affected by other factors such as extractive resources in each country. Table 1 estimates the factors conducive to both democratization and democratic breakdown. The estimation (done through a dynamic probit) generates two sets of parameters: The first parameter (the beta coefficient) estimates the probability of 100 | CARLES BOIX transition from authoritarianism to democracy. The sum of the two parameters, beta and alpha, indicates the probability that an already democratic system will not break down.9 Model 1 in table 1 excludes per capita income, while model 2 includes it. The coef- ficients of the independent variables are in line with the theoretical expectations. Higher levels of economic equality (in the countryside in the form of more family farms and, in general, through higher literacy levels) and more nonspecific assets (through more human capital and a more diversified economic structure) increase both the chances of a democratic transition and the stability of democratic regimes. To clarify the results in table 1, table 2 simulates (based on model 1), the annual probability of democratic transition and democratic breakdown as a function of eco- nomic structure, that is, the average of urbanization and industrialization and the percentage of family farms. The probability of having a democratic transition is driven by the type of economic structure (table 2a). As the economy becomes more TABLE 2. a. Predicted probability of transition from authoritarianism to democracy by urban and industrial structure and distribution of property, 1850­1980 Average of urban population and Percentage of family farms manufacturing sector 0 25 50 75 100 0 0.009 0.010 0.010 0.011 0.011 25 0.013 0.013 0.014 0.015 0.015 50 0.017 0.018 0.019 0.020 0.021 75 0.024 0.025 0.026 0.027 0.028 100 0.031 0.033 0.034 0.035 0.037 b. Predicted probability of democratic breakdown by urban and industrial structure and distribution of property, 1850­1980 Average of urban population and Percentage of family farms manufacturing sector 0 25 50 75 100 0 0.200 0.120 0.066 0.033 0.015 25 0.115 0.063 0.031 0.014 0.006 50 0.060 0.030 0.013 0.005 0.002 75 0.028 0.012 0.005 0.002 0.001 100 0.012 0.005 0.002 0.001 0.000 Source: Simulation taken from Boix (2003), tables 2.3 to 2.5. Note: Simulation based on table 1, model 1. Index of knowledge distribution set at median value (31.8 percent). INEQUALITY AND POLITICAL REGIMES | 101 diversified, the chances of a transition increase from about 1 percent if less than a fourth of the economy is urbanized and industrialized, to around 3 percent if more than three-fourths is modernized. The role of the distribution of rural property in affecting regime transitions is positive but marginal. Both factors are more important in determining the robustness of democracies (table 2b). The probability of a democratic breakdown in any given year reaches 20 percent in highly unequal and underdeveloped countries. Yet as either rural equality or industrialization increases, the authoritarian threat disappears. In a highly rural economy, the probability of a democratic breakdown falls to 1 percent as one moves from a country controlled by landowners (Russia before the Stolypin reforms and the Soviet Revolution, Spain for most of the 20th century, as well as most Latin Ameri- can nations) to a highly fragmented property system (like the one prevailing in Nor- way, where at the turn of the 20th century family farms represented almost four- fifths of all land). Similarly, even when the distribution of property remains highly unequal, the chances of an authoritarian backlash disappear as most of the economy ceases to be based on fixed assets. Political Regimes since 1950 To complement the results presented so far, table 3 reproduces the simulated impact of inequality (measured through Gini coefficients as reported by Deininger and Squire (1996) and the nature of assets (the size of the agri- cultural sector and the weight of fuel exports) on the annual probability of transiting from an authoritarian regime to a democracy. These estimations are derived from dynamic probit regressions done over a panel data of yearly observations of all sov- ereign countries since 1950.10 Table 3a simulates the effect of inequality and per capita income. A high Gini index (the maximum in the sample is 0.66 in Zimbabwe in 1950), the yearly probability of a democratic transition is close to 0. By contrast, for a highly equal society, that prob- ability rises to over 0.10--even for very low levels of development. These results (jointly with parallel estimations on the probability of a democratic breakdown) can be employed to calculate the long-term political dynamics of any country given its underlying conditions. To exemplify this, consider the regime path of several countries with differing levels of economic development and inequality: country A has a per capita income of $2,000 and a high level of income inequality (a Gini index of 0.65); country B has a per capita income of $2,000 yet little inequality (a Gini index of 0.35); country C has a per capita income of $6,000 and high inequality; finally, country D is both developed and relatively equal. Assuming that all countries start out as author- itarian, their political dynamics are sharply different. The level of inequality shapes the chances of making a democratic transition. Whereas in the two relatively equal coun- tries (B and D) a democratic regime emerges after 15 to 20 years of dictatorship, in the unequal countries the authoritarian regime remains in place for more than 90 years (up to 120 years in country A). Once democratic regimes are established their stability levels differ, driven by the level of development. In the two developed coun- tries, democracy lasts forever. In the two underdeveloped countries, democracy collapses relatively quickly, generally after 1 or 2 years. Then, again, in the unequal 102 | CARLES BOIX TABLE 3. a. Predicted probability of transition from authoritarianism to democracy by per capita income and income inequality, 1950­90 Per capita income (1985 $) Gini index 1,000 3,000 5,000 7,000 9,000 11,000 70 0.000 0.000 0.000 0.001 0.001 0.002 60 0.001 0.001 0.002 0.003 0.006 0.009 50 0.004 0.006 0.010 0.015 0.022 0.033 40 0.016 0.023 0.034 0.049 0.069 0.094 30 0.052 0.072 0.097 0.129 0.168 0.213 20 0.134 0.174 0.220 0.272 0.330 0.393 b. Predicted probability of transition from authoritarianism to democracy by size of agricultural sector and income inequality, 1950­90 Share of agricultural sector over GDP Gini index 0 20 40 60 70 0.000 0.000 0.000 0.000 60 0.000 0.000 0.000 0.000 50 0.005 0.003 0.002 0.001 40 0.033 0.024 0.018 0.013 30 0.143 0.115 0.091 0.071 20 0.383 0.333 0.287 0.243 c. Predicted probability of transition from authoritarianism to democracy by size of oil sector and income inequality, 1950­90 Fuel as percentage of exports Gini index 0 20 40 60 80 100 70 0.000 0.000 0.000 0.000 0.000 0.000 60 0.002 0.000 0.000 0.000 0.000 0.000 50 0.009 0.000 0.000 0.000 0.000 0.000 40 0.037 0.003 0.000 0.000 0.000 0.000 30 0.111 0.013 0.001 0.000 0.000 0.000 20 0.256 0.048 0.004 0.000 0.000 0.000 Source: Simulation taken from Boix (2003), tables 2.3 to 2.5. INEQUALITY AND POLITICAL REGIMES | 103 FIGURE 2. Civil Wars before 1945, Worldwide 0 CHN COL 10 CHN CHN AFG COL RUS HON COL PER CHN IRN COL RUM PER 20 FIN SAL SPN SPN ARG MEX ARG ARG MEX USA 30 HUN FRN CHL FRN industrialization URU GRC 40 SPN SPN and 50 AUS 60 70 urbanization 80 verageA 90 100 100 90 80 70 60 50 40 30 20 10 0 Percentage of family farms Source: Author's depiction. country, the authoritarian regime lasts for another century while in the equal nation a new transition occurs after less than 20 years of dictatorship. Table 3b considers the effect of inequality and agriculture. Again, a smaller Gini facilitates the introduction of democracy. Finally, table 3c observes the interaction of inequality and the weight of oil. The presence of an oil economy acts as a heavy con- straint on the possibility of democratization. For economies in which fuel exports represent more than one-third of all exports, the probability of democratic transi- tions is 0, regardless of the distribution of income. Political Violence Figures 2 through 5 show the distribution of civil war onsets and guerrilla warfare onsets across the world by the average level of industrialization and urbanization (on the x-axis) and the percentage of family farms (on the y-axis).11 Figure 2 plots all the country-year observations from 1850 to 1944 (the black dots) and the cases in which a civil war (as defined by the Correlates of War project) started, marked with the abbreviated name of the country in which it took place. The graph shows considerable dispersion in the combination of how industrialized coun- tries are and how unequal their agrarian sectors are.12 Latin American countries are clustered in the upper right-hand corner. The dotted line in the upper left-hand area corresponds to the United Kingdom; it traces a story of continuous industrialization without much change in a considerably concentrated (yet progressively more irrele- vant) agrarian sector. A symmetrically opposite case is that of Norway--where family farms accounted for 64 percent of the cultivated land in 1850 and about 84 percent in 1939, while industrialization remained sluggish. Most civil wars occur in countries where the agrarian sector is still dominant and land is distributed unequally (basically within the triangle to the right of a diagonal that goes from no industrialization and 104 | CARLES BOIX FIGURE 3. Civil Wars after 1945, Worldwide 0 RWA RWA NIR YAR 10 BUI BUI UGA UGA DRV SUD AFG CAM IND CHN SOM 20 SUD CAM MYAPAK NIGPHI PAK GUA GUACOL NIG DOM BOL 30 MYA GUA IRQ SRI NIG PAR SRI PHI industrialization SAL YPR 40 LEB NIC LEB and IRN IRNOR NIC 50 TUR CUB PER RUM YGS IRQ 60 COL TAJ ARG LIB 70 urbanization LIB GRG 80 verageA 90 100 100 90 80 70 60 50 40 30 20 10 0 Percentage of family farms Source: Author's depiction. less than 50 percent of the land to middle levels of industrialization with no family farms at all. Although the American Civil War falls outside that area, this is proba- bly the outcome of using statistical averages for the entire country since the war was (arguably) linked to the high inequality-high specificity nature of the American South in the mid-19th century (Moore 1966). The Austrian civil conflict of 1934 and the Greek War of 1944 are the only conflicts that fall outside the boundaries of most the- oretical expectations. Figure 3 reproduces the distribution of country-years (dots) and civil wars (letters) after 1945. Again, most civil conflicts fall squarely within the area defined by high inequality and high asset specificity. Several cases that are closer to the middle (that is, farther away from the upper right-hand corner) have considerable oil resources, so conflict there may be related to asset immobility. Figures 4 and 5 depict the distribution of guerrilla warfare before and after 1945. The occurrence of guerrillas is more widespread than systematic civil wars, but the pattern is still similar. Violence is heavily concentrated in unequal agrarian economies, including those of: pre-World War II Southern and Eastern Europe, Czarist Russia, Central and South America, Cuba, mid-20th century China, Vietnam Cambodia, and most of the sub-Saharan and Middle Eastern states. Table 4 presents results on a multivariate analysis of the factors that may influ- ence the eruption of civil war and guerrilla warfare to control for the impact of important variables in the literature, such as per capita income, population, political regime, geography, and ethnic and religious composition. Two models were run for each type of violence. The first includes data prior to 1950 (beginning with 1850 for civil wars and 1919 for other violent events); this dataset maximizes the number of observations, which range from 8,300 to 6,200 INEQUALITY AND POLITICAL REGIMES | 105 FIGURE 4. Guerrilla Onsets before 1945, Worldwide 0 10 SAU AFG CHN ALB HONHON HON BUL ALB RUS CHN RUS BUL BUL BUL RUS CHN RUM TUR IRN PER COL COL 20 POL POL RUM DOM FIN NIC PER ECUMEXMEXPAR B MEX FIN PAN MEX SAF 30 HUN POR VEN POR POR VEN POR industrialization IRE IRE GRC JPN SPNITA SPN 40 ITA and NOR CHL AUS FRN CHL CHL CHL 50 GMY ARG GMY GMY 60 USA 70 urbanization UK 80 UK verageA 90 100 100 90 80 70 60 50 40 30 20 10 0 Percentage of family farms Source: Author's depiction. FIGURE 5. Guerrilla Onsets after 1945, Worldwide 0 NEP NEP NEP LES RWA LES 10 THI YAR YAR YAR CAM UGA BNG THIUGA THI YUGCAMTHI YAR YUGBNG 20 CAM BUL CAM VIE BOL 30 POL PAR BOL YGS POL YGSROK POL POL PAR ROK TUR industrialization SRI PER 40 LEB LEB VEN GRC IRE PER VEN and CON TUNMEX TUR PHI MOR PER 50 IRE TUR EGY IRQITA IRQ FRN SPN SWZ FRNFRN ALG SPN NIC 60 FRNFRN DEN GDR ITA SWZ DJI ITA SPN GRC FRN JOR IRQ FRN ITAITA FRN MEXIRQ SPN 70 ITA ARG USAUSA CANBEL IRQ VEN urbanization BEL USA COL URU BEL CFR USA ISR BUL USA CANDJI ISR ISR CHL USA ISR ISR SPN 80 RUS SPN UKUK UK UK CHL UK UK verageA URU 90 LEB ISR UKISRISRISR UK 100 100 90 80 70 60 50 40 30 20 10 0 Percentage of family farms Source: Author's depiction. country-years. This model also includes the lagged value of the dependent variable, the percentage of family farms, the index of occupational diversification (that is, the average of industrialization and urbanization), the interaction of the two previous variables (which should capture the theoretical expectations of the model), the log value of per capita income, and the log value of population and democracy. The sec- ond model includes data only for the second half of the 20th century--but adds the 106 | CARLES BOIX TABLE 4. Civil Wars and Guerrilla Movements Civil War Guerrilla Warfare 1850­1994 1950­97 1919­97 1950­97 Constant ­2.046*** ­4.131*** ­2.787*** ­2.673*** (0.698) (1.009) (0.378) (0.545) Lagged dependent variable 2.953*** 3.505*** 1.382*** 1.233*** (0.084) (0.108) (0.050) (0.062) Percent of family farms t­1 0.005 0.021*** 0.004* 0.008** (0.004) (0.006) (0.002) (0.004) Index of occupational 0.005 0.013* 0.005** 0.010*** diversification t­1 (0.005) (0.007) (0.003) (0.004) Percent of family farms t­1 * ­0.026** ­0.043*** ­0.023*** ­0.027*** Index of occupational (0.011) (0.013) (0.005) (0.007) diversification t­1 Log (per capita income) t­1 ­0.154 ­0.107 ­0.024 ­0.083 (0.094) (0.121) (0.049) (0.070) Log (population) t­1 0.096*** 0.125*** 0.166*** 0.107*** (0.025) (0.045) (0.016) (0.023) Democracy t­1 0.142 0.242** 0.172*** 0.133* (0.100) (0.129) (0.057) (0.072) Ethnic fractionalization 0.641 1.726*** (0.940) (0.501) (Ethnic fractionalization) ­0.288 ­1.585*** (0.985) (0.552) Religious fractionalization 0.913 0.173 (1.006) (0.560) (Religious fractionalization) ­0.598 0.046 (1.227) (0.668) Percentage of Muslims 0.005* ­0.000 (0.002) (0.001) Percentage of Catholics 0.003 0.002* (0.002) (0.001) Percentage of Protestants ­0.002 ­0.008** (0.005) (0.003) Growth rate t­2 to t­1 ­0.184 ­0.068 (0.747) (0.438) Log (percentage mountainous) 0.051 0.072*** (0.045) (0.025) Oil exporter ­0.031 ­0.096 (0.179) (0.099) Noncontiguous state 0.283** 0.424*** (0.143) (0.077) Number of observations 8322 4207 6155 3906 Log likelihood ­649.42 ­357.02 ­2029.37 ­1313.62 Probability >chi2 0.0000 0.0000 0.0000 0.0000 Pseudo R2 0.5827 0.7909 0.2151 0.2629 Source: Boix 2003. Estimation: Probit Model. Standard errors in parenthesis. *** p<0.01; ** p<0.05; * p<0.10. INEQUALITY AND POLITICAL REGIMES | 107 following controls: ethnic fractionalization, religious fractionalization (percentage of Catholics, Muslims, and Protestants), the lagged growth rate, the log of the percent- age of the territory that is mountainous, a dummy for oil exports, and a dummy that captures whether or not the state is composed of noncontiguous territories. The esti- mation is done through probit analysis. As expected from the theory, the interactive term of family farms and nonagrar- ian assets is statistically significant and has a substantial depressing impact on the occurrence of civil wars. A simulation of the results is shown in table 5. In countries with either less than 20 percent of the land held by family farms or with an average urbanization and industrialization below 25 percent, the probability of a civil war onset is greater than 5 percent over the course of a five-year period. As both land equality and industrialization increase, the probability of a civil war declines quickly. In countries where family farms control more than 50 percent of the cultivated land and average industrialization and urbanization are also greater than 50 percent, the probability of a civil war occurring over a period of five years drops below 1 percent. Population increases the probability of a civil war. Democratic regimes slightly increase the probability of civil war onsets (although the coefficient is statistically sig- nificant only in column 2 of table 1). Per capita income becomes statistically insignif- icant: industrialization and land inequality now capture the part of the variation that other studies modeled through per capita income.13 Ethnic fractionalization and reli- gious fractionalization are not statistically significant. The proportion of Muslims has a small positive effect on civil wars. And contrary to a substantial literature, nei- ther economic crises nor oil resources induce more civil wars. Geography has a par- tial effect: the coefficient of mountainous terrain is positive but not significant; by contrast, noncontiguous states have a stronger chance of facing civil wars. Columns 3 and 4 of table 4 examine the covariates of guerrilla warfare, and those results parallel those for civil wars. The effect of inequality and asset specificity is very similar in statistical significance--and substantial in size for both guerrilla and civil war. Table 6 simulates the probability of a guerrilla movement over a five-year period. For low levels of family farms and industrialization, the probability fluctu- ates around 20 percent. In fact, it increases with each value separately, which may TABLE 5. Predicted Probability of Civil War Onset over 5 Years by Size of Agrarian Sector and Landholding Inequality Index of occupational Share of family farms over total cultivated land (percentiles) diversification 10 30 50 70 90 10 0.07 0.07 0.08 0.09 0.10 30 0.06 0.05 0.04 0.03 0.03 50 0.05 0.03 0.02 0.01 0.01 70 0.04 0.02 0.01 0.00 0.00 90 0.04 0.01 0.00 0.00 0.00 Source: Boix 2003. Note: Simulation based on table 4, column 1. 108 | CARLES BOIX TABLE 6. Predicted Probability of Guerrilla Warfare Onset over 5 Years by Size of Agrarian Sector and Landholding Inequality Index of occupational Share of family farms over total cultivated land (percentiles) diversification 10 30 50 70 90 10 0.20 0.22 0.24 0.25 0.27 30 0.22 0.20 0.18 0.16 0.15 50 0.24 0.18 0.14 0.10 0.07 70 0.26 0.17 0.10 0.06 0.03 90 0.28 0.15 0.08 0.03 0.01 Source: Boix 2003. Note: Simulation based on table 4, column 3. capture the fact that societies with family farms may organize more easily. Nonethe- less, as both variables increase, the probability of conflict drops, falling to below 10 percent at the median values of both variables and to below 5 percent for values common in developed countries. Conclusions This paper has argued that both the distribution and nature of economic assets and the political balance of power among different social groups are key factors in the emergence of democracies and authoritarian regimes. Stable democracies occur when inequality is low or wealth is either mobile or difficult to tax. Excessive differences in the economic situation among the rich and the poor push the former to restrict the franchise to avoid the redistributive consequences of a fully democratic system, unless capital mobility restrains the ability of the poor to expropriate this wealth. Similarly, this paper has shown that, given some uncertainty about the power of the repressive capacity of the state (and provided that inequality and asset specificity are high), political agents become more prone to engage in political conflicts--and some countries experience systematic civil unrest. This model has considerable empirical purchase on the history of political development. It accounts for both the antidemo- cratic nature of landlords in agrarian settings and the mostly benign role of the com- mercial and industrial bourgeoisie in the process of extending universal suffrage in Europe. It explains why economic development, by equalizing conditions and reduc- ing the weight of immobile assets, fosters democracy. And, at the same time, it explains why the ruling elites of most wealthy oil producers resist the liberalization of their government's institutions. The theory of political regimes laid down so far opens up a key question that is addressed briefly here: How did the underlying economic and political conditions that trigger a political transition to democracy come into place? The achievement of conditions favorable to democracy is mainly the result of long-run political and eco- nomic processes (and sometimes externally imposed reforms). In agricultural, pre- industrial societies, inequality arises as a result of the expropriatory actions of INEQUALITY AND POLITICAL REGIMES | 109 authoritarian bandits, be they the Bourbons, the Sauds, or the Somozas. In exchange for the protection they may offer against similar warlords, bandits exact a heavy price in the form of wealth accumulation at the expense of vassals and peasants. To sustain their political and economic advantage, they repress their contenders and crush any popular revolts. Hence equality of conditions and democratic procedures only emerge in frontier societies and in farmer communities that, for geographic rea- sons (or perhaps due to the nature of their production endeavors), forestall the emer- gence of predators. In unequal societies, the capacity of enlightened domestic elites to implement distributive reforms that reduce inequality and subsequently foster democracy is very limited. And the reason is simple. If these societies could pass a reform (that is, if they were able to overcome resistance to change), this would mean that a democratic arrangement would have been possible in the first place--and that the reform was unnecessary to begin with. Those that oppose democracy oppose it for its distributional consequences--and have the same incentives to block any reformist program that is directed toward creating the social preconditions needed for a successful democracy. In other words, a redistributive strategy (typically a land reform) can only be imposed from abroad; and this is how one should interpret the long-run beneficial impact of the postwar reforms imposed by the United States in Japan, the Republic of Korea, and Taiwan. The only other path to democracy comes from industrialization, which leads to the systematic generation of new inventions and products and changes the nature and distribution of wealth. The mobile nature of these new assets makes their holders more sanguine about the consequences of democracy. Moreover, the growing value of human capital facilitates the spread of wealth to new social segments and equal- izes conditions. As assets become more mobile and their distribution more equal, uni- versal suffrage is introduced. Provided there are no exogenous shocks, democracy should then become a self-sustaining equilibrium. By engendering political competi- tion and policy transparency, representative institutions reduce the ability of politi- cians to expropriate other political actors. Moreover, democracy tends to foster higher levels of equality through extensive education and redistribution schemes. Notes 1. This paper is based on work developed in Boix (2003, 2004). 2. For a discussion on democracy as a self-enforcing mechanism, see Przeworski (1991), Weingast (1997), and Boix (2003). 3. These conditions are formally proved in Boix (2003), chapter 1. 4. In fact, all real-world income distribution matches this assumption. 5. Without this restriction, the median voter would always set the tax equal to 1, and hence expropriate from individuals with an income equal to or larger than the average income. 6. In fact, this mistake explains why the Meltzer-Richards model inadequately explains the level of taxes in the empirical arena. 7. As the costs of repression shift, the probability of an authoritarian or a democratic regime varies. 110 | CARLES BOIX 8. The level of social heterogeneity may vary also according to religious and ethnic prefer- ences; as fragmentation along those lines increases, heterogeneity should go up as well, and the chances of democracy should decline. Although religious and ethnic heterogene- ity have a negative effect on democratization and democratic consolidation, their impact is much lower than income inequality (Boix 2003, pages 83­84). 9. See Amemiya (1985, chapter 11) for the estimation and asymptotic properties of this dynamic probit model. For a previous application of the model to democratic transitions, see Przeworski and others (2000). 10. The regressions include the following control variables; per capita income; percentage of the population of each country that belongs to Catholicism, Islam, and Protestantism; level of religious fractionalization; level of ethnic fractionalization; and lagged growth rate. 11. This analysis employs data on civil wars from two datasets: the first from the Correlates of War (COW) project developed by Singer and Small (1993), which includes data from 1816 through 1992; and the second built by Fearon and Laitin (2003), which spans 1945­99. The data on guerrillas, from Banks (1997), cover the period 1919­97. 12. For purposes of this discussion (to investigate the origins of violence) the distribution of agrarian property measure is adequate, since it is only inequality in the agrarian sector that matters: As assets become less fixed or less specific, the incentives to engage decline-- even in relatively unequal societies. 13. Dropping the index of occupational diversification (alone and in the interaction) makes per capita income statistically significant and doubles the size of its coefficient. References Amemiya, Takeshi. 1985. Advanced Econometrics. Cambridge, MA: Harvard University Press. Banks, Arthur S. 1997. "Cross National Time Series: A Database of Social, Economic, and Political Data." http://www.databanks.sitehosting.net. Boix, Carles. 2003. Democracy and Redistribution. New York: Cambridge University Press. ------. 2004. "Political Violence around the World." Paper prepared for the Conference on Order, Conflict, and Violence at Yale University, April 30­May 1, 2004. Boix, Carles and Sebastian Rosato. 2001. "A Complete Data Set of Political Regimes, 1800­1999." Department of Political Science, The University of Chicago, Chicago, IL. Bourguignon, François, and Christian Morrisson. 2002. "Inequality among World Citizens: 1820­1992." American Economic Review 92 (September): 727-45. Deininger, Klaus, and Lyn Squire. 1996. "New Data Set Measuring Income Inequality." The World Bank Economic Review 19 (3): 565-91. Fearon, James D., and David Laitin. 2003. "Ethnicity, Insurgency, and Civil War." American Political Science Review 97 (February): 75-90. Heston, Alan, Robert Summers, and Bettina Aten. 2002. Penn World Table Version 6.1. Cen- ter for International Comparisons at the University of Pennsylvania (CICUP). Maddison, Angus. 1995. Monitoring the World Economy, 1820­1992. Paris: Organisation for Economic Co-operation and Development. Meltzer, Allan H., and Scott F. Richards. 1981. "A Rational Theory of the Size of Govern- ment." Journal of Political Economy 89 (5): 914-27. INEQUALITY AND POLITICAL REGIMES | 111 Moore, Barrington. 1966. Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World. Boston, MA: Beacon Press. Przeworski, Adam. 1991. Democracy and the Market. New York: Cambridge University Press. Przeworski, Adam, Michael Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development: Political Institutions and Well-Being in the World, 1950­1990. New York: Cambridge University Press. Romer, T. 1975. "Individual Welfare, Majority Voting, and the Properties of a Linear Income Tax." Journal of Public Economics 4 (February): 163-85. Singer, J. David, and Melvin Small. 1993. "The Correlates of War Project: International and Civil War Data, 1816­1992." Correlates of War Project, The University of Michigan, Ann Arbor. Statistical Office of Finland. Various years. www.stat.fi. Vanhanen, Tatu. 1997. Prospects of Democracy: A Study of 172 Countries. London: Routledge. Weingast, Barry R. 1997. "The Political Foundations of Democracy and the Rule of Law." American Political Science Review 91 (June): 245-63. Income Inequality, Economic Transformation, and Growth in Russia EVGENY GAVRILENKOV According to the State Statistics Committee (Goskomstat) in Moscow, gross domes- tic product (GDP) in Russia grew fast in 2003 (by 7.3 percent), as did industry (by 7.0 percent). This eclipses the much slower rate of growth in 2002 (4.7 percent for GDP and 3.7 percent for industry). On the basis of these figures, GDP growth in the postcrisis years 1999­2004 comes to 38 percent, or an average of 6.7 percent per year. In the first half of 2004 growth was also high: GDP grew 7.4 percent (year-on-year), while in the second half growth slowed substantially, bringing the annual growth fig- ure well below 7 percent, which is still high compared to many other countries. So in 2004, the overall Russian GDP exceeded the 1998 level by about 47 percent--but was still 15 percent lower than in 1990. Yet private consumption reached the 1991 level by 2001 (see figure 1), and in 2004 grew by nearly 40 percent compared to 1991. FIGURE 1. Private Consumption Outstripped GDP and Exceeded the 1991 Level 160 140 Private consumption GDP 120 100 80 60 40 20 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: Data from Goskomstat, Statistical Yearbook, various issues. Evgeny Gavrilenkov is managing director and chief economist of Troika Dialog and vice director of scientific research at the Higher School of Economics in Moscow. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 113 114 | EVGENY GAVRILENKOV FIGURE 2. Change in Income Inequality and Inflation (percent) 3000 100 Inflation (l.h.s.) (left axis) 2500 Change in inequality (right axis) 80 2000 60 1500 40 cent Per 1000 20 500 0 0 ­20 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: Data from Goskomstat, Statistical Yearbook, various issues. However, the difference between the two periods (1991 and 2004) arises from the fact that income inequality increased substantially over the past 14 years. According to the Russian Statistical Committee, the gap between incomes of the poorest and richest deciles of the Russian population widened from about 4.4 times as of 1990­1, to nearly 15 times that amount by the end of 2004. The fastest growth in income inequality was experienced in the early 1990s, during the period of high inflation and privatization. In 1992, the gap between the incomes of the richest and poorest deciles widened to 8 times. In 1993, the gap widened even more (to 11.2 times) and reached 15.2 times in 1994. (Figure 2 shows the change in percentages of income inequality and inflation.) Practically all economies in transition experience rapid increase in income inequal- ity, especially during first years of reforms, while later on, the situation stabilizes. How- ever, relative to the Eastern European countries, income inequality in Russia has increased more dramatically, even though the initial, prereform, conditions were largely similar. The results of a survey on recent developments related to poverty issues in East- ern Europe and the Commonwealth of Independent States (CIS) can be seen in the Eco- nomic Survey of Europe 2004, No. 1 (UNECE 2004). This study found no strong ties between income inequality and growth. Moreover, in some cases, countries with high income inequality grew faster than the countries with lower degree of inequality. In some cases, inequality continued to widen on the back of economic recovery, which means that--apart from the degree of inequality--other more important factors affect economic growth (such as external conditions and economic policies, for example). This paper attempts to broach those issues with respect to Russia's recent growth, which, in spite of a slowdown in the second half of 2004, remains quite high and takes place on the back of high income inequality. In spite of many unresolved problems, Russia from 1999 until mid-2004 demon- strated a healthier macroeconomic performance than many countries. Equally impor- tant is the fact that Russia financed this growth primarily from its own sources INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 115 (mostly from oil export revenues) and without a massive inflow of foreign direct investment or external borrowing. External borrowing did, however, increase signif- icantly in 2003 and remained relatively high in 2004, although it decreased relative to the previous year. As a result, foreign financing in 2003­4 played a much more important role than ever before: Foreign debt of nonfinancial enterprises in Russia rose by over 22 percent from January to September 2004, and exceeded $56 billion. Not only the growth numbers in Russia were impressive in 2003 and early 2004. More important was the recognition that some structural changes had become more visible in recent years. The macroeconomic performance in 2002 (in particular the deceleration of growth in that year) already clearly indicated that Russia could no longer rely on the advantages of "easy" growth--and that a repeat of the growth pat- tern that emerged after the 1998 crisis would be impossible. A rapid rise in incomes in recent years had shifted consumer demand toward higher-quality goods that could not be produced in Russia. So domestic manufacturers throughout the market real- ized that, to compete with imports, they needed to offer better (perhaps more expen- sive) products, and therefore must invest in new productive capacities. The need to resuscitate the exhausted growth mechanism that had emerged from the 1998 crisis (based on increased capacity utilization) became obvious. Thus, greater investment activity was one of the major recent drivers of growth. Other factors also contributed to the acceleration of growth in 2003 and early 2004. Notable were higher oil prices (which caused the money supply to surge), low interest rates, and a rapid increase in domestic demand. High oil prices also were a catalyst of both higher domestic demand (consumer and investment) and economic growth in general. As a result of the changing growth model, investment activity began to rise across the board in 2003, not only in the oil and gas sector (as had always been the case). Medium-size companies oriented toward the domestic consumer market now set their sights on more aggressive growth. On the back of the liquidity surge in the financial system and low real interest rates, they sought to raise funds by issuing ruble corporate bonds and borrowing directly from domestic banks. According to the central bank, the broad monetary base in Russia expanded by more than 55 per- cent in 2003, and growth of the money supply also exceeded 50 percent. Meanwhile, nominal lending rates in 2003 dropped to around 12 percent, the level of inflation reported for the year (and the upper limit of the government's target). Thanks to the zero real interest rates, bank loans to the private sector swelled by around 45 percent and the ruble corporate bond market by over 90 percent. Overall, investment grew by 12.8 percent in 2003, and private consumption by 7.5 percent. Meanwhile, high oil prices and increased physical volumes of exports bumped up the current account balance to $35.8 billion, versus $29.1 billion a year earlier. This enabled the central bank to collect about $26 billion in international reserves and to expand the money supply considerably. As a result, the monetization of the economy (M2 to GDP ratio) exceeded 25 percent, versus only 22 percent in 2002 and 14 percent in 1999. In 2004, oil prices increased even more--bringing Rus- sia's current account balance to nearly $60 billion (over 10 percent of GDP). Both international reserves and money supply grew even more, and monetization again 116 | EVGENY GAVRILENKOV rose, exceeding 26 percent of GDP. However, this is still a very low level compared to that of most countries in both developed and emerging markets. Growing mone- tization, which takes place on the back of growing savings, is an important trait of the Russian growth model. In this regard, it looks as if the high degree of income inequality and savings of a relatively narrow group of the wealthiest economic agents became an important part of investment potential in the country. Growing savings and slowing velocity of money circulation increased capitalization of the banking sector and financial system in a broader sense, so that the role of the financial sys- tem in Russia's growth has recently increased. Meanwhile, in the second half of 2004, those trends were broken as a result of deteriorating government policies. Increased pressure on business undermined confi- dence, and investment and economic growth slowed as a result. On the back of this decrease in confidence, savings stopped growing as quickly as before, which on the one hand stimulated consumer spending (and to some extent supported economic growth) but on the other hand resulted in high inflation. Slow growth and decreased money demand were accompanied by increased taxation of the oil sector, which enabled the government to accumulate about $20 billion in its stabilization fund dur- ing a relatively short period of time. But this money did not appear in the private financial system, and the rise in monetization slowed in 2004. Impact of the Energy and Financial Sectors It is also worth mentioning that when oil prices faltered in early 2003, and the econ- omy faced lack of liquidity flow, the burden was partly taken up by increased for- eign borrowing. And due to increased foreign borrowing and certain structural changes, Russia's dependence on oil prices began to slip at the end of 2002, but is still strong and will remain strong in the future (see figure 3). Thus, it can be expected that energy exports will remain vital to the Russian economy for years to come. It would certainly take massive investment in the non-energy sector to bring about more change in the structure of Russia's exports and economy in general. This has not yet happened on a large scale, and growth rates thus still appear closely tied to the oil price. That is, the higher the price for liquid hydrocarbons, the greater the investment by oil exporters; and the more money absorbed by domestic manufactur- ing due to higher domestic demand, the higher the growth. Meanwhile, this depen- dence on oil prices slipped even more in 2004, although in a negative sense. Due to deteriorating policies, Russia grew well below potential in 2004. This dependence on commodities exports, which developed and gathered strength over several decades, will not go away overnight. And to break this dependency will require massive and efficient investment in the non-energy sectors. Russia has not yet seen such investment on a large scale, in spite of the government's intentions. The breakdown of investment by source of finance (see table 1) shows that enterprises continued to rely on their own funds, although recently less heavily than before 2003. The share of borrowed money (bank loans and loans from other enterprises) increased in 2003 and in the first quarter of 2004 compared to previous years, sug- INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 117 FIGURE 3. Growth Rates Closely Tied to Oil Price 40 Urals, $ p/bbl GDP growth, % 30 20 10 0 ­10 1998 1999 2000 2001 2002 2003 2004 Source: Data from Goskomstat, Statistical Yearbook, various issues; Troika Dialog estimates. Note: $ p/bbl, dollars per barrel. TABLE 1. Sources of Investment Financing: Own Funds Predominate, Although Companies Borrow More 1st half, 2004 2003 2002 2001 Own funds 48.6% 45.6% 45.0% 49.4% Bank loans 7.3% 5.3% 5.9% 4.4% Loans from other enterprises 6.0% 9.2% 6.5% 4.9% Budget 16.2% 18.7% 19.9% 20.4% Off-budget funds 0.7% 1.1% 2.4% 2.6% Other 21.2% 20.1% 20.3% 28.3% Total 100.0% 100.0% 100.0% 100.0% Source: Goskomstat monthly bulletins, 2001­4. gesting that the financial system was becoming increasingly important for Russia's growth. Some decline of the share of borrowed money in 2004 was a result of a dete- riorating investment climate that resulted from increased pressure on the oil giant Yukos, a banking crisis, and slowing of reforms, which increased uncertainty and contributed to the acceleration of the capital flight. That own funds were the single biggest source of fixed investment from 2001­4 means that money stayed mostly in those sectors where it was generated, contribut- ing little to economic diversification. The share of borrowings (bank loans and loans from other enterprises), although increasing (especially in 2003), remains relatively low. Generally speaking, the higher this share, the greater the economy's opportunities for diversification (provided that the borrowings are allocated to non-energy sectors). It is expected that the share of borrowed money will increase further, although one can scarcely expect borrowings to make a much higher share of Russian investment in the 118 | EVGENY GAVRILENKOV near future. Cash flows in Russia are concentrated in a few export-oriented indus- tries, while the financial system is still too weak to reallocate capital to other sectors. Its "weakness," however, does not stem from the inadequacy of financial institutions alone; there are other macroeconomic reasons, such as a low monetization rate. Low monetization constrains growth, endangers stability, and limits the choice of economic policies. It is obvious that monetization cannot be raised artificially but must be fuelled by growing confidence in economic policies, an improving invest- ment climate, and higher economic growth--as well as by stronger incentives to save money domestically rather than offshore. Figure 4 illustrates this point. Real money supply expanded much faster than GDP when the economy grew (in 1997 and from 1999 onward) and contracted much faster than GDP (during the 1998 crisis) on the back of high inflation. It is also expected that if the Russian economy continues to grow in the long run, its remonetization will continue. Two other charts (figures 5 and 6) highlight the link between monetization and economic growth in a different way. Since Russian economic growth resumed in the late 1990s, purchasing power parity-based GDP per capita has risen from about $6,000 to $8,000, and monetization has increased from 14 percent to approximately 25 percent of GDP. This trend is expected to continue beyond 2004, with monetiza- tion growing by about 2 to 3 percent of GDP annually for some years to come. Finally, there are the conclusions of cross-country analysis. The richer the country, the higher is its monetization rate. Increasing monetization is good for the economy. First of all, as mentioned earlier, it translates into higher capitalization of the banking sector, which in turn facilitates loan issuance. Historical data show that, in real terms, outstanding credits have grown as fast as the monetization rate, with deposits keeping up at much the same pace. FIGURE 4. Real Money Supply (Average Annual) Grows Faster than Economy (percent) 40 GDP growth 30 Real M2X growth 20 10 0 ­10 ­20 ­30 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: Data from Goskomstat, Statistical Yearbook, various issues; Central Bank (www.cbr.ru). INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 119 FIGURE 5. Economic Growth Fuels Monetization 28 2004 26 2003 24 2002 entc 22 2001 per, 20 1997 18 M2X/GDP 1998 2000 16 1996 1999 14 1995 12 5,000 5,500 6,000 6,500 7,000 7,500 8,000 8,500 9,000 PPP-based GDP per capita, $ Source: Data from Goskomstat, Statistical Yearbook, various issues, 1995­2004; Central Bank (www.cbr.ru). Note: PPP, purchasing power parity. FIGURE 6. Monetization and Economic Development in Selected Countries 140 Japan 120 100 UK cent USA per, 80 Czech Republic Slovak Republic 60 Korea M2X/GDP Poland 40 Bulgaria Slovenia Hungary Romania Mexico 20 Ukraine Russia `04 Russia `99 0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 GDP per capita, $ PPP Source: Data from Goskomstat, Central Bank, OECD, and IMF (2002). Note: PPP, purchasing power parity. 120 | EVGENY GAVRILENKOV The Impact of Consumption and Saving Apart from increased investment activity, Russia's strong economic performance over the past few years has largely been the result of steady consumer demand as well. Exports also expanded, so the growth was quite balanced. Private consumption did fall in 1999, in the aftermath of the 1998 crisis, but it recovered rapidly and has since 2000 seen an annual growth rate in excess of 7 percent. The fastest growth was recorded in 2001 (9.5 percent) and in 2004 (about 12 percent). As mentioned ear- lier, by 2004 consumption grew fast--and was 34 percent above the 1998 level, slightly less than the growth rate in GDP over the same period (38.3 percent, due to rapid growth in investment). A deep contraction in real incomes in 1999 also con- tributed to a fall of consumption in that year and to a slightly slower cumulative growth of consumption over the postcrisis period. The consumer market is also expected to do well beyond 2004. In spite of political problems in Russia (which became obvious in 2004), invest- ment in productive capacity continues to grow faster than GDP--and is likely to do so for another couple of years. However, once the trend is eventually broken, private consumption will have a greater direct effect on the country's economic growth. As a proportion of GDP, private consumption is still quite low in Russia, at just over 50 percent. This has been the case for some years now, and many emerging market and transition economies boast a higher figure. The anomaly stems from the fact that Russia has a strong current account balance and thus net exports (in terms of national accounts). As a result, gross national saving in Russia is much higher than elsewhere, while consumption is lower. Russia's current account and net exports as a percentage of GDP are likely to shrink in years to come, while the percentage of gross investment and consumption FIGURE 7. Private Consumption in Transition Economies and Emerging Markets 80 80 Bolivia 70 Mexico 70 RB de Bulgaria Poland Venezuela Hungary GDP of 60 Croatia 60 Brazil cent Slovakia Czech Rep. Per 50 Russia 50 China Malaysia 40 40 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 GDP per capita, $ Source: IMF (2002), Troika Dialog estimates. INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 121 continue to grow. A continuation of this would move Russia up and to the right- hand side on figure 8. At the same time, many developed market economies report a level of private con- sumption even lower than Russia's (as a percentage of GDP). Countries such as Den- mark, Finland, the Netherlands, and Sweden show comparable figures, while New Zealand and the United States show a far higher result than the Scandinavian coun- tries and continental Europe in general. This raises the question: Why should Russia want to score a higher private consumption-to-GDP ratio? Can Russia not join its Scandinavian peers on the side of expected economic growth alone? This scenario is probably unlikely. For this reason, one needs to look more closely at total consumption (private plus public). Reinsert public consumption into the formula, and most countries differ little-- with the notable exception of the United States, which enjoys massive inflows of for- eign capital. The countries with a low level of private consumption thus show a higher level of public consumption as a percentage of GDP; and these are the countries with well developed public sectors, a high degree of social protection, and strong social safety nets. Here, Russia again loses out to developed market economies (figure 9). So, the expectation is that total consumption will in the long run grow as a per- centage of GDP--and will do so on the back of a shrinking current account balance and falling net exports. One would not expect this to be achieved via rapid expan- sion of the public sector. On the contrary, it is private consumption that will pick up pace. Plans afoot in Russia for the reform of housing, education, and health care services suggest that people will have to pay more for these, and the proportion of unpaid services will diminish. The recent move toward monetization of benefits goes in line with this, and as a result, the services sector will grow faster than manufac- turing (as has been the case in recent years). FIGURE 8. Private Consumption in Russia and Selected Developed Countries 80 80 70 USA 70 Greece GDP UK of 60 Korea Australia Italy 60 New Zealand cent Canada Japan Slovenia Per 50 Russia Finland Netherlands 50 Denmark Sweden 40 40 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 GDP per capita, $ Source: IMF (2002), Troika Dialog estimates. 122 | EVGENY GAVRILENKOV FIGURE 9. Total Consumption in Developed Countries 90 90 USA UK Greece 80 Italy 80 GDP New Zealand Canada Sweden of Australia Japan Slovenia Denmark cent 70 Korea Finland Netherlands 70 Per Russia 60 60 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 GDP per capita, $ Source: IMF (2002), Troika Dialog estimates. TABLE 2. Annual Average Income per Population Quintile, dollars 1997 1998 1999 2000 2001 2002 2003 1 565.1 375.0 242.7 282.6 353.5 422.6 559.6 2 1,023.0 656.3 420.6 506.7 656.6 784.8 1,029.3 3 1,480.9 937.6 598.6 735.7 972.3 1,162.1 1,528.9 4 2,172.7 1,343.9 853.4 1,067.1 1,439.4 1,720.5 2,268.4 5 4,501.3 2,937.7 1,929.2 2,280.3 2,891.5 3,456.0 4,606.7 Total 1,948.6 1,250.1 808.9 974.5 1,262.7 1,509.2 1,998.6 Source: Goskomstat, Statistical Yearbook, various years, 1997­2003. Faster growth in the services sector is also only natural, as growth in incomes leads to a further shift in consumer demand--from relatively inexpensive goods toward higher-quality goods and services, something that has already happened in Russia in recent years. Analysis of annual incomes and how they have changed over time shows that people there were earning more (on average) by 2003 than before the cri- sis (table 2). Only the bottom 20 percent of the population continued to lag behind the 1997 level, meaning that income inequality widened a little. This has been the result of government policy: Although public sector wages have been on the rise, the public sector has not been the subject of any restructuring. Employee numbers in this sector have risen, as has the wage per employee--but not as quickly as the govern- ment had hoped. However, in the private sector, the wage per employee has grown more rapidly, while the number of those who are employed has seen less aggressive growth. Indeed, some sectors have even seen their workforce contract. On the back of increased dollar-denominated incomes, which in 2003 overtook the precrisis level, private consumption also grew faster, up 31 percent over 1997­2003. In real terms, ruble-denominated incomes have grown even faster than INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 123 TABLE 3. Structure of Household Consumption by Income Quintile Food Total Cash Natural form Non-food Alcohol Services In kind 1 64.3% 50.4% 13.9% 18.3% 1.3% 15.9% 0.2% 2 61.2% 48.7% 12.5% 21.3% 1.7% 15.4% 0.4% 3 56.4% 46.0% 10.4% 25.5% 2.0% 15.5% 0.6% 4 48.3% 40.5% 7.8% 32.8% 2.0% 16.3% 0.6% 5 38.7% 32.8% 5.9% 42.3% 2.3% 16.1% 0.6% Total 48.3% 39.9% 8.4% 33.1% 2.0% 16.0% 0.6% Source: Goskomstat data (2003). Note: "Natural form" refers to goods produced and consumed by households. dollar-denominated incomes (up 12.9 percent); the reasons for the difference in trends are the 1998 devaluation and subsequent gradual appreciation of the ruble. Private consumption, meanwhile, has grown much faster than real incomes--at the expense of savings--and due to the fact that household consumption in "natural form" (that is, goods produced and consumed by households) has comprised a sub- stantial part of total food consumption over the years. A review of the structure of household consumption by income groups shows quite clearly that consumption in "natural form" in the high-income quintile is a far smaller share of the total than in the low-income groups (table 3). As is also natural, the high-income quintile spends more on nonfood products than the low-income quintiles. If the fundamentals remain strong and Russia pursues a reasonable long-term eco- nomic policy--one that promotes economic growth--the natural conclusion to be drawn is that demand for nonfood products and services will continue to grow, in both low and high-income groups. At the same time, there is unlikely to be substan- tial growth in the demand for food, since the Russian population is declining. This scenario will be accompanied by a dramatic change in the structure of demand. With incomes growing, one would (as mentioned earlier), expect to see rapid growth in demand for high-quality goods, including foodstuffs. This change is, in fact, already underway. Demand for the cheapest staple foodstuffs, such as bread and sugar, has declined since 1999, while demand for foods high in protein, such as meat and eggs, has risen (table 4). A similar trend has been seen in nonfood goods. Demand for finished goods has grown rapidly, while demand for fabrics, for example, has declined. Up until the 1998 crisis, the vast majority of finished products were imported. Since then, how- ever, domestic production in Russia has been catching up. In spite of the ongoing real ruble appreciation, a natural and inevitable phenomenon, local output of consumer goods is expected to grow more important. This will prompt rapid growth in the retail trade, while mergers among trading companies will lead to the rise of strong retail chains. Wholesale trade and logistics will also be a growth industry. The con- sumer boom in Russia is here to stay. It is worth mentioning here that income inequality in Russia could be reduced, if not the major drawbacks in its social policy, especially with regard to pensioners. 124 | EVGENY GAVRILENKOV TABLE 4. Growth of Retail Outputs in Russia Indexes (1990=100) Retail (1990 = 100) 1991 1999 2003 Food 98 84 115 Meat 98 70 95 Fish 79 88 98 Cheese 79 115 221 Eggs 100 58 73 Sugar 72 56 66 Bread 108 64 63 Alcohol 104 122 167 Non-food 93 88 130 Cotton 65 27 25 Wool 75 69 79 Silk 77 62 40 Knitwear 93 152 205 Sewn products 88 86 104 Footwear 89 99 142 Perfume 84 106 180 Cars 81 106 136 Tobacco 130 127 171 Total 95 87 124 Source: Data from Goskomstat, Statistical Yearbook, various issues. Following the 1998 crisis, wages and pensions in the country grew largely at the same rate, while more recently, in 2003­4, this trend was broken. Real wages grew (on average) about 11 percent per year in 2003­4, while real pensions grew 4.5 percent in 2003 and about 5.5 percent in 2004. This widened income inequality and reduced consumer demand from low-income groups, since average pensions in Russia remained at around $70 at the end of 2004. It is also important to note that real pen- sions grew at a much slower pace than retail sales more recently (see figure 10). All of these changes obviously created the potential for social unrest. Meanwhile, given that average nominal pensions in Russia are so low (about 28 percent of the average nominal wage), pensioners are more sensitive to the rapid growth of housing tariffs and utility charges than are the employed. If payments of this type were deducted from the average nominal pension in Russia, it would appear that real growth of the remaining part of that pension ("disposable" pension) could be close to zero in recent years. Sluggish demand from low-income groups has also limited the growth of the Russian economy, since the demand from these groups is not for imports but for relatively cheap domestic products. A move to narrow the income inequality in Russia by raising pensions appears to be inevitable. In recent years, the government has increased spending on state admin- istration, national security, the legal system, and law enforcement much more rap- idly than on other budget categories, including social policy. Moreover, this rise in spending has not been accompanied by any real restructuring of those sectors, which INCOME INEQUALITY, ECONOMIC TRANSFORMATION, AND GROWTH IN RUSSIA | 125 FIGURE 10. Retail and Pensions, Year-on-Year Percent 14 Retail 12 Pension 10 8 6 4 2 0 1st quarter 2nd quarter 3rd quarter 4th quarter 1st quarter 2nd quarter 3rd quarter 2003 2003 2003 2003 2004 2004 2004 Source: Data from Goskomstat monthly statistical bulletins (2003­4). accounts for its high inefficiency. But if the expected rise in pensions is not accompa- nied by active restructuring of the public sector, economic growth may not acceler- ate. And the rise in nominal pensions may remain only nominal, not real. Conclusions Income inequality in Russia has widened dramatically since 1991, as a result of high inflation in the early 1990s, the privatization model that was chosen, and the gov- ernment's economic policy in general. This degree of inequality reduces potential for growth of domestic consumption, yet the country's investment potential is high due to highly concentrated incomes. The growth model based on higher capacity utiliza- tion that emerged after the 1998 crisis has been exhausted, leading to greater depen- dency than ever before on investment activity. Rapid economic growth can be secured in the coming years only in an environment of diversified investments and a high rate of growth for investment in productive capacities. Private consumption will also grow quickly, in tandem with growing incomes. References Central Bank of the Russian Federation (www.cbr.ru/statistics). IMF (International Monetary Fund). Various years. International Financial Statistics. Goskomstat (State Statistics Committee, Russian Federation). Various years. Statistical Yearbook. ------. Various issues. Monthly statistical bulletins (www.gks.ru). Troika Dialog estimates. Various years. Author's calculations, published in Troika's desk notes and monthly economic bulletins (www.troika.ru). UNECE (United Nations Economic Commission for Europe). 2004. Economic Survey of Europe 2004, No. 1. Geneva: UNECE. Justice and Transition VLADIMIR GLIGOROV Social conflicts are not particularly characteristic for transition economies--at least not open conflicts such as mass strikes, demonstrations, riots, or massive economic disruptions. This observation leads to the hypothesis that issues of allocation may dominate those of distribution in post-communist developments.1 European integra- tion, however, does present post-transitional economies with a dilemma when it comes to the institutionalization of the system of social justice. The issues of social justice and fairness will, in time, emerge as important for other reasons as well. Distribution versus Allocation Not all problems of development and institutional reform are the same, and there are at least three ways to approach them.2 One problem that is perhaps typical of development is that it can be seen as charac- terized by multiple equilibria. In transition economies there is often an underdeveloped equilibrium and a growth equilibrium, which may or may not be Pareto-ranked. The problem of moving from one of these to the other is not essentially connected with the issue of social justice or that of equity of the distribution of expected benefits. Typical problems with the take-off of an underdeveloped economy are often connected with the lack of institutions, the low level of skills, and the wrong policies--that is, with the problems of allocation rather than with sharp and prevailing distributional conflicts.3 The ideal type of case is one with two equilibria: a stagnant and a growth equilib- rium that are both Pareto-ranked. In other words, in this scenario no one is "worse- off" and most are "better-off" in the growth equilibrium than in the stagnant one-- but the latter is still an equilibrium. That is, there are no internal reasons or mechanisms to move out of the stagnant equilibrium and switch to the growth Vladimir Gligorov is a senior economist at the Vienna Institute for International Economic Studies (Wiener Institut für Internationale Wirtschaftsvergleiche, WIIW) in Vienna, Austria. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 127 128 | VLADIMIR GLIGOROV equilibrium. Theoretically, this type of a developmental problem can exist due to some kind of institutional or policy failure, even if the growth equilibrium is not only more efficient but also more fair, more just, or more equitable. Here it perhaps makes sense to comment on the difference between two widely discussed concepts of fairness: those related to the distribution of opportunities and those related to capabilities. The problem of just distribution requires resources or outcomes to be determined by the influence of nature (circumstances), society (insti- tutions), and individual choice and effort. If it is assumed that people or individuals have capabilities that can be fully revealed or articulated in proper or fair social circumstances, they can be compen- sated (procedurally or in kind) for natural differences in capabilities, and if there are reward mechanisms that connect efforts and outcomes justly, then policies that make opportunities available to everybody are all that is needed.4, The responsibility for 5 the outcomes is then a matter of individual choice. This is the approach of equality of opportunities, and it fits well with the idea that the problem of development is sim- ply that of solving coordination difficulties or other institutional and policy failures. So the equality of opportunity thesis might be stated as follows: If opportunities are opened up to all (that is, equalized) a society will move to the Pareto-superior or growth equilibrium. If, however, it is assumed that these circumstances of fairness and equal opportu- nities exist, the problem of the development of capabilities remains. Those capabili- ties are, in the most fundamental sense, determined by nature, society, and individ- ual choice. Assuming that there are circumstances of equal opportunities within a society, that still may not be enough for the development of capabilities; and this is the problem usually addressed by development and growth economics. There could be multiple equilibria (that is, an equilibrium of equal opportunities) at each level of development of capabilities. A capabilities thesis might then be stated as follows: Opportunities do not neces- sarily develop capabilities. The capability approach can be seen as not too concerned by outcomes such as welfare, utility, or even equality. It may also not be primarily concerned with proce- dural justice, as is the case with the idea of equality of opportunity. This approach is mainly concerned with the ancient idea of the human life worth living that is based on a specific conception of human beings. In that sense, it is a version of outcome justice: the outcome being the truly "human" being. A second type of problem for development is one that is often found in develop- ing countries that are moving toward the growth equilibrium. This has been studied extensively, for example, in the case of Latin America. On the development path there is a set of social alternatives that are neither equilibria nor Pareto-ranked. This may be because every alternative is characterized with a certain level of social injus- tice and thus produces or sustains social conflicts. In those circumstances, the issue of equity or social justice may be central to the whole process of development. In terms of opportunities and capabilities, it is the case where lack of opportunities cur- tails the development or sheer use of capabilities. JUSTICE AND TRANSITION | 129 The third type of development problem is perhaps most characteristic of countries in transition. In the case of the post-communist transitions, sensitivity to the issues of social justice or distributional equity are not that pronounced--because the start- ing point is one of inequity and is also not an equilibrium state of affairs. Thus, tran- sition provides for the increase in efficiency and, in that sense, is seen as a Pareto improvement. Of course, there are winners and losers, but their conflict of interest does not (as a rule) have the intensity of the social conflicts usually observed when the issue of equity in distribution is predominantly at stake. Thus, post-communist transition is a move from a state of social injustice to a state of increased efficiency: The increase in efficiency (allocative justice) by itself may be seen as leading to an improvement in social justice. In that sense it may be that, in post-communist transitions, issues of allocation dominate those of distribution. Or, stated differently, allocative justice solves some problems of distributional justice and does not, at least initially, open up the usual issues of distributional justice. Justice and Conflict The typology described in the previous section assumes certain things about justice, fairness, or equity. For instance, in the case of multiple equilibria, each of these equi- libria may be just--although the concept of justice may differ from one equilibrium to another. For instance, a traditional social system may not be based on any notion of social justice, especially not on that of equality of opportunity, while the growth equilibrium may increase the importance of justice in distribution. The possibility that ideas of what is equitable may differ explains the existence of at least three types of social conflicts over social justice.6 One type of conflict is between what could be called the communal idea of justice (both in terms of allocation and of distribution) and an emerging conception of social justice. This was the theme studied extensively in the early sociological literature, where the institution of custom was contrasted with the institution of contract as the source of just allocation and distribution; this has been revived in a certain sense, with the reintroduction of the concept of social capital.7 A second type of conflict is between the concept of equity or social justice that is accepted by the society--and the performance of the actual institutions that deter- mine the distribution in a society. In that sense, it can be argued that the institutions of that society may lack legitimacy. Clearly, communist states lacked legitimacy in that regard. The predominant concept of justice went against the principles on which the main institutions of the communist societies were based. Similarly, the social con- flicts that characterize typical distributional problems in developing countries can be of this kind. A widespread concept of social justice may exist but not be embedded in the existing social institutions. Thus, social norms may conflict with the norms on which formal institutions are based. As it is sometimes stated, informal norms may conflict with the formal ones; or, in other words, morals may conflict with institutions. 130 | VLADIMIR GLIGOROV Here, there are several cases that may be of interest. One is that of social resent- ment due to the belief that institutions are not just. The Aristotelian idea that it is not necessary to know what is just in order to recognize injustice is important.8 The very removal of injustice (for example, through liberalization or increase of freedom) may be seen, at least temporarily, as an increase in social justice. Another case is one that may follow from the former one: The new state may be seen as unjust too, so there is a move from one unjust state of social affairs to another. Finally, a tyranny of social justice may exist, which is seen as being inimical to individual freedom. In the latter case, the very increase in freedom or liberty may be seen as leading to the increase of social justice. A third type of social conflict over just distribution arises when the ideas of what is just or equitable differ or change during the process of development. For example, the concept of equity based on the equality of resources may conflict with the con- cept that stresses the equality of opportunity.9 The latter may be interpreted as a ver- sion of negative freedom, while the prevailing concept may be that of positive free- dom. In other words, along the path of development--and along the path of social progress in general--a choice may have to be made between different concepts of equity or social justice. Thus, there are at least three types of social conflicts centered around the issues of equity. One is between communal and social justice. Another is between the prevail- ing social idea of equity with the performance of the existing social institutions; this is the conflict over legitimacy. The third is between existing and emerging concepts of social justice; this is the conflict over the appropriate institutionalization of one or the other concepts of social justice. In the case of post-communist countries, the initial conflict was the one over the legitimacy of communist institutions. The new emerging conflict may be, for instance, between the meritocratic concept of justice and the one based on social equality.10 The former may be that of equality of opportunity--and it may conflict with the belief in the inherent worth of equality of resources. In general, the idea that equality of opportunity is the proper concept of social justice can always be contested (even if it can be defined unambiguously), because it does not or cannot adequately take into account the two other factors that can influence the justness of a distribution. The first factor is that of luck or other external influences. The other is that of the very possibility of rewarding efforts justly, in such a way that the rewards are proportionate to the individual effort. In other words, there is a question as to whether individual contributions can be measured properly. The former is the problem of the influence of nature, which provides for equal treatment of those who are unequal. That can be seen as unjust. The latter is the issue of how the product of a common effort can be divided according to individual contributions. In general, that is not really possible. And these two problems may explain the fact that equality, in the sense of equality of resources or income, has such an important role in almost every concept of equity or fairness. JUSTICE AND TRANSITION | 131 Transitional Justice The lack of significant social conflicts in transition economies is that much more remarkable in light of the steep initial decline of production, sharp rise of unemploy- ment, significant increase in inequality, restructuring of public obligations, and com- prehensive redistribution of resources that occur. The initial decline of overall wel- fare and the increase of poverty are also well documented, although there has been some debate about the figures due to the reliability and inadequacy of the official sta- tistics. In any case, there is no doubt that a significant redistribution takes place dur- ing the course of transition.11 As argued earlier, one reason for the lack of significant social conflicts is that of allocative justice. After the initial contraction, the industrial production and the gross domestic product (GDP) have grown at a relatively high level in many transitional economies. Wages and other incomes have grown steadily, social security payments have remained significant, and the overall public sectors have remained quite large, especially if compared with other countries at a similar level of development.12 In addition, political and social changes have increased the overall freedom and have had three important consequences from the point of view of social justice: 1. Political equality has increased. This may not be so important in authoritarian or dictatorial regimes, but in a communist regime economic inequality depends upon political inequality. Therefore, the introduction of political freedoms and demo- cratic rules bring significant increases in equality. 2. Individual freedom has increased. Again, in the case of communist states, the sys- tem was designed to work against the rule of law. Therefore, the increase in indi- vidual freedom and the possibility of social association, in addition to a commit- ment to the rule of law, have increased the feeling of social justice. 3. The political power of citizens has increased. Finally, as can be seen from the fre- quent changes in government, the power of the citizens has increased, which again has an effect of the increase in social justice. Here, a more general point may be in order. Today it is increasingly argued that democracy is less important than the rule of law for growth and development. More specifically, political equality is less important than the protection of private prop- erty. This is not true, however, for post-communist transitions in Europe. Commu- nist states lacked both political equality and the rule of law (and were inimical to them both). Therefore, authoritarian implementation of the rule of law, in the restricted sense of the protection of private property, lacked legitimacy. Democrati- zation has proved to be the necessary condition for the introduction and strengthen- ing of the rule of law and for legitimizing the redistribution of resources and the increase in inequality. In other words, the increase of political liberty was a substitute for social conflicts over just distribution of resources. However, in cases in which this did not happen, violent political conflicts and significant social conflicts could be observed. 132 | VLADIMIR GLIGOROV Yet democracy is not without distributional consequences; and two stylized facts may be of interest. First, the public sector is quite large in most European economies in transition, and the structure of public expenditures indicates that the issues of equity are relevant in transition. In general, public expenditures as a percentage of GDP are higher than in other countries at similar levels of development. Also, trans- fers and subsidies are significant--and public employment is quite high. Second, changes in the government are frequent, which perhaps indicates that social dissatis- faction is channeled politically rather than through social conflicts such as strikes or demonstrations. Indeed, in a number of cases, an increase in unemployment has given rise to the change in government. Thus, issues of social justice are important. In early discussions about post-communist transition, it was assumed that issues of social justice would be important, both because of the high level of social equal- ity and because of the increased awareness of the distribution problems brought on by the process of privatization. Also, and perhaps most important, it was feared that the liberalization of the labor markets and the increase of unemployment would threaten the whole process of transition. These assumptions turned out to be mostly wrong. Even though privatization produced as many scandals as expected, the whole process proved sustainable and, indeed, proceeded with remarkable speed. The issues of justice in privatization have played a role, but not as much as was expected.13 Also, as stated earlier, increased unemployment was an important political issue but did not lead to sharp social conflicts that would place the newly emerging political and economic institutions at risk. The reasons for these developments might also be attributed to the lack of institu- tions for social conflict--and to the fact that communist institutions were not believed to be fair. There is a difference between equality and fairness in this respect. Social equality may be desirable, but the actual institutions that deliver it may not be deemed to be fair. For instance, if prices are repressed and there are shortages of con- sumer goods, then the increased equality brought about by rationing may not be seen as fair--both because of the way in which queues are administered and because of the incentives for corruption that are inherent in rationing. Liberalizing prices may in that case increase the level of inequality, but the resulting distribution may be con- sidered more fair. Liberalization of prices in general may have the same effect. Even in the case of the labor market, liberalization may be seen as leading to a more equitable distribu- tion of jobs, although the rate of unemployment may increase significantly. To see how this might work, a simple conceptual analysis may be helpful. First assume that there is no unemployment in a communist economy--but that a significant amount of employment is involuntary. Also assume that in a transition economy there is involuntary unemployment as well. The prevailing judgment of the relative fairness of the labor situation before and after transition will not depend on just the rate of unemployment or even the rate of involuntary unemployment--but on the relative rate of voluntary versus involuntary employment and the share of those who were involuntarily employed becoming involuntarily unemployed (and on whether they prefer one of these over the other). If the distribution of jobs was not truly voluntary in a particular communist economy (and the number of people who were involuntar- JUSTICE AND TRANSITION | 133 ily employed was smaller in the transition economy), then the increased unemploy- ment and the increased income inequality may be seen not as less fair but rather as an increase in social fairness. Fairness, however, is not necessarily social justice. If the unemployment is struc- tural, the perception may change in time, because the point of reference will change as well. Similarly, the perception of increased social inequality being fair may change in the absence of other reasons to justify them. The concept of social justice that a society may come to accept will then become important. Merit and Luck It has been argued by a number of authors, most recently by Alesina and Angeletos, that the concept of social justice in Europe is different from the one found in the United States. In Europe, it is believed that social justice is needed to correct for the influence of luck, nepotism, and corruption; while in the United States it is believed that institutions screen people fairly and reward them based on their merits. That may explain the fact that the welfare state is much larger in Europe than in the United States. In the case of the countries in transition, the meritocratic conception of justice may initially prevail, but it may eventually change in line with the dominant concept of justice in the European Union (EU). As a consequence, social conflicts may become more important as the countries in transition integrate into the EU.14 The case of the former German Democratic Republic may provide a good example. Unlike most other European post-socialist countries in transition, where meritocratic justice was emphasized through processes of liberalization and privatization, in reunified Ger- many a version of equality of opportunity and of income was adopted, with huge transfers flowing from the West to the East. This was perceived as neither just nor fair; it proved to be inefficient as well. Although on a smaller scale, similar problems may appear as more of the transi- tion countries join the EU. Issues of distributive justice may become more important than those of allocation of effort and income according to merit. This may not lead to the level of dissatisfaction with integration observed in Germany, because there is no place from which transfers could flow. That, however, may prove to be a prob- lem for the stability of the enlarged EU--precisely because of the frustration with the limited amounts of social and other transfers that should instead highlight the signif- icance of social justice, equality of opportunity, and the insignificance of historical legacies for the capabilities, opportunities, and welfare of individual citizens.15 Conclusions Unlike in the case of distributional conflicts, transition opens up opportunities and enables better allocation of capabilities in the context of increased political freedom. That may, in turn, dampen interest in social conflict and in distributional justice in 134 | VLADIMIR GLIGOROV general. In time, however, those should be expected to emerge, perhaps influenced by the prevailing conception of social justice in the EU, which is historically more redis- tributional and less meritocratic. Annex 1: Pareto Choices Figure A1.1 illustrates the various Pareto concepts presented in this paper, and can be used to examine some of the problems commonly associated with those concepts. A and B are two actors (persons, families or other social groups and social actors, firms, parties, states, or international organizations). They are to choose from four social policy alternatives: w, x, y, and z. It is assumed that the actors will rank these alternatives, although not necessarily in terms of welfare. The policy alternatives can be valued as "better than," "more useful than," or "larger than"--or whatever is chosen as the criterion of comparison. It is also assumed that these alternatives are social in the sense that they can be realized only if both actors act; refraining from action is considered an action as well. The actors do not have to cooperate, however; and they may experience conflict. Thus, the actors may produce, trade, contract, fight, or engage in just about every possible social relationship. Four alternatives are considered, though three are enough to illustrate the prob- lems of Pareto comparisons and choices. This is not really restrictive. Given the social FIGURE A1.1. Pareto Comparisons A w x y z B Source: Author's depiction. JUSTICE AND TRANSITION | 135 character of the problem, it is natural to assume that two individual rankings will yield at least four social possibilities: a status quo, a social improvement, and two points at which one actor is better-off and the other is worse-off. These four social states can be characterized in Pareto terms and in terms of equilibrium in the follow- ing way: Pareto-optimal states: w and z. This is because a move from either state to any other three would make one actor better-off and the other worse-off. A state is Pareto-optimal if there is no alternative state in which at least one of the actors is better-off and none is worse-off. Pareto-inferior state: y. Because y is worse than x for both actors. Pareto-superior state: x. Because x is better than y for both actors. Equilibrium: w, z and y. Because the move from the former two states to some other state is not acceptable to one of the actors. State y is the most interesting, because it can be an equilibrium even though it is a Pareto-inferior state. To see this, the preferences of the actors A and B should be clearly defined. Assuming that they prefer more to less (with P standing for preference), the actors will rank the four alternatives in this way: A: wPxPyPz B: zPxPyPw, which gives the structure of preferences that are characteristic of the so-called "prisoners' dilemma." If the two actors are to choose from the four alternatives x, w, y, and z, they will-- in accordance with the argument for the noncooperative outcome in the prisoners' FIGURE A1.2. Prisoners' Dilemma Actor A Cooperates Does not cooperate Cooperates x w Actor B Does not cooperate z y Source: Author's depiction. 136 | VLADIMIR GLIGOROV dilemma--choose y. They will do this even if they initially find themselves at x. Thus, the Pareto-inferior social state is an equilibrium, which then leads to the following characterization of x: Disequilibrium: x. Though x is Pareto-superior to y (and thus the move from y to x is a Pareto improvement), x is not an equilibrium state of affairs. The reasoning here is that, at y, both actors will prefer w or z to x and will not agree to forgo the possibility of reaching their best choice in order to settle for the second best. They cannot reach an agreement so will settle for y. The same reason- ing will destabilize x. Attempting to reach their best choices, the two actors may not agree to stay at x but move to y instead. In terms of this classification, it could be argued that a reform move from y to x is equal to a miracle; the move from y to either w or z is the canonical case of devel- opment and may be free of social conflicts if compensations can be found for the los- ing actor; and a movement from x to y is akin to a disaster. A move from w to z (or vice versa) is characteristic of social conflicts with significant equity problems. It is interesting to note that, within a prisoners' dilemma, there is an arrow prob- lem. If either w or z is removed, three ranked alternatives are left (if it is assumed that w is removed), as follows: A: xPyPz B: zPxPy. FIGURE A1.3. Arrow Triple A x y z B Source: Author's depiction. JUSTICE AND TRANSITION | 137 Agreement that is not dictatorial is impossible, because although both actors agree that x is preferable to y, they do not agree on any of the other choices. Indeed, if they were to rank the alternatives, they would end up with the social ranking that reflects only the individual ranking of one or the other of the actors--but not with a com- mon ranking. Thus, again, the Pareto-superior choice cannot, in general, be made. Both configurations of preferences, those of the prisoners' dilemma and of the Arrow paradox, are at the heart of Amartya Sen's famous Pareto-liberal paradox, which is at the foundation of his stress on capabilities rather than on opportunities.16 This can easily be seen. In the case of arrow preferences, and if one person is given a right to determine a certain social outcome, the same right cannot be given to another person--because there is room only for a dictatorial solution, and only one person can have all the rights with nobody else having any rights at all. Thus equality of opportunity understood as equal rights makes no sense within the theory of social choice. Equally, if individual rights are granted to two people facing a prisoners' dilemma, they will end up choosing an outcome that is Pareto-inferior; indeed, the distribution of preferences in the canonical case of Sen's paradox is that of the pris- oners' dilemma. Indeed, Sen's paradox (and most other problems related to the prin- ciples of justice) are simply interpretations of the difficulties related to social coordi- nation and the aggregation of preferences that are well known from game theory and the theory of social choice. Sen's paradox is particularly useful because it highlights the conflict between liberty and welfare that he has shown is crucial for development. Annex 2: Institutions, Social Capital, and Path Dependence Institutions are today mostly defined by "rules of the game" or less precisely as "the way things are done." But this procedural definition is only one part of what is needed for an instrumental definition. Instrumental rationality is generally character- ized this way: If y is to be achieved, x should be done, requires the specification of the goals in order for procedures (for example) to be meaningful. This is perhaps easier to see within the interactive setting of game the- ory, both because the "rules of the game" definition comes from there and because it is inherently social in its reliance on interdependencies. In that context, it is easy to see that the idea of institutions being "rules of the game" does not mean that much, because the rules do not by themselves define the game. At least the specification of a pay-off structure is necessary to establish at least an idea of the game form that is being considered. That is why some argue that a meta-game should be specified and that various institutions should then be seen as strategies chosen by social actors to play their particular games. Of course, the spec- ification of this meta-game is not an altogether straightforward task. Most games of this sort that have been proposed beg the question of where that meta-game comes from; and if the game is to be chosen by its players, then that choice would have to be made within a meta-meta-game, and so on ad infinitum. Other problems with this 138 | VLADIMIR GLIGOROV game-theoretic definition of institutions come from the general problems that game theory faces. Particularly important is that of multiple equilibria, which are often at issue when cultures and institutions are mentioned. It is argued in the literature on social capital that culture could serve as a vehicle to move from a particularistic to generalized trust.17 This begs a number of ques- tions, one being that of how society is constructed. Here, the traditional sociological distinction between community and society is often relied upon--the former based on the culture of collectivism and the other on the culture of individualism. However, the traditional literature highlighted the role of the legal norms in the creation of society, while trust played a more important role in communities. In that context, the notion of social capital may be a way to bridge the gap between communities and the wider society--in a sense constructing social bonds from communal ones. Thus, social trust somehow emerges from people bowling together or from similar activi- ties in local communities. Yet it is not clear how useful the idea of social capital really is if conceptualized in this way. These communal and societal bonds may complement each other, but they may also be substitutes for each other. In other words, it may be the case that the weakness of social bonds leads to the increased importance of communal trust. Thus, it may not be the case that strong communities support strong societies; they may instead be a signal that the societies where they are found are, in fact, weak. In the game-theoretic context, it may not be rational to trust or build up the rep- utation to being trustful, because trust and trustfulness may introduce incentives to cheat. This is why the supply of trust is almost always dependent on interest in the outcome, some type of collateral, and the effectiveness of enforcement. Thus, apart from interest, two key issues are the availability of collateral and the supply of enforcement. Without going into these issues, the point to be made here is that trust is very much endogenous to other economic and social factors.18 Less demanding and easier to conceptualize is the idea of institutional inertia or path-dependence. It relies on the problem that the fixed cost may present for invest- ing in a change or a reform. An established institution has an advantage over an alter- native institution because the costs of the maintenance of the former may be much lower than those needed to introduce the latter. There is an element of a public good in an institution. Thus, there could be an undersupply of new institutions--and the idea of the institutional inertia supports the activist role of the state in reforms and transitions. Annex 3: Some Statistics After the initial transitional recession, the growth in some transition economies was driven by increases in productivity. This is illustrated in figure A3.1 for the (transi- tion) New Member States (NMS) of the European Union, and in figure A3.2 for Bul- JUSTICE AND TRANSITION | 139 FIGURE A3.1. Growth and Productivity in the New EU Member States (percent) NMS-8 (CZ, HU, PL, SK, SL, EE, LV, LT) 140 130 120 110 100 90 Value added 80 Employment 70 60 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: WIIW data (various years). Note: Czech Republic (CZ), Hungary (HU), Poland (PL), Slovak Republic (SK), SL, Slovenia (SL), Estonia (EE), Latvia (LV), and Lithuania (LT). FIGURE A3.2. Growth in the New EU Member States, Bulgaria, and Romania (percent) BG & RO 120 110 100 90 80 Value added 70 Employment 60 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: WIIW data (various years). Note: NMS = new member states; Czech Republic (CZ), Hungary (HU), Poland (PL), Slovak Republic (SK), Slovenia (SL), Estonia (EE), Latvia (LV), Lithuania (LT), Bulgaria (BU), Romania (RO). garia and Romania, where the same process emerges somewhat belatedly. The decrease of employment and increase of unemployment is certainly one indicator of increased inequality. In addition, adverse development in employment have been more pronounced among women, the young, and those who have been out of work for longer periods of time; this is illustrated in tables A3.1-A3.5. Countries have performed differently in different periods, but most of them do not differ very much in their political per- formance or in the level of social conflicts. 140 | VLADIMIR GLIGOROV TABLE A3.1. Employment Rates in New EU Member States/Candidate Countries, LFS (percent of working age population, ages 15­64) 1996 1997 1998 1999 2000 2001 2002 2003 Czech Republic 69.3 68.7 67.3 65.6 65.0 65.0 65.4 64.7 Hungary 52.1 52.2 53.7 55.6 56.3 56.2 56.2 57.0 Poland 58.4 58.9 59.0 57.6 55.0 53.4 51.5 51.2 Slovak Republic 61.9 60.8 60.6 58.1 56.8 56.8 56.8 57.7 Slovenia 61.6 62.6 62.9 62.2 62.8 63.8 63.4 62.5 Bulgaria 54.0 54.1 53.7 51.2 50.4 49.7 50.6 52.5 Romania 65.5 65.4 64.2 63.2 63.0 62.4 57.6 57.6 Estonia 64.9 65.4 64.6 61.5 60.4 61.0 62.0 62.9 Latvia 57.1 59.8 59.9 58.8 57.5 58.6 60.4 61.8 Lithuania 60.3 62.6 62.3 61.7 59.1 57.5 59.9 61.1 European 60.3 60.7 61.4 62.5 63.4 64.1 64.2 64.3 Union (15) European n.a. 60.6 61.2 61.9 62.4 62.8 62.8 62.9 Union (25) Source: WIIW data (various years). Note: LFS, labor force survey; n.a., not applicable. JUSTICE AND TRANSITION | 141 TABLE A3.2. Employment Rates in the NMS/Candidate Countries by Gender, LFS (percent of working age population, ages 15­64) a. Male 1996 1997 1998 1999 2000 2001 2002 2003 Czech Republic 78.1 77.4 76.0 74.0 73.2 73.2 73.9 73.1 Hungary 59.5 59.7 60.5 62.4 63.1 62.9 62.9 63.5 Poland 65.2 66.8 66.5 64.2 61.2 59.2 56.9 56.5 Slovak Republic 69.2 67.7 67.8 64.3 62.2 62.0 62.4 63.3 Slovenia 66.0 67.0 67.2 66.5 67.2 68.6 68.2 67.4 Bulgaria 57.7 58.0 57.5 55.1 54.7 52.7 53.7 56.0 Romania 72.6 71.9 70.4 69.0 68.6 67.8 63.6 63.8 Estonia 69.8 70.7 69.6 65.8 64.3 65.0 66.5 67.2 Latvia 62.3 64.9 65.1 64.1 61.5 67.9 64.3 66.1 Lithuania -- 68.4 66.2 64.3 60.5 58.9 62.7 64.0 European 70.4 70.7 71.2 72.0 72.8 73.1 72.8 72.5 Union (15) European n.a. 70.2 70.6 70.9 71.3 71.3 71.0 70.8 Union (25) b. Female 1996 1997 1998 1999 2000 2001 2002 2003 Czech Republic 60.6 59.9 58.7 57.4 56.9 57.0 56.3 56.3 Hungary 45.2 45.4 47.2 49.0 49.7 49.8 49.8 50.9 Poland 51.8 51.3 51.7 51.2 48.9 47.7 46.2 46.0 Slovakia 54.6 54.0 53.5 52.1 51.5 51.8 51.4 52.2 Slovenia 57.1 58.0 58.6 57.7 58.4 58.8 58.6 57.6 Bulgaria 50.4 50.3 49.9 47.5 46.3 46.8 47.5 49.0 Romania 58.4 59.1 58.2 57.5 57.5 57.1 51.8 51.5 Estonia 60.2 60.4 60.3 57.8 56.9 57.4 57.9 59.0 Latvia 52.4 55.0 55.1 53.9 53.8 55.7 56.8 57.9 Lithuania n.a. 57.2 58.6 59.4 57.7 56.2 57.2 58.4 European 50.2 50.8 51.6 52.9 54.1 55.0 55.6 56.0 Union (15) European n.a. 51.1 51.8 52.9 53.6 54.3 54.7 55.0 Union (25) Source: EUROSTAT; WIIW incorporating national LFS data (various years). Note: LFS, labor force survey; n.a., not applicable; --, not available. 142 | VLADIMIR GLIGOROV TABLE A3.3. Youth Employment Rates in New EU Member States/Candidate Countries (percent of population, ages 15­24) a. Total 1997 1998 1999 2000 2001 2002 Czech Republic -- 41.5 38.3 36.4 34.4 32.3 Hungary 29.6 33.9 34.9 33.5 31.8 30.1 Poland 28.9 28.5 26.6 25.3 24.0 21.7 Slovak Republic n.a. 35.0 31.0 29.0 27.7 27.0 Slovenia 40.4 37.5 34.3 32.8 30.5 30.6 Bulgaria -- -- -- 19.7 19.7 19.4 Romania 36.5 35.5 33.7 33.1 32.6 28.7 Estonia -- 35.5 30.0 28.3 28.1 28.2 Latvia -- 33.3 32.3 29.6 28.8 31.0 Lithuania -- 34.0 32.5 26.5 22.9 23.8 European 37.2 38.2 39.4 40.4 40.8 40.6 Union (15) European 36.4 37.1 37.6 38.0 38.1 37.6 Union (25) b. Male 1997 1998 1999 2000 2001 2002 Czech Republic . 47.3 42.3 39.3 37.2 35.3 Hungary 33.4 37.6 38.5 37.3 35.7 32.9 Poland 33.9 32.7 30.5 28.3 26.6 24.2 Slovak Republic -- 38.0 32.9 29.8 28.9 28.7 Slovenia 44.2 39.5 36.3 35.7 34.1 34.4 Bulgaria -- -- -- 21.8 20.0 20.4 Romania 40.4 39.4 37 32.8 35.2 31.4 Estonia -- 40.0 34.9 32.6 33.9 34.6 Latvia -- 37.7 36.9 34.7 32.8 36.4 Lithuania -- 39.7 36.8 29.6 25.2 27.1 European 40.8 41.8 42.9 43.9 44.2 43.7 Union (15) European 40.3 40.8 41.2 41.5 41.4 40.7 Union (25) c. Female 1997 1998 1999 2000 2001 2002 Czech Republic . 35.8 34.3 33.5 31.4 29.2 Hungary 25.8 30.2 31.2 29.7 27.8 27.3 Poland 24 24.3 22.8 22.2 21.5 19.3 Slovak Republic -- 32.1 29.0 28.2 26.5 25.3 Slovenia 36.4 35.4 32.4 29.7 26.8 26.5 Bulgaria -- -- -- 17.7 19.4 18.4 Romania 32.7 31.6 30.4 30.5 30 26.1 Estonia -- 32.0 26.0 24.0 21.9 21.6 Latvia -- 28.8 27.6 24.4 24.6 25.4 Lithuania -- 28.1 28 23.2 20.6 20.5 European 33.6 34.6 35.9 36.8 37.3 37.4 Union (15) European 32.5 33.4 34.1 34.6 34.7 34.4 Union (25) Source: WIIW data (various years). Note: -- not available. JUSTICE AND TRANSITION | 143 TABLE A3.4. Youth Unemployment Rates in New EU Member States and SEES, LFS (percent of population, ages 15­24) 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Croatia -- -- 26,9 28,4 30,9 36,5 38,4 41,5 35,5 35,9 Macedonia, FYR -- -- 69,5 74,2 70,9 62,9 59,9 56,1 58,4 65,7 Serbia & Montenegro -- -- 50,3 50,2 51,5 53,1 49,6 -- -- -- Czech Republic 8,7 7,8 7,2 8,6 12,4 17,0 17,0 16,6 16,0 17,6 Hungary 19,4 18,6 18,0 15,9 13,5 12,4 10,1 11,2 12,6 13,4 Poland 32,5 31,2 28,5 24,8 23,2 31,3 35,1 41,0 43,9 43 Slovakia 27,3 24,8 20,9 21,7 23,6 32,1 35,2 37,3 36,1 32,2 Slovenia 22,2 18,8 18,8 17,6 18,5 17,9 16,8 18,1 16,7 17,4 Bulgaria 44,9 37,7 33,5 34,8 32,2 34,0 35,3 39,5 36,5 28,3 Romania 22,5 20,6 20,2 18,0 18,4 18,9 18,6 17,5 21,7 18,5 Source: CANSTAT 2003­4, national LFS statistics (various years). Note: EU, European Union; LFS, labor force survey; --, not applicable; SEES, Southeast European states. TABLE A3.5. Long-term Unemployment in New EU Member States and SEES, LFS (percent of total unemployed more than 12 months) 1996 1997 1998 1999 2000 2001 2002 2003 2002 2003 Albania 76,0 84,0 89,0 90,2 89,6 91,8 93,1 92,6 93,1 92,6 Croatia 41,1 44,2 44,7 49,2 52,1 55,9 54,8 53,4 54,8 53,4 Macedonia, FYR 80,7 83,2 82,9 83,8 83,3 86,9 84,5 85,1 84,5 85,1 Serbia & Montenegro 78,8 74,1 79,5 79,6 81,7 74,8 75,6 -- 75,6 -- Montenegro 85,5 82,2 83,9 81,6 80,1 85,0 86,0 -- 86,0 -- Serbia 78,1 73,3 77,9 79,3 81,9 73,4 74,3 -- 74,3 -- Czech Republic 28,2 28,1 29,5 36,0 48,3 57,1 50,5 50,1 50,5 50,1 Hungary 48,3 43,8 41,6 44,6 44,0 41,8 40,1 34,7 40,1 34,7 Poland 39,1 37,9 37,5 32,9 37,9 43,1 49,9 49,5 49,9 49,5 Slovakia 51,6 50,3 50,7 46,9 53,9 57,9 60,0 64,4 60,0 64,4 Slovenia1 52,2 56,7 57,7 57,7 62,4 61,9 57,7 56,5 57,7 56,5 Bulgaria1 -- 60,4 59,1 56,8 58,2 62,2 66,2 65,3 66,2 65,3 Romania1 51,3 47,7 41,9 44,3 51,5 49,5 53,1 60,7 53,1 60,7 Source: EUROSTAT, national LFS statistics for respective countries (various years). Note: EU, European Union; LFS, labor force survey; --, not available; SEES, Southeast European states. 1. Unemployed more than 11 months. 144 | VLADIMIR GLIGOROV Notes 1. Even without the benefit of a comprehensive study of social conflicts in transition coun- tries, one might observe that social conflicts should be distinguished from severe political ones (see Gligorov 1994). It may also be observed that pre-transition socialist or commu- nist economy and society were characterized by deep conflicts and pervasive repression that periodically incited mass expressions of dissent. Today, however, it is often assumed that these were rather harmonious (albeit inefficient) economies and societies. In other words, it still seems to be commonly believed that efficiency was traded for social equality. 2. See annex 1 for a simplified clarification of various concepts used in this paper. 3. For an overview, see Hoff (2000); for a discussion of poverty traps, see Azariadis (2004). 4. For instance, in the Aristotelian sense that rewards are proportionate to efforts ("equals should be treated equally and those that are unequal should be treated unequally"). 5. To call policies equal in this context is almost tautological. 6. For discussion on conflicts over justice, see Hampshire (2000). 7. More on this concept is presented in annex 2 at the end of this paper. 8. Aristotle could be credited with the idea (even though it may not be explicit in his writ- ings) that everybody knows what injustice is, even if nobody knows what justice is. John Rawls argued that there is no moral value to differences in ability that are the product of the natural lottery. 9. The notion of equality of opportunity has to be made precise if it is not to mimic the famous line in George Orwell's Animal Farm that "everybody is equal, only some are more equal than the others." Whether that is at all possible is an open question. 10. For a discussion, see Alesina and Angeletos (2004a and b). 11. For evidence on the rise in inequality, see Milanovic (1998). For evidence on employment and unemployment across countries and social groups, see annex 3 at the end of this paper. 12. For a discussion of these issues, see Gligorov (2000). 13. For an early assessment, see Gligorov (1992). 14. For more discussion on this, see Gligorov (2004). 15. This generalizes to the issue of global justice in the sense that it is hard to see it emerging without the significant strengthening of global institutions. Clearly, meritocratic justice is hard to conceptualize globally, while social justice is hard to enforce. For an interesting discussion of this, see Nussbaum (2003). Annex Notes 1A. The connection between the two is emphasized in Petit (2001). See also List (2004). For a recent detailed criticism of Sen's capability approach, see Pogge (2004); for a different capability approach, see Nussbaum (2003). 1B. For more on this and on social capital in general, see Durlauf and Fafchamps (2004) and Fafchamps (Forthcoming). For the history of the concept, see Farr (2004). 2B. For more on this topic, see especially Rigobon and Rodrik (2004). JUSTICE AND TRANSITION | 145 References Alesina, A., and G-M. Angeletos. 2003. "Fairness and Redistribution." NBER Working Paper 9502, National Bureau of Economic Research, Cambridge, MA. ------. 2005. "Corruption, Inequality, and Fairness." NBER Working Paper 11399, National Bureau of Economic Research, Cambridge, MA. Azariadis, C. Forthcoming. "The Theory of Poverty Traps: What Have We Learned." In S. Bowles, S. N. Durlauf, and K. Hoff, eds. Poverty Traps. Princeton, NJ: Princeton Uni- versity Press. Bicanic, I., V. Gligorov, and I. Krastev. 2005. "State, Public Goods, and Reform." In G. McMahon and J. M. Fabelli, eds. Understanding Market Reforms: Volume 1: Philoso- phy, Politics and Stakeholders. Hampshire, U.K: Palgrave Macmillan. Durlauf, S. N., and M. Fafchamps. 2004. "Social Capital." NBER Working Paper 10485, National Bureau of Economic Research, Cambridge, MA. Fafchamps, M. Forthcoming. "Social Capital and Development." Journal of Development Stud- ies. http://www.economics.ox.ac.uk/members/marcel.fafchamps/homepage/manche.pdf. Farr, J. 2004. "Social Capital: A Conceptual History." Political Theory 32 (1): 6-33. Gligorov, V. 1992. "Justice and Privatisation." Communist Economies and Economic Trans- formation 4 (4): 45-58. ------. 1994. Why Do Countries Break Up? The Case of Yugoslavia. Uppsala, Sweden: Acta Universitatis Upsaliensis. ------, ed.. 2000. "State in the Balkans." Economic Reconstruction of Southeast Europe. Vienna, Austria: Vienna Institute for International Economic Studies (WIIW). ------. 2004. "A Note on Dissatisfaction with Integration." WIIW Monthly Report 7: 8-11. Hampshire, S. 2000. Justice Is Conflict. Princeton, NJ: Princeton University Press. Hoff, K. 2001. "Beyond Rosenstein-Rodan: The Modern Theory of Coordination Problems in Development." In B. Pleskovic, N. Stern, eds. Annual World Bank Conference on Devel- opment Economics 2000. Washington, DC: World Bank. List, Ch. 2004. "The Imposibility of a Peretian Republican? Some Comments on Pettit and Sen." Economics and Philosophy 20: 65-87. Milanovic, B. 1998. Income, Inequality, and Poverty during the Transition from Planned to Market Economy. Washington, DC: World Bank. Nussbaum, M. 2003. "Beyond the Social Contract: Toward Global Justice." The Tanner Lec- tures on Human Values. http://www.tannerlectures.utah.edu/. Pettit, Ph. 2001. "Capability and Freedom: A Defence of Sen." Economics and Philosophy 17 (1): 1-20. Pogge, Th. 2004. "Can the Capability Approach Be Justified?" http://aran.univ-pau .fr/ee/page3.html. Rigobon, R., and D. Rodrik. 2004. "Rule of Law, Democracy, Openness, and Income: Esti- mating the Interrelationships." NBER Working Paper 10750, National Bureau of Eco- nomic Research, Cambridge, MA. Building Efficient Welfare States: Lessons Learned Welfare State, Redistribution, Growth, and Poverty Reduction ROLPH VAN DER HOEVEN Jan Tinbergen, the first Nobel Prize laureate in economics, argued in one of his last writ- ings, that "Redistribution is the core political issue of the 20th century" (UNDP 1994). Indeed, over the past century, industrialized countries have built up a welfare state and now spend a substantial amount of their gross domestic product (GDP) in redistribu- tion to other parts of their populations. The amounts range from roughly 15 percent in the United States to 35 percent in Northern European countries. At the beginning of the 21st century, three questions are often posed with regard to the welfare state: · Can the current system of welfare states and redistribution be sustained and con- tinue, or will it diminish as a consequence of growing globalization? · Can the concept of a national welfare state as developed in industrialized coun- tries be expanded to a system of a global welfare state based on international sol- idarity and global redistribution? · Is the current system of welfare states and redistribution in industrialized countries applicable in developing countries; and, if so, can it contribute to faster poverty reduction in these countries? This paper deals mainly with the third of these; but in order to put that question in perspective, it is necessary to briefly deal with the first two questions. A Diminishing Welfare State? The fear of a diminishing welfare state is based on the argument that a welfare state is costly and inefficient and that it cannot, there- fore, be sustained fiscally--not only because of internal political reasons, but also because of the fact that globalization sets welfare states in competition with each other, leading to a so-called a "race to the bottom" (Deacon 2000). Although most authors indeed acknowledge such political and systemic pressures to reduce the Rolph van der Hoeven is manager of the technical secretariat of the World Commission on the Social Dimension of Globalization at the International Labour Office in Geneva, Switzerland. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 149 150 | ROLPH VAN DER HOEVEN welfare state, many authors take a nuanced view on whether the welfare state is undergoing a rapid decline. Atkinson (1999) has pointed out that those who criticize the welfare state often do so on the basis of a theoretical framework that remains rooted in a model of perfect competition and perfectly clearing markets. However, such a theoretical framework does not reflect reality and cannot theoretically address any of the contingencies for which the welfare state was created and exists. In effect, the whole purpose of the welfare state is missing from the theoretical model of mar- ket clearance; hence policy options have to be based on a different theoretical model that explicitly includes market, including labor market imperfections. Atkinson's arguments measure up well against recent observations by Rodrik (2000), who strongly argues that a well-functioning welfare state can offer advantages in facing the challenges of globalization: As countries open up and become more engaged in international trade and capital markets, they need a strong national welfare state in order to buffer the external shocks that come with greater integration in such mar- kets. And, indeed, it can be observed that countries at the top of the Kearns classifi- cation of the most open and globalized countries also have the highest rates of social spending to GDP.1 Furthermore, social spending has remained a superior social good in most countries, and often debate is more about the content of spending and mechanics of transfers than on the principle of an active welfare state. As argued by Standing (2002), in times of upheaval, systems of regulation, social protection, and redistribution break down. This happened in the 1980s and 1990s, but economic and social realities call for new forms of regulation, protection, and redistribution. This debate on the nature and content of the welfare state in industrialized countries is ongoing. However, in reviewing the question of whether developing countries should also roll back their welfare states, the issue is often that these countries have too lit- tle rather than too much of a welfare state. Hence, we are in no way constrained by concerns about the size of the welfare state in highly developed countries in shaping ideas for a welfare state in developing countries. An Internationalization of the Welfare State? There is also a stronger plea among various policy analysts that the accepted notion of redistribution within national boundaries needs to be taken further on a global scale--and needs to be part of a system of global governance in which the United Nations system and its special- ized agencies, including the World Bank and the International Monetary Fund, must all cooperate in an integrated system of global redistribution within the context of a larger system of global governance. This is, for example, well-argued in the recent report of the World Commission on the Social Dimension of Globalization (WCSDG/ILO 2004). However, despite this report and other pleas for greater inter- national redistribution, research and development of policy alternatives on issues related to global redistribution is still in an embryonic stage: An efficient system for international redistribution of global resources (including development aid) requires a system of international governance, where economic, social, and political concerns are treated in an integrated manner, and where the rights of individuals and their families are acknowledged with regard to development and acted upon by the inter- national community--even in cases where such rights are not fully acknowledged by WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 151 individual governments. Development aid needs to become a global public good for all people, irrespective of the varying behavior of their governments (Kaul, Grunberg, and Stern 1999). A system of global economic, social, and political rights should complement the current malfunctioning global economic and social framework, which is basically founded on two principles, namely the existence (or the potential) of a national welfare state (including "national redistribution") and an international system of free trade and free capital flows (globalization) that determines the national framework for economic and social policies (Kapstein 1999). But, as is becoming increasingly clear, the issue of redistribution systems in a globalizing and liberalizing world is politically not settled at all.2 Hence, all the more reason to focus attention on the development of welfare states and issues of national redistribution in poorer countries as well. Is the Concept of Welfare State and Redistribution Relevant for Developing Countries? The third question posed at the beginning of this paper was whether the welfare state and issues of redistribution in industrialized countries are of rele- vance for developing countries. Strictly speaking, one would agree with Atkinson and Hills (1991), who argue that systems of social security in industrialized countries are of less relevance for developing countries. But one could also argue that the practical application of different elements of a welfare state evolves with the notion of welfare and social security in the course of development process itself. Dreze and Sen (1989), for example, distinguish between "the task of preventing a decline in living stan- dards" and "the enhancement of living standards and the expansion of basic capac- ity of the population." Both issues are relevant to the concept of providing welfare, but their relative weight depends on the level of development and the socioeconomic structure of the society. Such an approach makes the concept of a welfare state more dynamic--and as relevant for the poorer as for the richer countries. The notion of the welfare state has then to be an integral part of the notions of good governance, a notion which is so often restricted only to the notion of "good economic manage- ment of a country." So it makes sense to consider the issue of redistribution in the dynamic concept of a welfare state. But before considering issues of redistribution in more detail, we must still clarify one point: It is often argued, especially in an international context, that issues of national redistribution are difficult to address, as historic and cultural factors determine (to a large extent) the magnitude of inequality in a given country. Therefore, at the ILO we carried out a static and dynamic analysis, in which we asked the questions how much lower, or how much higher respectively, would the World Bank's global estimate of people living in poverty at the beginning of the 21st century be, if at the beginning of the 21st century each country had the lowest com- munity quotient (CQ) or highest level of inequality that each country had itself expe- rienced since the 1960s. In this way, both historical and cultural characteristics of each country would be acknowledged. The results, the details of which are described in Lübker (2002) and reproduced here in figure 1, were quite revealing. Under the equal scenario, the global estimate of poverty would decrease from 1.275 billion to 0.900 billion, while in the most unequal scenario it would increase to 1.370 billion. 152 | ROLPH VAN DER HOEVEN FIGURE 1. Population living on US$1/day or less under different distribution scenarios 1400 1300 1370 1200 1275 1100 people 1000 of 900 900 800 800 700 Thousands 600 500 400 Actual Best Best distribution Worst distribution and positive distribution growth aspect Source: Lübker (2002). (Another experiment was also carried out in which the lowest inequality scenario was supplemented with a parameter reflecting the outcome of recent research which found that lower initial inequality would lead to higher growth, and hence lower poverty. Under that scenario, adding to the lower inequality an estimate of the growth-enhancing effect of lower inequality, the global estimate of poverty would have dropped another 100 million (to 0.800 billion). These simulated numbers tell us two things: First, that inequality does matter; and that even when respecting ini- tial and historical trends, there is an important case to be made for lower inequality. And second, that current inequality is rather high compared to the last decades. As redistribution matters, the following section presents analysis and discussion of the options for redistributive policies in various groups of developing countries, since these could have an important impact on poverty reduction and growth. A short, his- toric overview on the relation between growth, distribution, and poverty is followed by some static modeling exercises, with three experiments of different distribution and growth scenarios and their effects on poverty. This is followed by a discussion of some typical distributive policy options in developing countries. Growth and Distribution3 Of the many issues central to the development process, few have been characterized by the shifts, reversals, and reaffirmations that have plagued the analysis of the inter- action of growth, poverty, and inequality. Evidence that inequality and poverty rose in many countries during the 1980s and 1990s, including some of the OECD coun- tries, rekindled smouldering controversies.4 From the 1950s into the 1970s, analytical emphasis was on probable trade-offs between growth and income distribution. This derived in part from the famous "inverted-U hypothesis" which postulated that inequality would rise in the initial WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 153 phases of development, then decline after some crucial level was reached.5 Growth theories could be cited in support of the hypothesis, such as the Lewis model of "eco- nomic development with unlimited supplies of labour."6 Kaldor's growth model, in which capitalists have higher marginal propensity to save than workers, also implies that redistribution to profits raises the growth rate. This model is most appropriate for developed countries in which the functional distribution of income largely con- sists of wages and profits, and of less relevance to the developing countries consid- ered here.7 The Chenery and Ahluwalia model of "distribution with growth" that came into fashion in the mid-1970s distinguished social groups by asset ownership or mode of access to assets.8 The interaction between growth and distribution was modeled through "income linkages" between the groups; that is, via the labor and commod- ity markets. In simulation experiments with this model, progressive redistribution of income and assets led to substantial improvements in the incomes for poverty house- holds, and non-poverty households as well, via increases in aggregate productivity. As a consequence, in the 1970s emphasis shifted to the identification of redistribu- tive mechanisms to reduce poverty without hampering growth. This focus proved to be short-lived, however, and was abandoned with the rise of neo-liberalism and the Washington Consensus in the late 1980s. In the Washington Consensus approach, growth itself would be the vehicle for poverty reduction, achieved through "trickle-down" mechanisms that were not always clearly speci- fied.9 The perceived ineffectiveness of redistributive measures under the Washington Consensus also led some to advocate targeting public expenditure to the poor--and to judge effectiveness by the accuracy of that targeting. However, targeting of expen- ditures in developing countries is fraught with difficulty. Amartya Sen argued against targeting public spending for several reasons: information asymmetries reduce the effectiveness of targeting in the presence of "cheating," the prospect of losing tar- geted subsidies may reduce beneficiaries' economic activities, targeting may nega- tively effect the self-respect of the poor, and the sustainability of targeted programs is doubtful since the potential beneficiaries are politically weak.10 To the list can be added the formidable measurement problem of identifying who qualifies as benefici- aries, a serious difficulty in industrial countries and virtually intractable in most devel- oping countries. Targeting public spending is more likely to be effective if the poor are a small proportion of the population; that is, if poverty is not a major problem. For countries in which poverty is widespread, the administrative cost, identification, mon- itoring, and delivery of programs may outweigh the benefits. And this is particularly the case if a country is experiencing (or recently has experienced) conflict. In the 1990s, both the neo-liberal analysis and the earlier view of a trade-off between growth and equity were challenged by a number of studies.11 In particular, doubt was cast upon the sanguine view that orthodox macro policies were, by their nature, poverty-reducing. Much of the work on the relationship between growth and income distribution from the 1990s is basically empirical and concluded that during recessions inequality rises and that on average positive growth rates are distribution-neutral, while lower initial inequality raises the likelihood that growth will reduce poverty.12 154 | ROLPH VAN DER HOEVEN A strand of theoretical arguments from the 1990s involves so-called political econ- omy arguments against inequality and, by implication, poverty.13 This analysis pre- dicts a negative relationship between income inequality and growth on the grounds that higher initial inequality would: a) lead to increased public expenditure, because it prompts a demand for redistributive policies, and b) incite political instability that undermines growth. This excursion into political science is somewhat dubious.14 For example, it is not at all clear how a society with the power relationships to generate inequality would, at the same time, produce an underclass with the political power to force redistributive policies upon a government. On somewhat firmer analytical ground, it is also argued that inequality has a negative impact on growth--through imperfect capital markets, to which the poor have limited access.15 In other words, if capital markets discriminate against the poor, potentially profitable activities by the poor are constrained by lack of credit. However, the imperfect capital markets argument has practical limits, in that it presumes the poor to be self-employed or to have the option to become so. While this may apply to a portion of the households in poverty, empirical evidence suggested that during the 1990s, those in the lowest income quintile in Latin America, at least (and perhaps elsewhere), were increasingly in wage employment. The idea that most low-income wage earners could escape poverty through self-employment somewhat challenges the imagination, as well as historical trends. Overall, the literature of the 1990s was relatively limited in its theoretical contri- bution, and most striking in that it demonstrated, yet again, the ambivalence of econ- omists toward the issues of inequality and poverty. On the one hand, the mainstream literature, with its emphasis upon the efficiency of markets, had a predilection to view inequality and poverty as accidental or occasional outcomes of a deregulated growth process. On the other hand, the persistence and severity of poverty in many, if not most, developing countries brought forth periodic arguments for their alleviation. The shifts in emphasis in the literature reflect the difficulty of reconciling these two. However, there seems to be a growing consensus that countries with an "initial condition" of relatively egalitarian distribution of assets and income tend to grow faster than countries with high initial inequality. This is an extremely important con- clusion, because it means that reducing inequality strikes a double blow against poverty: A growth path characterized by greater equality at the margin directly ben- efits the poor in the short run, yet the resulting decrease in inequality creates in each period an "initial condition" for the future that is growth-enhancing. Thus, any growth path that reduces inequality reduces poverty through redistribution and via "trickle down." How Does Redistribution Contribute to Poverty Reduction? This section presents the results of a simulation on the impact on poverty in 50 coun- tries of three simulation exercises, corresponding to different distributional out- comes: 1) a 1 percent distribution-neutral increase in per capita GDP; 2) a 1 percent WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 155 increase in per capita GDP, distributed equally across income percentiles; and 3) a 1 percent redistribution of income from the richest 20 percent to the poorest 20 per- cent. The effectiveness of the outcomes in reducing poverty is judged by the time period required to achieve a given target. In all simulations a "headcount" measure of poverty is used (for details see Dagdeviren, van der Hoeven, and Weeks 2004). The results of the simulation suggest a typology of countries differentiated by the general strategy that is most conducive to poverty reduction. On the basis of these cal- culations, the 50 countries fall into three categories (see table 1). In category 1, the "income redistribution countries," both growth strategies require more than one year to reduce poverty as much as a straight redistribution. For 34 of the 50 countries (68 per- cent), straight redistribution is the most effective method of poverty reduction. In category 2 there are 13 "redistribution with growth" countries, for which redis- tribution is not the most effective poverty reduction strategy, and equal distribution growth is more effective than distribution-neutral growth. For these countries one or both of the growth strategies at least matches the redistribution poverty reduction in less than a year, and the time period for equal distribution growth is the shorter. TABLE 1. Typology of Countries According to Effects of Redistribution Policies of Poverty Category 1: Income Redistribution Countries (34) 1 Venezuela 13 Egypt 24 Costa Rica 2 Colombia 14 Lithuania 25 Moldova 3 Mexico 15 Hungary 26 Kyrgyz Rep 4 Algeria 16 Tunisia 27 Romania 5 Brazil 17 Bulgaria 28 China 6 South Africa 18 Jordan 29 Sri Lanka 7 Morocco 19 Philippines 30 Guinea 8 Dom Rep 20 Czech Rep 31 Pakistan 9 Russian Fed 21 Thailand 32 Indonesia 10 Kazakhstan 22 Mauritania 33 Slovak Rep 11 Panama 23 Chile 34 Senegal 12 Turkmenistan Category 2: Equal Distribution Growth Countries (13) 35 Botswana 40 Kenya 44 India 36 Zimbabwe 41 Nicaragua 45 Rwanda 37 Nigeria 42 Honduras 46 Niger 38 Ecuador 43 Guatemala 47 Belarus 39 Lesotho Category 3: Distribution Neutral Growth Countries (3) 48 Nepal 49 Madagascar 50 Zambia 156 | ROLPH VAN DER HOEVEN These countries are characterized either by low per capita income or relatively equal distribution (or some combination of the two). Finally, there is category 3 comprised of three "trickle down" countries, for which growth as such is the most effective vehicle for poverty reduction. The defining characteristic of the trickle-down coun- tries is that they have more than 50 percent of their population in poverty as a result of their low per capita income. However, it does not follow that all low-income countries would fall into this category. If low income is combined with a relatively equal distribution, as for Niger, equal distribution growth may be more effective in reducing poverty, if only marginally so in that specific case. The simulation exercises demonstrate that for the overwhelming majority of middle- income countries, poverty reduction is most effectively achieved by a redistribution of current income. For these same countries, redistribution with growth would be the second-best option, and distribution-neutral, or status quo growth, a poor third. Low-income countries require a growth strategy, and, for most, redistribution with growth would be more effective than status quo growth. Policy Effectiveness for Redistribution with Growth The major element required to introduce and effectively implement a redistributive strategy in any country is the construction of a broad political coalition for poverty reduction. The task of this coalition would be the formidable one of pressuring gov- ernments for redistribution policies while also neutralizing opposition to those poli- cies from groups whose self-interest rests with the status quo. How such a political coalition might come about is beyond the scope of this paper. The focus here is on a less fundamental, but crucially practical issue: the policies that could bring about a redistribution strategy. To be policy relevant, this consideration of redistribution mechanisms must move beyond a listing of possibilities to an analysis of the likely effectiveness of these, which depends (among others) on the initial conditions and socioeconomic structure in each country and can determine the extent to which redis- tribution policies might contribute to poverty reduction, as shown in the previous section. Perhaps the most important determinant of the effectiveness of the various mea- sures and specifics of each redistribution strategy is the structure of an economy. This structure will depend on the level of development, which will (to a great extent) con- dition the country's production mix, the endowments of socioeconomic groups, the remuneration to factors, direct and indirect taxes on income and assets, prices paid for goods and services, and transfer payments. These elements of the distribution sys- tem are initial conditions that delineate the scope for redistributive policies. In this analytical context, the implementation requirements of redistributive policies can be summarized in a simple theoretical framework. Define the following terms so that: Y denotes the income of a household, V is transfer payments, T is taxes, k is a set of assets (including human capital), w is a set of rates of return (including wages), p is the price set of goods and services, q is the quantity of those goods and services, and S is household saving. WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 157 Then, by definition it follows, Y = (V ­ T) + wk = pq + S Transfer payments Minimum wages, Subsidies for basic Facilitate future (unemployment low-wage needs goods, asset acquisition: compensation, subsidies, other public sector `village banks' and pensions, child labour market infrastructure other financial benefits, aid to regulations, public investment (p); services for the disabled) and employment child nutrition poor progressive taxes schemes (w); credit programmes (q) (on income and programmes wealth) for the poor; land reform, education (k); Effective in Effective in Effective in most Effective in most middle-income middle-income countries countries countries and some low- income countries The effectiveness of tax and expenditure policies (V and T) to generate secondary and tertiary distributions more equitable than the primary distribution depends upon the relative importance of the formal sector. This is for the obvious reason that gov- ernments can most effectively apply progressive income taxes to wage employees and corporations. All empirical evidence shows that the formal sector wage bill and profit shares increase with the level of development. Along with the importance of the for- mal sector goes a high degree of urbanization, and working-poor urban households are more easily targeted than either the rural poor or urban informal sector house- holds. The experience of a number of middle-income countries has demonstrated the effectiveness of basic income payments for poverty reduction, with an example being the basic pension paid to the elderly in South Africa. A tax- and expenditure-based redistribution strategy is most appropriate for middle- income countries, because their per capita incomes are high relative to the absolute poverty line. These are also the countries whose economic structures make taxation and expenditure instruments effective for redistribution. Such countries would include the larger ones in Latin America (Argentina, Brazil, Chile, Mexico, and Venezuela), several Asian countries (the Republic of Korea, Malaysia, and Thailand), and virtually all formerly socialist countries of Central and Eastern Europe. To a certain extent, specific economic structures allow for effective use of taxation for redistribution in a few low-income countries that would typically be relevant only for middle-income countries. If the economy of a low-income country is dominated by petroleum or mineral production, then a large portion of national income may be generated by modern sector corporations. This allows for effective taxation, even though administrative capacity of the public sector may be limited. The tax revenue can be redistributed through poverty-reduction programs, though not through trans- fer payments if the labor force is predominantly rural. Examples of mineral-rich, low-income countries with the potential to have done this, albeit unrealized, are Liberia, Nigeria, and Zambia. 158 | ROLPH VAN DER HOEVEN Interventions to Change the Distribution of Earned Income (wk in the equation presented earlier), which alter market outcomes, will also tend to be more effective in middle-income countries. The most common intervention is a minimum wage, though there are many other policies to improve earnings from work. Further mechanisms include public employment schemes and tax subsidies to enterprises to hire low-wage labor. Some of these would be effective in low-income countries (employment schemes), but others might still be less effective because of enforcement problems (minimum wage), targeting difficulties, and narrowness of impact (wage subsidies). Land reform might achieve poverty reduction for rural households, but the rela- tionship between land redistribution and level of development is complex. On the one hand, low-income countries are predominantly rural, so if land ownership is con- centrated, its redistribution could have a substantial impact on poverty. Further, the more underdeveloped a country, the less commercialized tend to be poor rural house- holds. Therefore, the benefits to the poor from land redistribution in low-income countries are less likely to be contingent on support services. On the other hand, lack of administrative capacity and so-called traditional tenure systems represent substan- tial constraints to land redistribution in many low-income countries, especially in the sub-Saharan countries. The usual approach to land redistribution presupposes pri- vate ownership, such that it is clear from whom the land will be taken and to whom it will be given. There are few sub-Saharan countries in which private ownership is widespread, making redistribution difficult or impossible without prior clarification of ownership claims. While land redistribution is probably not an effective poverty reducing measure for most low-income countries, a few notable exceptions in Asia (India and Vietnam, for example) suggest that it should not be ruled out in all cases. For middle-income countries, experience in Latin America has shown that govern- ments can effectively implement land redistribution. However, the high degree of com- mercialization of agriculture in middle-income countries requires that redistribution be complemented by a range of rural support services, including agricultural extension, marketing facilities, and other measures. Perhaps more serious, the relevance of land reform for poverty reduction tends to decline as countries develop and the rural popu- lation shrinks relatively and absolutely. For example, at the end of the 20th century in the five most populous Latin American countries, 20 percent or less of the labor force was in agriculture. Minimum wages may be more relevant than land redistribution in reducing poverty among the landless and near-landless in such countries. Interventions that directly affect the prices and access to goods and services (pq) could potentially be quite powerful instruments for poverty reduction. Subsidies to selected commodities have the administrative advantage of not requiring targeting, only identification of those items that carry a large weight in the expenditure of the poor. While multilateral adjustment programs typically require an end to such sub- sidies on grounds of allocative efficiency or excessive budgetary cost, the rules of the World Trade Organization do not (as long as subsidies do not discriminate between domestic production and imports). Whether subsidies would generate excessive fis- cal strain would depend on the products covered and financing. Again, the level of development of a country is of central importance for the effectiveness of subsidies. In low-income countries with the majority of the poor in the countryside, consumer WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 159 subsidies are unlikely to have a significant impact on the poor outside urban areas. Basic in-kind goods provision can be an effective instrument for poverty reduction even in very low-income countries, by delivering such items as milk to school chil- dren. To do so with a non-targeted program would require a progressive tax system, and thus would be more likely in a middle-income country. In all countries the poor suffer from poor health and inadequate education rela- tive to the non-poor. Expenditures on education and health have the practical advan- tage that programs that would help the poor are easily identified, though the specifics would vary by country. However, providing these services to the poor may in some countries be as politically difficult to implement as are more obviously controversial measures such as asset redistribution. The same point applies to infrastructure pro- grams directed to poverty reduction. To the extent that these would reduce public investment in projects favored by the non-poor, especially the wealthy, they may be no easier to implement than measures that appear superficially to be more radical. With these generalizations in mind, now consider poverty reduction redistributive policies. Table 2 provides a summary of the discussion, with poverty-reducing mea- sures listed by rows, and the three categories of countries across columns. The table indicates that for the "redistribution" countries, a redistribution of current income and assets is the most effective means of poverty reduction, and the methods to achieve this are feasible. For the "redistribution with growth" countries, the measures for redistribution of current income and assets are less feasible, but instruments to achieve the more modest goal of redistributing the growth increment would be TABLE 2. Summary of Feasibility of Redistribution Instruments by Category of Country Feasibility of Redistribution Instruments Growth with Growth without Redistribution of redistribution redistribution Country category: current income + policies (middle + policies (very Redistributive assets (middle- most low-income low-income instrument: income countries) countries) countries) Progressive taxation Yes Yes, for some No countries Transfer payments Yes Yes, for some No countries Consumer subsidies Yes Yes Yes, for some countries Public employment Yes Yes Sometimes schemes Land reform Yes, but not Yes Not for most always relevant countries Education + health Yes Yes Yes Infrastructure + Yes Yes Yes public works Source: Dagdeviren and others (2004). 160 | ROLPH VAN DER HOEVEN feasible. Finally, most redistribution instruments would not be feasible, or only to a limited degree, for very low-income countries; but for these countries, a growth strat- egy with no redistributive mechanisms may be the most poverty-reducing path. Conclusions Poverty reduction has always been a priority of development policy, although some- times only at the rhetorical level. The end of the 1990s saw increased emphasis on bringing the benefits of growth to the poor. However, growth policies alone are a rather blunt instrument for poverty reduction, since the consensus of empirical work suggests that it is distribution-neutral at best. Along with emphasis on poverty reduc- tion, a shift occurred in the policy literature toward a more favorable view of poli- cies to redistribution income and assets. An integration of distributional concerns and a priority on poverty reduction could be the basis for a new policy agenda to fos- ter both growth and equity. This new agenda would be based on three analytical generalizations: 1) that greater distributional equality provides a favorable "initial condition" for rapid and sustainable growth; 2) that redistribution of current income and assets, or redistri- bution of an economy's growth increment is the most effective forms of poverty reduction for most countries; and 3) the mechanisms to achieve the redistributions are as feasible as other policies for most countries. The latter point perhaps deserves more elaboration. As shown here, implementing an agenda of redistribution is often a major challenge and can pose problems; but these problems should not be exaggerated. In many countries they may prove no more intractable than the problems associated with implementation of other eco- nomic policies. An effective orthodox monetary policy is difficult to implement if a country is too small or underdeveloped to have a bond market. For example, the absence of a bond market leaves the monetary authorities unable to "sterilize" for- eign exchange flows. Similarly, replacing tariffs by a value-added tax would be a daunting task in a country whose commerce was primarily through small traders. Lack of public sector capacity would limit the ability to execute a range of so-called supply side policies such as privatization, "transparency mechanisms," and decen- tralization of central government service delivery. The multilateral agencies have rec- ognized these constraints to adjustment programs--and typically made the decision that constrained implementation was preferable to non-implementation. The same argument can be made for a redistributive growth strategy: To achieve poverty reduction, it might be preferable to implement redistributive growth imperfectly than to implement the status quo imperfectly. Notes 1. See Foreign Policy (January 2002, 2003, and 2004 issues). 2. For further discussion, see van der Hoeven (2001). WELFARE STATE, REDISTRIBUTION, GROWTH, AND POVERTY REDUCTION | 161 3. This section and the following sections are based, in part, on Dagdevieren, van der Hoeven, and Weeks (2004). 4. See Cornia (1999) and van der Hoeven (2002). 5. See Kuznets (1955). 6. See Lewis (1954). 7. For a discussion, see Aghion, Caroli, and Garcia-Penalosa (1999). 8. See Ahluwalia and Chenery (1974). 9. See van der Hoeven and Saget (2004). 10. See Sen (1995). 11. See Milanovic (1998), Ferreira (1999), van der Hoeven (2000), and Weeks (1997). 12. For a discussion, see Ravallion and Chen (1997). 13. See Alesina and Rodrik (1994). 14. See Cramer (2000). 15. See Aghion, Caroli, and Garcia-Penalosa (1999). References Aghion, P., E. Caroli, and C. Garcia-Penalosa. 1999. "Inequality and Economic Growth: The Perspective of the New Growth Theories." Journal of Economic Literature 37 (December): 1615-60. Ahluwalia, M.S., and H. Chenery. 1974. "A Model of Redistribution and Growth." In H. Chenery, M.S. Ahluwalia, C.L.G. Bell, J.H. Duloy and R. Jolly, eds. Redistribution with Growth. Oxford, UK: Oxford University Press, pp. 209-35. Alesina, A., and D. Rodrik. 1994. "Distributive Politics and Economic Growth." Quarterly Journal of Economics 109 (2): 465-90. Atkinson, A.B. 1999. The Economic Consequences of Rolling Back the Welfare State. Cam- bridge, MA: MIT Press. Atkinson, A.B., and J. Hills. 1991. "Social Security in Developed Countries: Are There Lessons for Developing Countries?" In E. Ahmad, J. Dreze, J. Hills, and A. Sen, eds. Social Secu- rity in Developing Countries. Oxford, UK: Clarendon Press. Cornia, G. A. 1999. "Liberalization, Globalization, and Income Distribution." WIDER Work- ing Paper 157, WIDER/UNU, Helsinki. Cramer, Chris. 2000. "Inequality, Development, and Economic Correctness." Department of Economics Working Paper no. 105, School of Oriental and African Studies, London. Dagdeviren, H., R. van der Hoeven, and J. Weeks. 2004. Redistribution Does Matter: Growth and Redistribution for Poverty Reduction. In T. Shorrocks. and R. van der Hoeven, eds. Growth, Inequality, and Poverty--Prospects for Pro-Poor Economic Development. Oxford, UK: Oxford University Press. Deacon, B. 2000. Globalization and Social Policy. UNRISD (United Nations Research Insti- tute for Social Development),Geneva. Dreze, J. and A. K. Sen. 1989. Hunger and Public Action. Oxford, UK: Clarendon Press. Ferreira, Francisco H. G. 1999. "Inequality and Economic Performance." www. worldbank.org/ poverty/inequal/index.htm: World Bank. 162 | ROLPH VAN DER HOEVEN Kapstein, E. 1999. "Distributive Justice as an International Public Good: A Historical Perspec- tive." In I. Kaul, I. Grunberg, and M. Stern, eds. Kaul, I. I. Grunberg, and M. Stern eds. 1999. Global Public Goods--International Coopera- tion in the 21st Century. Oxford, UK: Oxford University Press. Kuznets, S. 1955. "Economic Growth and Income Inequality." American Economic Review 45 (issue?): 1-28. Lewis, W. A. 1954. "Economic Development with Unlimited Supplies of Labour." Manches- ter School of Economics and Social Studies 22: 139-81. Lübker, M. 2002. "Assessing the Impact of Past Distributional Shifts on Global Poverty Lev- els." Employment Paper 2002-37, International Labour Organization, Geneva. Milanovic, B. 1998. "Explaining the Increase in Inequality during the Transition." Policy Research Paper 1935. World Bank, Washington, DC. Ravallion, M., and S. Chen. 1997. "What Can New Survey Data Tell Us About Recent Changes in Distribution and Poverty." The World Bank Review 11(2): 357-82. Rodrik, D. 2000. "How Far will International Economic Integration Go?" In Journal of Eco- nomic Perspectives 14 (1): 177-86. Sen, A. 1995. "The Political Economy of Targeting." In Public Spending and the Poor. D. van de Walle and K. Neat, eds. Baltimore: John Hopkins University Press for the World Bank, pp. 11-24. Standing, G. 2002. Beyond New Paternalism: Basic Security as Equality. London: Verso. UNDP (United Nations Development Programme). 1994. Human Development Report. New York: United Nations. van der Hoeven, R. 1999. "Africa's Adjusted Labour Markets: Can Institutions Perform?" In van der Geest and van der Hoeven, eds. Adjustment, Employment, and Missing Institu- tions in Africa. London: James Currey. ------. 2000. "Labour Markets and Income Inequality: What Are the New Insights after the Washington Consensus?" WIDER Working Paper 209, WIDER/UNU, Helsinki. ------. 2001. "Assessing Aid and Global Governance." Journal of Development Studies 37 (6): 109-17. ------. 2002. "Poverty and Structural Adjustment: Some Remarks on the Trade-Off between Equity and Growth." In New Poverty Strategies: What Have They Achieved, What Have We Learned? P. Mosley and A. Booth, eds. London: Macmillan. van der Hoeven, R., and C. Saget. 2004. "Labour Markets and Income Inequality: What are the New Insights after the Washington Consensus?" In A. Cornia, ed. Inequality, Growth, and Poverty in an Era of Liberalization and Globalization. Oxford, UK: Oxford Univer- sity Press. Weeks, John. 1997. "Analysis of the Demery and Squire `Adjustment and Poverty' Evidence." Journal of International Development 9 (6): 827-36. WCSDG (World Commission on the Social Dimension of Globalization). 2004. A Fair Glob- alization, Creating Opportunities for All. Geneva: International Labour Organization. Global Inequalities: What Can Be Done to Reduce International Inequities? Rights-Based Development JEREMY HOBBS Rights-based development implies a broad understanding of equity that is not limited to income distribution, and which embraces the provision of health, education, and other basic services. Equity also encompasses land rights, cultural rights, minority rights, and democratic rights. And the extreme global inequalities that exist both between countries and within countries contravene universal human rights on all levels. Oxfam contends that there is a collective responsibility to address these inequalities. It is increasingly obvious to most informed people that tackling global inequity is the essential challenge facing the planet today. Inequality and poverty are intrinsi- cally linked to the other big issues, including the environment--and since Septem- ber 11, 2001--global security as well. Yet, extraordinarily, despite increasing understanding of globalization and the increasing interconnections between all people, global poverty and inequality still rate very low in importance on national agendas and in opinion polls. While recent polls indicate that there is broad support for greater economic justice (or fairness), that support is "a mile wide but an inch deep." Growth with Equity Can Reduce Poverty Oxfam agrees with policy experts who believe that economic growth is a vital factor in reducing poverty and injustice. It is also clear from the World Bank's work on sus- tained poverty reduction in East Asia that economic growth must be balanced with measures that ensure equity and sustainability. These include agrarian reform, heavy investment in education, and strategic protection of key industries until they are com- petitive in world markets. Oxfam believes that responsible governments and international institutions should ensure the achievement of most fundamental aspects of human development. Jeremy Hobbs is executive director of Oxfam International in Oxford, U.K. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 165 166 | JEREMY HOBBS Services that are essential to the attainment of the fundamental rights of all mem- bers of society should not be left to the uncertainties and risks of the market. Peo- ple's rights to a livelihood, to decent social services, to a secure existence, to par- ticipation in political life, to respect for their diversity and differences--and to social and cultural innovation--must lie at the heart of public policy and economic planning. Some structural inequalities cannot be changed, such as geography and the avail- ability of natural resources. Most others can be changed, such as the global trading system, developing world debt, and the failure to meet agreed targets of aid. Struc- tural adjustment policies may have been sound theoretical economics, but in many cases they have also reduced public infrastructure, the quality of human resources, and the capacity of governments to manage. Conflict also causes great inequalities, either by plunging people into poverty as happened in central Europe during the 1990s, or by maintaining people in poverty, as in sub-Saharan Africa. The international community's failure to protect civilians in low-intensity conflict or in more extreme situations (such as in Rwanda or Dafur) unless there is a strong vested interest, is in itself iniquitous--and allows and pro- longs global inequity. Trade With trade, the inequities are extraordinarily stark. Rich countries preach liberaliza- tion of developing countries while practicing protectionism on a massive scale. On average, tariffs between rich countries are a quarter of those imposed on poor coun- tries. And within the low averages of tariffs of rich countries, there are huge hidden disparities on products where poor countries have an advantage, particularly in agri- culture and textiles. Ninety-seven percent of the world's farmers live in the develop- ing world, yet the developing world accounts for less than 37 percent of agricultural exports. It is estimated that 3.8 billion people either work in agriculture or depend on the sector. And they mostly live in poverty. The negative impact of the agricultural subsidies of rich countries is well known; together, the United States and European Union provide US$70 billion in agricultural subsidies. These create huge surpluses that are dumped on the world markets and undermine both the viability of farmers and the food security of poor countries. Less well known is the extent to which these subsidies are highly inequitable within their own country; the biggest benefits by far go to large farms and agribusinesses, at great cost to consumers. Currently, England's richest farmer receives over £300k for annual wheat subsi- dies, while the average small cereal grower in the U.K. may receive around £500. In Europe, the biggest recipients of sugar subsidies are the refiners and big agribusi- nesses. In the United States, 25,000 cotton farmers receive US$3.8 billion in subsi- dies. And the impact of such subsidies on world prices threatens the livelihoods of some 10 million farmers in West Africa alone. RIGHTS-BASED DEVELOPMENT | 167 It is estimated that the total value of protection and subsidies is US$1 billion per day, or seven times that of global official development assistance (ODA). Yet, if Africa lifted its share of world trade by only one percent, it is estimated that some 70 million people could be lifted out of poverty. Oxfam has documented cases where forced liberalization--whether through the World Bank/International Monetary Fund (IMF) conditionalities, the World Trade Organization (WTO), or free trade agreements--has caused immense hardship (in Haiti, Mexico, and Peru for example). In general, Oxfam's research shows that rapid liberalization of the imports of poor countries has had a negative impact on poverty reduction, partly because it has exacerbated inequity by creating rural poverty rather than growth of urban economies. Slow liberalizers, such as China and other East Asian countries, have done far better. Commodity prices in agriculture also create huge inequities. The failure of the cof- fee market is a good example of how the big roasters have forced coffee prices down, (below the cost of production), forcing millions of people off their farms--even though the roasters have made record profits. While trade cannot be a panacea for inequity, changing the rules of trade in favor of poor countries would make a big difference. The Doha Development Round was supposed to put development at the heart of the WTO negotiations and to redress some of the imbalance from the Uruguay round. The Doha round has struggled so far, with U.S. intransigence on the TRIPS agreement (Trade-Related Aspects of Intel- lectual Property Rights) delaying the negotiations for nearly a year, followed by the collapse of the talks at Cancun--because of a clash between the developed world and the developing countries led by the G-20 nations. So what can be done to reduce international inequities in trade? The meetings con- cluded in Geneva in July 2004 have made enough progress to keep the round alive. Agreement to end agricultural export subsidies and export credits is a big step for- ward, although the text still lacks a firm date and there is continued likelihood that the Europeans and the United States will keep options open on how domestic subsi- dies are treated. Without discipline on the WTO's box system, trade-distorting sub- sidies could be reclassified in such a way as to maintain current levels. Brazil has won two important WTO panel decisions (one on cotton and one on sugar) that have added momentum for the end to export subsidies. On agricultural market access, developing countries are being asked to accept a blended formula that will cut their tariffs much harder and faster than for the rich countries. On special and differential treatment for developing countries, little progress was made, while Europe and others secured agreements around so-called sensitive products. For the Doha Development Round to make a difference in global equity, deeper Common Agricultural Policy reform is needed (particularly in France and Germany), as well as reform of the U.S. farm bill, increased market access for developing coun- tries, and realistic accommodation of the needs of developing countries and the less developed countries (LDCs) to manage the pace and scale of liberalization-- especially with regard to strategic products and food security. 168 | JEREMY HOBBS The Millennium Development Goals It is obvious that over the past few years, the development community has not won the political argument for achieving the United Nation's Millennium Development Goals (MDGs), but there is a groundswell of support for change. Politicians often say there are no votes on aid (although they will also say that few votes are lost in sup- porting aid). And it is a rare development minister who can persuade his or her finance minister to increase aid. For most donor countries (apart from the Nordic states), to reach the United Nation's ODA target of 0.7 percent of gross domestic product (GDP) means at least a doubling of their ODA. Stated differently, the devel- oping world has missed out on billions of promised aid--a massive under-invest- ment. And much of the aid that has been given has been of poor quality and often in the interests of companies in the donor countries. Resourcing is currently well off-target for the MDGs, and many regard 2005 as the last chance to get back on track if the 2015 goals are to be met. The leadership of the U.K. in 2005 is shaping up as the best opportunity in years for the future of the MDGs, with the chancellor and prime minister showing deep commitment to aid and to Africa. With the G-7 nations meeting in Britain in 2005, and with Britain holding the EU presidency, there are opportunities to build signifi- cant momentum around the MDGs. Many civil society organizations are working together to lift the debate and secure some results backed by contributions. This is a huge opportunity to advance the cause of equity, which development institutions such as the World Bank need to embrace. What Oxfam Is Requesting in 2005 On aid specifically, Oxfam would like for: · All DAC countries (member countries of OECD's Develoment Assistance Com- mittee) and the EU to set time-tabled targets to reach 0.7 percent of GDP in for- eign aid by 2013 · Rich countries to announce an increase of at least US$50 billion in foreign aid at the U.N. Millennium Summit in September 2005 · Rich countries to agree on increased resources given to the International Develop- ment Association (IDA) by 10 percent, as part of the IDA's 14 negotiations in early 2005 · Rich countries to give support to the International Finance Facility (IFF) in early 2005, as a clear way of raising vital finance now, in addition to setting targets for 0.7 percent of GDP · Rich countries to fully fund the Education for All Fast Track Initiative and the Global Health Fund RIGHTS-BASED DEVELOPMENT | 169 · No repayments of IFF to be taken from aid budgets, and for aid to be transpar- ently and accountably governed · Fully implemented commitments made in the Rome Declaration in February 2003 to harmonize the operational policies, procedures, and practices of donor institutions · All donors to commit to giving at least 70 percent of aid to low-income countries by 2010, with the other 30 percent directed to the poorest in other countries · Donors to immediately implement the Organisation for Economic Co-operation and Development (OECD) agreement to untie aid to LDCs; in addition, this would then be supplemented by the further untying of food aid and technical assistance. Debt Nongovernmental organizations have held a consistent position over recent years-- that any calculation of debt sustainability should be linked to the poverty needs a country has, especially the financing gaps it faces in reaching the MDGs. This defi- nition of debt sustainability has gained wide acceptance in the United Nations, among African governments, finance ministers of the highly indebted poor countries (HIPCs), and even creditor governments such as that of Ireland. The Monterrey Consensus states that "future reviews of debt sustainability should also bear in mind the impact of debt relief on progress towards achievement of the MDGs." And it is clear, based on this criterion, that the debts of the HIPC countries are not sustainable. Uganda currently has a debt-to-exports ratio of 300 percent, and Ethiopia will not reach the HIPC target of 150 percent--widely disputed as not enough--until 2020, even after top-up debt relief (if received at completion). On average, HIPC countries in Africa are still spending 15 percent of their annual revenue on servicing debt, with some (Gambia, for example) spending over 25 per- cent. In 2004 legislation, the U.S. Congress agreed that in countries with critical poverty needs this figure should not be more than 5 percent. Yet debt relief under HIPC has worked very well in tackling poverty. Debt relief is a source of high quality resources to be spent on poverty reduction, as the money is gov- ernment revenue and is guaranteed over the long term. And as such, it is the equivalent of 20-year budget support--and far more reliable than foreign aid. This means, for example, that it can be used for crucial recurrent costs such as salaries for key workers. Poverty reducing expenditures in African HIPC countries have increased 6 per- cent, on average, and as much as 14 percent in some countries. In Benin, 43 percent of HIPC debt relief went to education in 2002, allowing the recruitment of teachers for empty posts in rural areas. Fifty-four percent went to health; and of that amount, 20 percent was used to recruit health staff for rural clinics, with the remainder allo- cated to implementing HIV/AIDS and anti-malarial programs, improving access to safe water and increasing immunization. In Mali, a monthly stipend of 25,000 CFAs is provided to over 5,000 community teachers using HIPC relief. 170 | JEREMY HOBBS In Niger, a special program that focuses on rural education, health, food security, and water systems has been fully financed through HIPC. So far, this has mainly been used to build classrooms and rural clinics. In Malawi, HIPC resources have been used to, among other things, train 3,600 new teachers each year. In Burkina Faso, HIPC relief has been spent on education (39 percent), health (33 percent), and rural roads (28 percent).1 But the amount of debt relief given so far is not nearly enough. Ten out of the 14 African HIPC countries where data are available still spend more paying back for- eign debt than they do on health services. This is true for Ghana, for example, where 11 percent of government spending goes to pay debt, compared to 9 percent for health care.2 In Zambia, the government spends more on servicing its debt than it does on education--and has faced widespread strikes since the IMF banned pay increases to civil servants. A teacher's basic take-home pay is $60 dollars per month. The IMF has also insisted that the Zambians cannot afford to employ 8,000 newly trained teachers. The cost of these "unaffordable" wage increases would be $106 million per year: In 2004, Zambia paid $227 million in debt repayments. Thus, the U.S. government's proposal to cancel 100 percent of multilateral debt is very encouraging--provided it is agreed that the cancellation is funded with new money and not offset against inflows (as proposed). Specifically, Oxfam wants to see: · Multilateral debt relief of 100 percent for low-income countries, announced at the World Bank's 2004 annual meetings, as a first step toward the Millennium Plan · Revaluation of IMF gold with resources used to finance debt cancellation · Any resources necessary for debt cancellation to be in addition to overseas aid funds · Future debt sustainability to be linked to financing requirements for the MDGs, and grants given instead of loans where necessary · Transparent allocation of aid and debt relief according to poverty needs, with the agreed aim to use aid and debt relief to support functioning and effective govern- ments in delivering policies grounded in national choices--along with the involve- ment of parliaments, reflecting the priorities of poor people · An end to policy conditionality that dictates detailed structural adjustment reforms, overly conservative macroeconomic policies, or specific reform paths. Conditions on resources must be restricted to those based on financial accounta- bility and broadly agreed poverty reduction goals such as the MDGs or human rights standards. · Poor country governments must demonstrate a domestic commitment to poverty reduction--through participatory planning and increased, transparent spending on poverty reducing expenditures. RIGHTS-BASED DEVELOPMENT | 171 · Rich countries and the World Bank and IMF to actively promote and support a complete end to user fees in health and education, in all poor countries · An end to the IMF role as the gatekeeper for debt relief and aid to poor countries. Conclusions If the international community can grasp the opportunities presented by both the Doha Development Round and the United Nations' MDGs, there is a chance that significant inroads can be made against poverty and inequity. The challenge for the World Bank, political leaders, and civil society is to build the momentum and public awareness necessary to rise to these challenges. The 2006 World Development Report should, therefore, provide the arguments and research to demonstrate how addressing equity--nationally and globally--can reduce poverty, and why it is in everyone's interest to do so. Notes 1. These figures are taken from Hinchcliffe (2003). 2. This is also true for Cameroon (percent of GDP spent on debt: 2.3, and health: 1.5 in 2002­3), Ethiopia (percent of GDP spent on debt: 1.3, and health: 1.1 in 2003), Gambia (percent of GDP spent on debt: 4.9, and health: 3.9 in 2003), Guinea (percent of GDP spent on debt: 2.7, and health: 0.8 in 2002), Madagascar (percent of GDP spent on debt: 1.1, and health: 0.9 in 2002), Malawi (percent of GDP spent on debt: 2.5, and health: 0.9 in 2002), Mauritania (percent of GDP spent on debt: 4.9, and health: 3.4 in 2003), Senegal (percent of GDP spent on debt: 3, and health: 2.1 in 2001), Uganda (percent of GDP spent on debt: 1.6, and health: 1.2 in 2003), Zambia (percent of GDP spent on debt: 3.9, health: 2, and education: 2.8 in 2001). These figures are from Hinchcliffe (2003) and NEF (2003). References Hinchcliffe, K. 2003. "Notes on the Impact of the HIPC Initiative on Education and Health Public Expenditures in African Countries." Unpublished paper prepared for the Human Development Department, Africa Region, Oxfam International, Oxford, UK. NEF (New Economics Foundation). 2003. Real Progress Report on HIPC. London. Rising Labor Force Participation as a Source of Pro-Poor Growth ANDREW M. WARNER This paper is about the channels through which economic growth reduces poverty-- in the cases where growth does indeed reduce poverty. Many analysts point out that growth does not invariably reduce poverty; and there are, in fact, cases where eco- nomic growth is not accompanied by an increase in the incomes of poorer groups. But these episodes are relatively rare. In the 29 countries in the well-known data organized by Denninger and Squire (1996) that experienced significant economic growth (defined as growth of 2 percent or more that lasts at least five years), there is one case where real income growth of the poorest 20 percent of the population actu- ally declined. This case was Colombia, between 1970 and 1991. In Colombia during this period, real economic growth per person was 2.16 percent; yet growth of the real income of the bottom 20 percent of the population was ­0.99 percent. In the same dataset there were 28 out of 29 cases where growth for both the econ- omy and the real income of the bottom 20 percent was positive; and in 15 of these, growth of the bottom 20 percent was higher than growth of the economy. Therefore, while growth does not invariably reduce poverty, there are many cases where growth significantly improves the real income of the poor and some in which growth improves the relative position of the poor. In light of this kind of evidence, and several previous studies that have docu- mented the neutrality proposition, a natural follow-on question is not whether growth helps the poor--but how growth helps the poor.1 This paper identifies four possible channels: Two of these have to do with trends in the overall level of wages and profits and differences in wage trends for different income groups; and two of these concern increases in the employment of poor people and changes in employ- ment by economic sector. Which of these channels tends to be more quantitatively significant in terms of their impact on poverty? This paper suggests that the employ- ment shifts tend to be more significant than wage or profit shifts. In particular, the Andrew M. Warner is the senior economist at the Millennium Challenge Corporation. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 173 174 | ANDREW M. WARNER rise in labor force participation of poorer workers appears to be especially important. These conclusions are drawn from a study of two fast-growing countries, Mauritius and Thailand. It is worth noting that data across many countries show that labor force partici- pation rates often rise during growth, sometimes dramatically. In Thailand, labor force participation rose from 43 to 56 percent of the population between 1977 and 1990. In Singapore, the increase was 33 to 50 percent, and in Taiwan 34 to 44 per- cent. Since the poverty rate in Thailand in 1975 was estimated at 31 percent of the population, the additional 12 percent of the population that entered the recorded labor force could represent a significant fraction of poor people. While it is not known where these new workers came from, the presumption in less-developed countries should be that many were poor or near-poor workers on the margin of for- mal or full-time employment. In addition to raising formal participation rates, growth may also increase the intensity of work through rising hours and through a shift from part-time to full-time work. All such employment shifts are analytically separable from trends in average wages or profits as channels through which growth can impact poverty and the distribution of income. Of all the possible shifts in employment in the course of growth, the shift into for- mal employment is likely to be particularly relevant for the poor. It is true that the very poor are almost always employed in some fashion, because not working endan- gers survival. Indeed, formal unemployment as we know it in richer countries is a luxury available only for middle-income or relatively rich workers in poorer coun- tries. However, although the poor are always employed, they typically work in mar- ginal activities, and are poorly paid and poorly recorded. In this context, a rise in measured labor force participation is likely to be pro-poor to the extent that it is associated with a transition of marginally employed workers into more intensive or higher-paying employment. In contrast, the impact on the poor of employment shifts between sectors is more ambiguous. The best known and studied of these employment shifts--agriculture to industry--is likely to be pro-poor for those migrants that find regular urban jobs. For the migrant population as a whole, there is likely to be a lot of variation; but the shift is ambiguous for those who obtain only marginal urban employment. In the classic Lewis model, the incomes of these workers rise as they obtain work in the formal sec- tor; alternatively, the incomes of such workers may not rise if they merely wait in the urban informal or semi-industrial sector. This paper is organized as follows. The next section sets the stage analytically by illustrating how value added (of any economic entity) can be separated into contri- butions from participation rates, wages, profits, and shifts across sectors. Cross- country evidence is then presented to show that changes in participation rates have indeed been an important phenomenon in growing countries. The final two sections document the employment shifts that occurred in both Mauritius and Thailand--and investigate the distributional implications of these employment shifts, by examining the available data on poverty incidence and wages in both countries. RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 175 Analytics The channels examined in this paper may be clarified by examining the relationship from national income accounting which states that GDP must equal the sum of wage and non-wage incomes throughout the economy. This identity is written in equation 1. GDP = (E1w1 + E2 w2 + E3 w3 + ) N N Equation (1) In equation 1, E stands for employment, N is population, and w and are wages and profits. Three sectors are shown. The equation may be manipulated to show the relationship in terms of the employment to population ratio, E , shares of N employment, Si, average wages in the three sectors, and the ratio of profits to employment: GDP = E s1w1 + s2 w2 + (1 - s1 - s2 )w3 + Equation (2) N N E It is important to clarify that this equation is not a statement that causality runs in any particular direction. In fact, this equation would normally be one equation in a larger system in which the elements of equation 2 were simultaneously determined by exogenous factors. To present an example, a productivity boom in sector 2 would raise GDP on the left-hand side of equation 2, but might have as a counterpart on the right- hand side a rise in wages in sector 2 and some increase in profits. Over time there might be migration of labor from sector 1 and 3 to sector 2, reflected in changing shares, and average wages may adjust further as the general equilibrium of the economy sorts itself out. The point is simply that any event that boosts GDP on the left side of the equa- tion must be mirrored in terms of some combination of events on the right side of the equation. And furthermore, knowing which event on the right is associated with GDP growth helps us learn about the pro-poor nature of the growth, because the events on the right of the equation can have different implications for the poor. In summary, equation 2 suggests that national GDP growth may be associated with any one of four distinct events: a change in the employment population ratio; a change in the average wage within sectors; a change in employment shares between sectors; or a change in profits per employed person. Further, each of these changes could occur in a direction that is pro-poor. Specifically, growth will be pro-poor: A. If the elasticity of wages with respect to GDP is higher for lower-wage sectors; B. If growth is accompanied by a net transfer of employment from low- to high- wage sectors; C. If growth is accompanied by a rise in the ratio of employment to population of low-income workers; D. If the elasticity of profits with respect to growth is negative (and, of course, prof- its are earned by people of higher than average income). 176 | ANDREW M. WARNER Later sections of this paper will argue and present evidence that in Thailand, chan- nels C and (to a lesser extent) B were significant. It will also show that channels C and B were important in Mauritius. Data to evaluate channels A and B are more dif- ficult to obtain, but household survey data from Mauritius are available to shed some light on this issue. Evidence from this data will also show that channel A was signif- icant in Mauritius during the latter stage of its growth episode, between 1986 and 2002. In addition to this country-level evidence, the following section will present cross-country data showing that employment shifts (channel C) have been significant in a wide range of countries. Cross-Country Evidence This section provides two pieces of evidence that employment shifts are a common occurrence in growing economies. The term "employment shifts" stands for increases in labor force participation, reductions in part-time or semi-informal work, and labor movements between sectors. The first piece of evidence shows that increases in labor force participation correlate positively with increases in per capita GDP across countries. Specifically, figure 1 shows that virtually all countries with growth of 2 percent or more also experienced rising rates of labor force participa- tion. The data span the period 1960­95 and include all countries with more than 20 years of data. While there were a few countries such as Guyana and Nicaragua that declined economically and yet increased participation, it is difficult to find countries that grew rapidly without increases in labor force participation. (The closest in this data would be Ireland.) In other words, growth rarely happens without increases in FIGURE 1. Growth of GDP vs. Growth of Labor Force Participation 0.06 Taiwan Singapore Cyprus 0.04 Ireland Thailand Malaysia GDP Mauritis Iceland 0.02 Dominican Republic owth Gr 0.00 Haiti Guyana Nicaragua ­0.02 ­0.01 0.00 0.01 0.01 0.01 Growth LFPR Source: Author's calculations, using data from Bosworth and Collins (2003). Note: GDP, gross domestic product; LFPR, labor force participation rates. RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 177 participation, but increases in participation do sometimes occur without rapid growth. Figure 1 also shows the variation across countries. Malaysia and Thailand experienced roughly the same economic growth rate, but Thailand experienced much faster growth in labor force participation than Malaysia. Another way to examine the relation between labor force participation and GDP growth is to examine data that measure two kinds of labor-force attachment that tend to be semiformal or informal: self-employment and unpaid family labor. Both of these decline dramatically as countries grow, so that higher incomes are associated with reduced shares of the labor force engaged in self-employment and unpaid fam- ily labor. For example, figure 2 plots the International Labour Organization's meas- ure of self-employment and unpaid family labor (as a percent of the labor force) and shows that the extent of such work declines as country income rises. In sum, both kinds of evidence (in figures 1 and 2) support the general proposi- tion that increases in work intensity tend to accompany economic growth. As dis- cussed in greater detail in the following section, this increase is likely to be positive for the poor. The discussion now considers this more specifically with reference to Mauritius and Thailand. Employment Shifts in Mauritius and Thailand This section summarizes data on the employment shifts that accompanied fast growth in both Mauritius and Thailand. In both cases, it will be shown that increased participation in the labor force accompanied economic growth. In Thailand, the rise FIGURE 2. Self-Employment and Unpaid Family Workers (percent of LF vs. GDP per capita) Central African Republic 80 Malawi Nigeria 1990 Haiti Indonesia Thailand c. 60 Pakistan Turkey cent, per in Mexico Greece ec 40 Korea for labor Italy of 20 Japan e Shar Hungary USA 0 6 7 8 9 10 GDP pc, in logs Source: Author's calculations, based on ILO Yearbook of Labor Statistics. Note: GDP, gross domestic product; LF, labor force; PC, per capita. 178 | ANDREW M. WARNER in participation was followed by second shift: a migration of labor from agriculture into all economic sectors. In Mauritius, the rise in participation happened along with a shift from "unemployment" into employment. In other words, there were two simultaneous movements: from nonparticipation to formal labor force status, and within the labor force, from unemployment to employment. Inevitably, data avail- ability constrains the analysis. We do not track all of the variables in the decompo- sition (wages, profits, employed status) for all of the years in which fast growth occurred. Nevertheless, there are sufficient data to draw important conclusions. The employment shifts are analyzed in terms of their quantitative impact on growth rates as well as their impact on the pro-poor nature of the growth itself. Thailand 1977­97 Thailand's remarkable rise in labor force participation occurred between the late 1970s and 1990. During this period, the labor force par- ticipation rate increased from 43 percent to 55 percent of the population before falling slightly (to 50 percent) by 1999­2000. This represents an increase of 8 to 12 percentage points in a period of just over a decade: an unusual achievement (see figure 3). To gauge the potential impact on poverty, it is helpful to estimate the number of people this shift may have involved. One way to estimate this is to compare the observed increase in the labor force with the increase in the labor force that would have been expected due to population growth alone. The difference between the two is an estimate of the number of "additional" workers due to rising participation. The estimates in table 1 suggest that by 1990, Thailand had an additional 6.8 million workers in the labor force due to this shift. FIGURE 3. Thailand: Ratio of Employment to Population, 1977­90 0.55 employed 0.5 population the 0.45 of Fraction 0.4 1977 1980 1990 2000 Year Source: Author's calculations using data at http://web.nso.go.th/eng/stat/lfs_e/lfse.htm. RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 179 TABLE 1. The Rise in Labor Force Participation in Thailand Between 1979 and 1990 Labor force with Labor force Population Participation constant 1979 "Additional" Year (`000) (`000) rate participation rate workers 1979 19954 46005 43.37% 19954 0 1990 30843 55287 55.79% 23980 6863 Change 1979­90 +12.42 Source: Author's calculations. TABLE 2. Growth Accounting During the Period of Rapid Employment Growth in Thailand: 1979­90 Percentage Percent growth, Contribution contribution to Weights 1979­90 to growth total growth (1) (2) (3) = (1)*(2) Capital growth 0.35 12.90 4.52 66% Labor force growth 0.65 3.53 2.29 34% from population growth 0.65 1.83 1.19 17% from additional labor 0.65 1.70 1.11 16% force growth Total explained growth 6.81 100% Source: Author's calculations. Note: Additional workers defined in the text. Apart from the impact on poverty, this increase in participation was sufficiently large that it was itself a factor in explaining growth for the whole Thai economy. To provide some illustrative calculations, note that economic growth was 7.29 percent over the 1979­90 period. Of this, as seen in table 2, growth in inputs can account for 6.81 percentage points in growth, since (with weights of 65 percent and 35 per- cent respectively) labor force growth and capital growth account for 2.29 percent and 4.25 percent growth. If the growth in the labor force of 34 percent is further divided, 17 percent came from population growth and 16 percent from the rise in labor force participation. These data therefore suggest that the additional labor employed can account for about 16 percent of growth during this period: In sum, employment contributed approximately 1.11 percentage points to the average 7.29 percent growth over the full period. A contribution of 1.11 percentage points out of a total of 7.29 percent growth is moderate in terms of its impact on growth. Nevertheless, it is likely to be significant in terms of its impact on the economic welfare of those affected. This is because those out of the labor force at the start of the period were very likely marginal laborers earning very little in informal activities--or nothing at all if they simply were not working. 180 | ANDREW M. WARNER It is possible that improvements in measuring part-time or informal labor, rather than true increases in labor force participation, may account for the observed increase in labor force participation. However, the available evidence does not sup- port this idea. The Thai labor force survey does indeed report figures for "under- employment" and "seasonally inactive labor" (not shown). The underemployed are included in the definition of the labor force and are first measured in the 1983 survey. Data for seasonally inactive labor begin in 1992. Yet these items are not large fractions of the labor force when they are measured, suggesting that even if measured prior to these dates they would not have been quantitatively significant. Moreover, neither data series trends up or down over the period in which they are measured, which undercuts the idea that the quality of the measurement has been changing over time. The second major employment change was the movement from agriculture to other economic sectors. This transition out of agriculture is shown in figure 4. It can be seen from the figure that the transition really only started in 1990, and was espe- cially dramatic in the period 1990­97--when the percentage of employment in agri- culture declined from 64 percent to 45 percent. The magnitude of the effect of this transition on GDP growth depends on the incomes of the marginal workers who actually moved in both the sending and receiving sectors. Other things constant, a transition to higher-wage employment would be recorded in a growth accounting framework as an increase in total factor productivity, since GDP would increase with no observed increase in aggregate employment or capital. FIGURE 4. The Transition Out of Agriculture in Thailand, Measured by the Fraction of the Labor Force Employed in Agriculture 70 60 cent 50 Per 40 30 1952 1960 1970 1980 1990 2000 Year Source: http://web.nso.go.th/eng/stat/lfs_e/lfse.htm Note: Authors calculations using data from the Thai Labour Force Survey. RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 181 To illustrate the magnitudes involved in such a transition, and to estimate the likely effect on the growth rate, consider the calculations summarized in table 3. There were 17.7 million workers in agriculture in 1990. This declined to 14 million by 1997, for a total decline of 3.6 million. Suppose that the typical agricultural worker that later moved on was earning one-third of the average per capita GDP in 1990, and that this typical worker moved to another job in which he or she earned two-thirds of the average per capita GDP. This represents a doubling of the income of the individual worker, from $1,757 in 1996 U.S. dollars per year to $3,773 in 1997, or an increase of 100 percent. The net gain to GDP can be calculated by mul- tiplying the additional annual income per person of $2,255 by the 3.6 million per- sons affected, for a total increase in GDP of approximately US$6.4 billion. This rep- resents an additional 2.4 percent of total Thai GDP in 1990. Such an increase spread out over seven years would translate into an additional 0.34 percent in overall growth and total factor productivity growth. In the example just given, the transition out of agriculture accounts for a small fraction of the average annual growth rate. However, the size of this effect, although moderate from the perspective of the economy, could have been highly significant for those affected. In the example just presented, it represents a 100 percent increase in income for 3.6 million persons. Both major transitions--into employment and between agriculture and nonagriculture--are likely to be important for poverty and income inequality, and for understanding the extent to which growth in Thailand affected poverty. The rise in participation was very likely positive for the poor, since the previously non-employed were likely to have been disproportionately poor. It is not clear, how- ever, the extent to which this would have been reflected in poverty statistics or income distribution statistics. To the extent that the universe for income distribution statistics is all wage-earners, (measured) income inequality may even rise, as low- wage earners are added to the bottom of the distribution. There would simply be TABLE 3. Some Estimates of the Impact of the Agricultural Transition in Thailand during the 1990's 1990 1997 Number of workers in agriculture (`000) 17706 14081 Change 1997­1990 ­3625 Average income in 1990 (96USD per person) 5270 33% of average income 1757 67% of average income 3513 Additional income of those leaving agriculture 1757 Number of people affected 3625 Total rise in GDP (96US$) 6,367,917 Total Thai GDP (96US$) 267,202,071 Total rise in percent of 1990 GDP 2.4% As percent of individual income 100.0% Source: Author's calculations. 182 | ANDREW M. WARNER TABLE 4. Poverty Rates in Thailand (head count ratios) Levels Changes Year Thailand Rural areas only Thailand Rural areas only 1962 57.0% 61.0% 1975 31.0% 35.0% ­26.0% ­26.0% 1981 23.0% 27.3% ­8.0% ­7.7% 1990 27.0% 33.8% 4.0% 6.5% 1996 11.2% 14.9% ­15.8% ­18.9% 2000 14.2% 19.0% 3.0% 4.1% Source: National Economic and Social Development Board (2000) and Jitsuchon (2001), taken from table 6 (p. 19) Jitsuchon (2002). more people at the bottom of the income distribution who previously were not there because they had zero income. In contrast, the transition out of agriculture is likely to reduce income inequality (both in reality and in the statistics), since this represents the movement to higher incomes of people already present in the official statistics. These people are likely to be moving from a position at the bottom of the income distribution to somewhere closer to the mean, so income inequality should decline. The reported poverty statistics in Thailand are consistent with the idea that the agricultural transition during the 1990s had a greater impact on (recorded) poverty than the rise in participation in earlier years. As is shown in table 4, poverty did not change dramatically between 1975 and 1990, a period that corresponds roughly to the rise in the employment to population ratio. However, the poverty rate did decline dramatically after 1990, a period that corresponds to migration out of agriculture. Mauritius 1982­2002 In Mauritius, a significant source of growth during the 1980s was the rise in participation. This source of growth ran its course in the 1980s, however. To explain what happened in Mauritius, note that any growth in GDP per capita can always be expressed as the sum of the growth rate of GDP per employed person and the growth rate of the employment population ratio, based on the follow- ing identity: GDP GDP EMPL POP EMPL POP Table 5 shows these data for Mauritius. Growth rates are shown in the right-hand columns and levels in the left-hand columns. The remarkable fact here is that, during the 1980s, a significant part of GDP per capita growth was simply due to growth in the employment-population ratio rather than growth in output per worker. The summary at the bottom of table 1 shows that between 1982 and 1991, output per worker grew at only 1.29 percent. However, GDP per worker grew at 5.26 percent because employment was growing at 3.98 per- RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 183 TABLE 5. Decomposition of GDP into Mauritius into Labor Productivity and Employment, Mauritius Levels Growth Rates GDP per Employment- GDP per Employment- GDP per employed population GDP per employed population person person ratio person person ratio 1982 23,518 82,624 28% 1983 23,381 79,836 29% ­0.58% ­3.37% 2.79% 1984 24,375 79,741 31% 4.25% ­0.12% 4.37% 1985 25,787 80,665 32% 5.79% 1.16% 4.64% 1986 27,911 81,714 34% 8.24% 1.30% 6.93% 1987 30,985 84,544 37% 11.01% 3.46% 7.55% 1988 32,666 85,967 38% 5.43% 1.68% 3.74% 1989 33,844 87,438 39% 3.61% 1.71% 1.90% 1990 36,045 90,542 40% 6.50% 3.55% 2.95% 1991 37,313 92,687 40% 3.52% 2.37% 1.15% 1992 39,223 96,826 41% 5.12% 4.46% 0.65% 1993 40,565 99,798 41% 3.42% 3.07% 0.35% 1994 42,049 102,902 41% 3.66% 3.11% 0.55% 1995 44,118 107,492 41% 4.92% 4.46% 0.46% 1996 46,208 112,446 41% 4.74% 4.61% 0.13% 1997 48,096 117,289 41% 4.09% 4.31% ­0.22% 1998 50,314 122,010 41% 4.61% 4.03% 0.59% 1999 50,992 123,583 41% 1.35% 1.29% 0.06% 2000 55,404 134,986 41% 8.65% 9.23% ­0.57% 2001 57,910 140,545 41% 4.52% 4.12% 0.41% 2002 58,615 143,682 41% 1.22% 2.23% ­1.01% 2003 60,747 149,086 41% 3.64% 3.76% ­0.12% 1982­1991 5.26% 1.29% 3.98% 1991­2003 4.15% 4.04% 0.10% Source: Author's calculations. cent. Thus the bulk of growth during this period was due to fast employment growth. In contrast, most growth came from labor productivity growth after 1991. Employment growth may be further subdivided. Given that employment may be written as Employment = (1-u)*Labor Force, where "u" is the unemployment rate, any rise in employment may be driven by both a rise in the labor force or a reduc- tion in unemployment. Between 1982 and 1997 the labor force grew from 36 to 44 percent of the population, and the unemployment rate declined dramatically. So both forces contributed to the rise in employment. Further data on employment by sector show the role of migration of labor across sectors during the fast growth. Figure 5 shows the share of the labor force in each economic sector plus unemployment--and that the bulk of the workers who 184 | ANDREW M. WARNER FIGURE 5. Employment by Sector .6 Agriculture Industry, non-manufacturing Manufacturing, non-epz Manufacturing, epz .5 Services Unemployment rate .4 .3 Fractions .2 .1 .0 1982 1985 1990 1995 1997 Year Source: Author's calculations, based on data from http://ncb.intnet.mu/cso.htm . migrated to the export-processing sector came out of unemployment, not other economic sectors. This migration was sufficiently dramatic that the unemployment rate in Mauritius declined from 21 percent to 4 percent in the space of just five years: 1982­8. As in Thailand, these employment shifts in Mauritius were important contribut- ing factors to growth. As workers migrated into the export processing zones from unemployment, total GDP would have risen significantly. Since the new workers that were absorbed tended to be low-wage female garment and apparel workers, this movement of labor likely boosted welfare at the lower end of the income distribution and contributed to poverty reduction. There are no long time-series data on income distribution in Mauritius. This lim- its the scope for examining distributional change over very long periods. However, a household budget survey is available that provides data starting with 1986. The data from this survey are presented in table 6. These data suggest the following points about trends between 1986 and 2001. First, as a benchmark, this was a period dur- ing which GDP per person rose at an average annual rate of 3.68 percent. The best measure of wages in this table, monthly income per income earner, grew much faster, at 8.09 percent. At the same time, average household size was declining (4.7 persons to 3.9) and the average number of income earners was rising (1.5 to 1.9). These two trends together mean that average household income rose even faster than average wages (at 9.8 percent), and income per person rose fastest of all (at 11.19 percent). In table 6, the best counterpart to the rise in labor force participation discussed above is the rise in average number of earners per household. Apart from this, the table sheds light on other aspects of the decomposition summarized in equation 2 at the beginning of this paper. First, wages were rising faster than GDP, suggesting a TABLE 6. Household Income and Expenditure in Mauritius, 1986­2002 Index numbers Percent change 1986­7 1991­2 1996­7 2001­2 (1987­2002) 1986­7 1991­2 1996­7 2001­2 Household income Average household size 4.7 4.3 4.1 3.9 100 91.49 87.23 82.98 Income earners per household 1.5 1.6 1.9 1.9 100 106.67 126.67 126.67 Average monthly income (rupees) 3,496 6,503 10,179 14,232 9.81% 100 186.01 291.16 407.09 Median monthly income (rupees) 2,663 5,300 7,870 11,017 9.93% 100 199.02 295.53 413.71 1/2 median monthly income (rupees) 1,332 2,650 3,935 5,509 100 198.95 295.42 413.59 Percent of households with less than 1/2 median 15.7 13.7 14.2 13.6 100 87.26 90.45 86.62 monthly income Gini coefficient 0.396 0.379 0.387 0.371 100 95.71 97.73 93.69 Monthly Income per income earner 2,331 4,064 5,357 7,491 8.09% 100 174.39 229.86 321.39 Monthly Income per person 744 1,512 2,483 3,649 11.19% 100 203.32 333.77 490.60 Average earnings per month for all categories 3,343 5,767 8,701 6.58% 100.00 172.51 260.28 of employees GDP per employed person 81,714 92,687 112,446 140,545 3.68% 100 113.43 137.61 172.00 Income share Percent of total income going to: lowest 20% of households 5.6 6.4 5.9 6.2 100 114.29 105.36 110.71 highest 20% of households 44.2 43.5 46.2 44.8 100 98.42 104.52 101.36 Ratio of highest 20% to lowest 20% 7.9 6.8 7.8 7.2 100 86.08 98.73 91.14 Households consumption expenditure Average monthly expenditure (rupees) 3,035 5,225 7,846 10,129 8.37% 100 172.16 258.52 333.74 Median monthly expenditure (rupees) 2,531 4,136 6,136 8,058 8.03% 100 163.41 242.43 318.37 1/2 median monthly expenditure (rupees) 1,266 2,068 3,068 4,029 % of households with less than 1/2 median 16.6 11.9 13.3 12.4 100 71.69 80.12 74.70 monthly expenditure Source: Household Budget Survey, Mauritius. 186 | ANDREW M. WARNER trend in favor of income equality. If this household budget survey is an unbiased sam- ple of the economy, this implies that the wage share was rising at the expense of the profit share. Summary and Conclusions This paper has stressed first and foremost that increases in the share of the population employed are common in fast growing economies, as illustrated by cross-country data and extensive analysis of employment in Mauritius and Thailand. Such increases in employment can take a number of forms: a rise in the labor force as a share of popu- lation, a shift of unpaid family workers into more formal employment, the transition of the unemployed into the workforce, and more constant employment for part-time or seasonal workers. Since these workers are likely to have been drawn disproportion- ately from poorer groups, this shift is very likely to be positive for the poor. And because the opportunity cost of their labor is low, this shift is likely to be quantita- tively significant for the incomes of new workers and therefore a significant factor in poverty reduction. The paper has also pointed to three other channels through which growth can affect poverty and income distribution. These include a rise in the average wage within sectors, employment shifts from low- to high-wage sectors, or a rise in prof- its per employed person. Data on these channels are less readily available than data on labor force participation. However, the analysis shown here indicates that, in the cases of Mauritius and Thailand, the shifts in employment between sectors were not extensive. In Thailand, the only major shift was between agriculture and all other sectors. Outside of agriculture, employment shares remained roughly constant. In Mauritius, the major shift was not between sectors per se but rather between unem- ployment and employment. Again, within total employment the shares employed in different sectors did not change dramatically. Regarding growth in wage levels, data from recent stages of growth in Mauritius (1986­2001) suggest that average wage growth was more rapid than growth of the economy during this period. This implies that the profit share was declining, since employment was also rising as a share of the population during that time frame. Therefore, during 1986­2001, fast wage growth relative to GDP and rising partici- pation rates were both important channels through which growth affected the income distribution. During the earlier period of growth in Mauritius, the major event that can be documented with the available evidence is the strong rise in employ- ment during the early 1980s. In summary, based on cross-country data and a study of Mauritius and Thailand, rising rates of employment as a share of the population represent a significant chan- nel through which growth affects poverty and income distribution. To the extent that labor supply is wage-inelastic in poorer countries, this channel, rather than rising wages, is likely to be the main channel through which growth can reduce poverty. RISING LABOR FORCE PARTICIPATION AS A SOURCE OF PRO-POOR GROWTH | 187 Notes 1. The neutrality proposition has been shown in a numbers of papers using the data in Den- ninger and Squire (1996). See Roemer and Gugerty (1997), Gallup, Radelet, and Warner (1997), or Dollar and Kraay (2002). 2. See CEPAL (2002) and the essays in Bourguignon, Ferreira, and Lustig (2004) for related studies on other countries. References Bosworth, Barry, and Susan Collins. 2003. "The Empirics of Growth: An Update. Brookings Papers on Economic Activity 2003, No. 3, pp. 113-206 Bourguignon, François, Francisco Ferriera, and Nora Lustig, eds. 2004. Macroeconomics of Income Distribution in East Asia and Latin America. Washington DC: World Bank and Oxford University Press. Central Statistics Office, Port-Luis, Mauritius. (http://ncb.intnet.mu/cso.htm ) ------. Household Budget Survey. http://statsmauritius.gov.mu/hs/cpi/hbs.xls. CEPAL (Comisión Económica para América Latina y el Caribe). "Meeting the Millennium Poverty Reduction Targets in Latin America and the Caribbean." Santiago, Chile, Decem- ber 2002. Denniger, F., and L. Squire. 1996. "A New Data Set Measuring Income Inequality." World Bank Economic Review 10 (3):pp . http://www.worldbank.org/research/growth/dddeisqu.htm. Dollar, David, and Aart Kraay. 2002. "Growth Is Good for the Poor." Journal of Economic Growth 7(3): 195-225. Gallup, J., S. Radelet, and A. Warner. 1997. "Economic Growth and the Income of the Poor." Paper prepared for CAER II Project, Harvard Institute for International Development, Cambridge, MA. ILO (International Labour Organization). Yearbook of Labor Statistics. Geneva: ILO. Jitsuchon, Somchai. 2001. "An Inquiry into the Proper Measurement of Poverty in Thailand." Paper presented at the 26th. FAEA Conference, Bangkok, Thailand, . . . ------. 2002. "Sources of Thailand's Economic Growth: A Fifty-Year Perspective (1950­2000)." www.gdnet.org/pdf/jitsuchon.pdf. National Economic and Social Development Board (Thailand). http://www.nesdb.go.th/ econSocial/macro/macro_eng.php. Roemer, M., and M.K. Gugerty. 1997. "Does Economic Growth Reduce Poverty?" CAER Discussion Paper No. 4, Harvard Institute for International Development, Cam- bridge, MA. What Will Greater Integration Mean for Inequalities between and within the Richer and Poorer Countries of the "New Europe"? Social Solidarity in an Expanding Europe KEMAL DERVIS¸ I am very glad to be here at the Berlin Workshop meetings. This is a great group, the discussion has been very good, and I have learned a lot. Today I will talk about Europe and, as requested, try to present the European adventure in unification in a way that sheds light on some of the other issues that we have been debating. Despite all the indif- ference we encounter, and the negative criticism that is often expressed, I believe that we must consider the European project of economic integration to be one of the most successful--if not the most successful--social and political projects in recent history. I will always remember an event during a congress of young entrepreneurs from Europe that convened in Istanbul several years ago. There were guests there from other parts of the world as well, and about five minutes into the opening speech I was delivering, two young ladies in the second row stood up, holding hands. And one of them said: "She is German. I am French. And we [our countries] will never fight each other again, ever." The whole assembly applauded in response to this very emo- tional moment. My point is that, I believe Europeans too often take their unity for granted. But it is an enormous achievement that--on a continent where so much war, disaster, and killing have taken place--a zone of peace has been established. And the unity is quite solid. Today I think it is unimaginable that Europe will return to the bad, old ways of its past. There are many dimensions to this new zone of peace, but I believe it has been made possible, in no small part, thanks to the European welfare state, the solidarity movement, and the ideology of solidarity that has emerged more strongly in Europe than in other parts of the world. Social solidarity was, of course, much easier during the phase of rapid growth in Europe following the World War II, when the fiscal sit- uation was quite different from what it is today. But I want to emphasize that Europe, as a project, has been not only about peace but also about solidarity and the Kemal Dervis¸ was previously the Minister of Economic Affairs for the Republic of Turkey, and represented Istanbul in the Turkish Parliament. In May 2005 he was unanimously confirmed as Administrator of the United Nations Develop- ment Programme, a position he assumed in August 2005. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 191 192 | KEMAL DERVIS¸ social market economy, which even the right-of-center parties in Europe accept as a basic framework. So looking to the future, I believe these must remain part of the European project--especially with the enlargement of the European Union (EU) to include the poorer countries of Eastern Europe. And although this is a huge challenge and is more difficult to achieve than in the past, I don't think the challenge itself is bigger. Helping Greece and Ireland, and helping southern Italy and Spain was not easy. Enormous resources were deployed. But the fiscal situation is totally different today, and it is quite clear that with the four percent of GDP limit that now exists in terms of EU assistance, unification has become more difficult. Nonetheless, I don't believe we can turn our backs on the fundamental value of solidarity in Europe. Social policies must remain an important part of the European project. Clearly, some of the things that were discussed here yesterday--reforming the wel- fare state, the need for more efficient targeting of subsidies and social solidarity actions, and obviously, reducing the burden of the common agricultural policy are necessary reforms. And other needs may increase as well, including the burden of pensions and the requirements of coordinating defense and foreign policy. These evolving needs are important, and frankly, it is difficult to see how the EU can both meet the challenges of the 21st century and reduce its budget. And it is also difficult to say this, but if anything, the EU budget needs to increase over the medium term. And from the point of view of the monetary union, I believe that greater cohesion in fiscal policy inside the EU is absolutely necessary. Cohesion can only be achieved by giving the EU institutions better coordinating mechanisms and greater fiscal resources. I realize that this is a big and controversial topic, and I will not go into the details. But we find ourselves in this uncanny world situation where the United States has a deficit of five percent of its GDP. There is a huge imbalance in Asia (especially China), where exchange rates are kept at unrealistic and unsustainably low levels. And this helps perpetuate an imbalance with Europe, which is suffering because of the fixed exchange rate in China and persistent Asian surpluses. Europe is experienc- ing an excessive part of the adjustment burden in terms of the value of the euro, which, in turn, makes reforms in Europe much more difficult. But I believe that in crafting the EU stability pact and linking it to social policy, the real answer would have been to reform the pact to make it more intelligent, more flex- ible, and more medium-term--without giving up on principle--because the principles of the stability pact are absolutely necessary. But the pact should be reformed into a fiscal policy that could be adjusted at the European level. Frankly, I think there is already a case for a much more expansionary fiscal policy in Europe, for at least the next two or three years. However, I do not think this reform should be accomplished by violating the stability pact at the national levels. This type of violation is the worst thing that could happen, and unfortunately Germany and France--the two leaders of the European Monetary Union--have a good deal of responsibility for the mistakes that are made. But I think that the stability pact as formulated is too rigid and sim- plistic; and it incapacitates policy at a time when I think Europe actually needs more expansion in fiscal policy, to help with the adjustment that the world economy as a whole must undergo. Of course, I believe China also plays a key role in this; and as SOCIAL SOLIDARITY IN AN EXPANDING EUROPE | 193 long as China insists on its current reserve accumulation and unrealistic exchange rate, the overall adjustment will remain limited. It is against this background that we must view the challenges to social solidarity policy, which I believe must have a stronger European dimension to it rather than a purely national dimension. I agree very much with Michael Hoffman, who in yesterday's opening session said that there is the "democracy deficit" problem in all of this. People always want to feel that the taxes they pay are well used and reach their aims; and clearly the Euro- pean institutions have a huge problem in this regard. So, what I am saying needs to be done cannot, I am sorry to say, be done in the short run. But I believe that, as a medium-term objective, it is necessary to go beyond the national dimension and somehow bridge the gap between the citizens' need for control and confidence that their taxes are well spent and their perceptions of the functioning of European insti- tutions. Some of the institutional failures in transparency and control have been extremely damaging in this context. But the European dimension is needed, and I think Europe should not give up on the need to reconcile efficiency and social soli- darity. In some parts of Europe, in the Netherlands and Sweden, for example, the welfare systems have been reformed in an intelligent way, and unemployment is much lower than elsewhere in Europe. And yet, these are extremely generous soci- eties, where solidarity is strong. The right-wing conservative view of the world argues that being threatened in one's job or fearing the loss of one's job somehow enhances efficiency. But if we look at this argument carefully, the evidence to support it is not there. And I believe that a certain degree of job security is actually good for produc- tivity and efficiency. There are studies on this, and this may be an area that the 2006 World Development Report should carefully explore. Is there really this terrible trade-off between solidarity and efficiency? And if so, what form does it actually take? Of course, if you go to extremes and pay people unemployment benefits worth 70 percent of their wages forever, there is a strong case against that. But the other extreme is one where people constantly feel insecure and threatened. This type of work environment could hardly be characterized as productivity enhancing. In conclusion, let us look at the EU enlargement in terms of income levels and pro- ductivity. There are, of course, still differences among the EU member countries, but in the EU 15, GDP in terms of purchasing power parity (PPP) is about 23,000 euros. Then there are the new member countries: In the 10 new member countries, PPP- adjusted GDP is about 11,000, or roughly half that of the EU 15. Then you have Bul- garia and Romania, which are scheduled to become members in 2007--although most people already view them as part of the European Union. But these two coun- tries are significantly poorer than the other members, with purchasing power parity GDP per capita at a little higher than 6,000 euros (as opposed to 11,000 for the 10 new members). So Bulgaria and Romania are quite a bit poorer than the other mem- bers. Then there is Turkey, which is scheduled to begin negotiations toward EU membership. I will not go into the discussion about Turkey and Europe. But if all goes well, Turkey will join the EU sometime within the next 10 years. Turkey's PPP- adjusted per capita income is roughly equal to Bulgaria and Romania's. So in terms of the average standard of living, Bulgaria, Romania, and Turkey are about the same. 194 | KEMAL DERVIS¸ These three countries are somewhat poorer than the Czech Republic, Hungary, and Poland, but the difference is not very large. In comparison with the EU 15 average, however, Bulgaria, Romania, and Turkey stand at about 26 percent or so, which is a big gap. So with Bulgaria and Romania alone (excluding Turkey), inequality within Europe will increase--because of the large inequalities in per capita income. This, at a time when the European budget is under pressure, and when the whole set of struc- tural, regional, and agricultural funds are under debate. This, no doubt, poses a big challenge. Bulgaria and Romania will probably be treated like Poland; it is difficult to see a political outcome that is very different in the negotiations. And that is how EU integration seems to be going. There will be transition periods when newcomers do not fully benefit from all the budgetary funds in the EU. But in the end, I can't see Bulgaria and Romania being treated completely differently. Turkey, however, presents a bigger challenge. It is larger (although if you combine Bulgaria and Romania, not much bigger); and, of course, the time frame for Turkey's membership is much further into the future. So I think that in the EU negotiations with Turkey, there will be a lot of debate about the European budget, solidarity, and so forth. Turkey will also not become a member until the next EU budget cycle, which begins in 2014. In this context, one thing that is interesting (and which fol- lows up on the discussion in session 4) is that when you look at productivity per worker versus per capita income, you see a somewhat different picture--because the demographic structures of the countries are so very different. And here it is interest- ing to note that, when you look at productivity per worker rather than income per capita, it turns out that Turkey is at the level of the Czech Republic and Hungary-- even if you include agriculture. This is something that not many people realize. And in fact, the demographic structure (or the ratio of the employed population to total population) differs because the age structure in Turkey and Eastern Europe are quite dissimilar. A smaller group of people in Turkey supports a much larger population, which means per capita income is low; but productivity per worker in Turkey is actu- ally high, with productivity per worker in industry higher than in the Czech Repub- lic or Hungary. However, in Bulgaria and Romania, the situation is reversed: Productivity per worker is actually quite low. The demographic structure of the new 10 members plus Bulgaria and Romania is quite similar to the rest of Europe. These are "old" popu- lations (with large percentages of elderly people), and Europe cannot realistically expect much of a demographic bonus from the enlargement of the EU to the east, whereas in the case of Turkey, the country is just now reaching the peak of its demo- graphic transition, resulting in a much higher ratio of the active to the total popula- tion. In Turkey, the demographic bonus alone will be worth 1.5 percentage points per year of extra growth over the next decade. And if you look at the European labor markets and the structure of the European economy as a whole, this is something to think about: Because clearly, the biggest social and economic challenges in Europe are the pension problem, the health care problem for elderly people, and how to alle- viate these burdens. There are many dimensions to resolving these problems. One way is to increase the working age. This is, no doubt, one of the obvious things that SOCIAL SOLIDARITY IN AN EXPANDING EUROPE | 195 need to be done in Europe. Another component of the solution will have to be more immigration. And the real challenge will, of course, be to manage this immigration so that it helps fuel European growth. Europe must bring in people with reasonable skills that can help resolve, not augment, its problems. I have presented some thoughts on the challenges of social policy in the context of European integration and enlargement. The 2006 World Development Report will address these issues globally, so in that volume it will be important to analyze the European experience since the beginning of EU integration, and the challenges faced by the EU as it transforms itself from a union of 15 countries to one of nearly 30 countries. Integration and Inequality in the EU Accession Countries JAN SVEJNAR Over the past 15 years, the eight new European Union (EU) members from Central Europe and the Baltics have undergone an unprecedented transformation. Within this relatively short time span, they converted their centrally planned and state- owned economies into vibrant, albeit imperfect, market economies based primarily on private ownership. In the process, they abandoned their economic integration within the Soviet-induced Council on Mutual Economic Assistance (CMEA) and opened themselves up to international trade and capital flows. They also gradually integrated themselves into the EU by developing market institutions and legal frame- works consistent with those observed in Western Europe. The transition has been accompanied by dramatic swings in economic perfor- mance. In all the countries of this New Europe, gross output declined by an estimated 15­25 percent over the first two to four years; but this depression was followed by a relatively high (3­6 percent) and steady annual rate of growth in virtually all of these economies.1 Trade has been reoriented from the CMEA toward the EU, and in many countries the sum of exports and imports now exceeds their gross domestic product (GDP). Capital flows have been relatively free and have risen dramatically over the transition period. Foreign direct investment (FDI) per capita in these economies is now among the highest in the emerging markets. The Nature of Inequality of Opportunity and Outcomes The destruction of the centrally planned system, privatization of state-owned enter- prises (SOEs), and creation of new domestic and foreign-owned firms has had a Jan Svejnar is the Everett E. Berg Professor of Business Administration and Professor of Economics and Public Policy at the University of Michigan, in Ann Arbor. He is also a founder and chairman of the Center for Economic Research and Graduate Education of Charles University-Economics Institute of the Academy of Sciences of the Czech Republic (CERGE-EI), a fellow of the European Economic Association, and a research fellow of the Center for Economic Policy Research in London and the Institute for the Study of Labor (IZA) in Bonn, Germany. Berlin Workshop Series 2006 ©2006 The International Bank for Reconstruction and Development/The World Bank 197 198 | JAN SVEJNAR major effect on welfare. The communist system limited individual opportunities and possible outcomes, but the outcomes it permitted were relatively egalitarian. With the switch to a market economy, there was an increase in both opportunities and inequality of outcomes. Since the transition began, the rising trade volume and capital inflows have con- tributed in an important way to new job creation, while many old jobs were eliminated as a result of shifting demand and penetration of imports. As Jurajda and Terrell (2003) show, in the cases of the Czech Republic and Estonia, more than one-half of all workers changed jobs--moving from the old (state) sector to the new economy-- within the first five years of the transition. Some of this job mobility was voluntary (resignations) and some involuntary (layoffs). Wages have risen, on average, as has labor productivity, resulting in relatively competitive unit labor costs. Moreover, wages measured at current exchange rates still remain at a fraction of western levels, even after 15 years of adjustment. With some lag, aggregate employment has fol- lowed the path of output, first falling, then stabilizing, and in some countries even- tually rising. Elasticities of labor demand became significant as firms began maximiz- ing profits and adjusting their labor force (see Basu, Estrin, and Svejnar 2005, for a fuller account). Open unemployment was an unknown phenomenon under the Soviet system, and in most of the countries of the former Soviet bloc the unemployment rate went from zero to double digits within the first few years of the transition. In some transition economies, such as Poland and the Slovak Republic, unemployment has remained in the 14­20 percent range for over a decade. In others, such as Hungary and Slovenia, it has settled in the mid-to-high single digits. Long-term unemployment became a serious issue in all countries, and the groups that are overrepresented among the long-term unemployed are the same groups that are disproportionately present among the long-term unemployed in the Western market economies--unskilled workers, young and older people, minorities (such as the Romany), the handicapped, and so forth. Many workers, especially women, dropped out of the labor force altogether. The relatively high labor force participation observed under communism has thus decreased, with part of the decline reflecting an optimal outcome for particular households and part reflecting discouraged workers. Overall, labor force participa- tion in the countries of Central and Eastern Europe is now lower than in Western Europe. The rate of return on human capital, close to zero under the centrally planned sys- tem, rose dramatically as the transition unfolded. It has now begun to vary substan- tially across occupations, resulting in rapidly rising wage differentials (Munich, Sve- jnar, and Terrell 2005). A number of countries experienced serious bouts of inflation. The ability to pro- tect oneself from the extreme effects of the vagaries of both a disintegrating com- mand economy and nascent market system has varied substantially for individuals and social groups. In recent years, most new market economies have succeeded in reducing inflation to single digits--and a number of them to low single digits. In INTEGRATION AND INEQUALITY IN THE EU ACCESSION COUNTRIES | 199 many respects, the ability to eliminate inflationary pressures is one of the greatest policy achievements in these economies. The story of the evolution of income and consumption inequality is a fascinating one. Nomenclature apart, the Soviet-style centrally planned economies compressed income distribution, and the Central European countries in particular had some of the most egalitarian income distributions in the world. In general, their Gini coeffi- cients were below 30, and in some countries (the Czech Republic and Slovak Repub- lic, for example) they fell below 20. Many observers of the system noted that this high degree of equality, together with full employment, provided insufficient incen- tives and had a negative effect on effort. With the dynamic transformation mentioned above, income distribution was bound to become more unequal. The interesting point is that studies of the evolution of income and consumption inequality point to two very different paths during the transition. The introduction of market forces has generally resulted in major inequalities in all the transition economies. However, there has been a major difference between the Central Euro- pean and Baltic states on the one hand and the Commonwealth of Independent States (CIS) countries on the other hand.2 The former countries put in place comprehensive social safety nets with the key elements of unemployment compensation, social secu- rity benefits, and a functioning pension system.3 This safety net "rolled back" much of the market-induced inequality and resulted in actual Gini coefficients in the 25 to 35 range, thus making income distribution in the countries of New Europe compa- rable to the more egalitarian Western European economies. Conversely, in Russia and the other CIS countries, the safety net remained underdeveloped, and the system of taxes and transfers actually reinforced the effect of market forces toward more unequal income distribution. This regressive system resulted in Gini coefficients that various estimates place in the range of 40 to 60, placing Russia and many CIS economies among the countries with the most inegalitarian distribution of income. There are also parallel stories in the realm of health care and education. The com- munist system provided the population with universal health care and basic educa- tion, as well as higher education for a select group of talented and politically screened students. The system functioned relatively well until it began to disintegrate. During the transition the health care and education systems were maintained, but the qual- ity of service started to vary and the average quality decreased--less so in Central and East Europe than in Russia and the rest of CIS. Hence access to these services became differentiated in terms of quality, mirroring the increase in the inequality of the dis- tribution of income. The principal lesson provided by the transition economies is that when one starts from a government-induced, highly egalitarian system, the introduction of market forces results in greater inequality. But government programs can, to vary- ing degrees, reign in these effects. A complementary finding is that comprehensive government transfer programs, such as those maintained by the countries of New Europe, are very expensive and possibly unsustainable in the long run. The coun- tries of New Europe continue to redistribute about 40 percent of GDP, which is a higher percentage than that observed in other countries at a similar level of 200 | JAN SVEJNAR development. The percentage is comparable, however, to the share of GDP redis- tributed by Western European governments. The immediate practical problem is that in the past few years, most countries of New Europe have been running unsustainable budget deficits. So they will have to restructure some social pro- grams in order to re-establish equilibrium, let alone to reduce the share of current public finances in their GDP and launch new programs. Likely Outcomes of Further Integration With the elimination of the remaining trade and capital market restrictions, there will be more competition in the EU, and one can expect a continuation of the process observed to date. Since local firms are, on average, less efficient than foreign-owned corporations; and spillovers from foreign to domestic firms appear to be negative, one can expect additional exit of domestic firms from the market (see Sabirianova, Svejnar, and Terrell 2005, for more detail). This process will be enhanced by the completion of the original legal and regulatory reform, especially in the area of bank- ruptcy laws and proceedings. But the process will be mitigated by factors such as the entry of branches and subsidiaries of Western banks, which will result in a greater competition in the banking sector and better credit conditions for corporate and small and medium-size enterprise (SME) clients. The expected entry into the European Monetary Union requires a major reduction of national budget deficits, which will in turn require the restructuring of basic social expenditures (health care, unemployment, and social security benefits, and in some countries, the pension system). The first of these three items currently account for most of public expenditures, leaving the governments little discretion in the use of public funds. The main new effect will come with the integration of the local labor markets into the EU labor market. This is already happening with regard to Great Britain, but it will be phased in with many old EU members over the next five to seven years. Hence, even this effect, which will raise wages due to a reduction of labor supply and competitive pressures, is unlikely to have much effect in the short run. The big question is what will happen to wages and FDI. Wages have already risen, but so far the increase has been relatively moderate. Many countries now realize that they have a serious demographic problem (an aging population) and are selectively liberalizing their immigration laws. This will keep wages lower than they would be otherwise--and it will help these economies maintain competitiveness. FDI keeps on flowing to the transition economies, but some FDI is now leaving as well, moving to lower-wage areas further to the East. This is to be expected, and the question is whether the net effect will be positive (inflow) or negative (outflow) and whether the incoming FDI will increasingly generate higher value added per employee. The countries of New Europe are poor in natural resources, so they need to invest in human capital and restructure their educational systems. The awareness of this INTEGRATION AND INEQUALITY IN THE EU ACCESSION COUNTRIES | 201 need is growing. But ideas such as restructuring and the introduction of tuition (along with a system of student loans) have not yet found major political support. Notes 1. The exception is the Czech Republic, which experienced a recession in the late 1990s and so, on average, achieved a somewhat lower rate of economic growth. 2. These are the countries of the former Soviet Union, except for the Baltic states of Estonia, Latvia, and Lithuania. 3. These are Armenia, Azerbaijan, Belarus, Kazakhstan, the Kyrgyz Republic, Moldova, the Russian Federation, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan. References Basu, Swati, Saul Estrin, and Jan Svejnar. "Employment Determination in Enterprises under Communism and in Transition: Evidence from Central Europe." Industrial and Labor Relations Review 58 (3): 353-69. Jurajda, Stepan, and Katherine Terrell. 2003. "What Drives the Speed of Transition During Episodes of Massive Reallocation?" IZA Working Paper 601, Institute for the Study of Labor, Bonn, Germany. Discussion Paper No. 601, http://www.iza.org/. Munich, Daniel, Jan Svejnar, and Katherine Terrell. "Returns to Human Capital under the Communist Wage Grid and during the Transition to a Market Economy." Review of Eco- nomics and Statistics 87 (1): 100-23. Sabirianova, Klara, Jan Svejnar, and Katherine Terrell. "FDI Spillovers and Distance of Firms to the Frontier." Forthcoming. Journal of the European Economic Association Papers and Proceedings 3 (2­3): 576-86. Appendix 1 Programme International Bank for Reconstruction and Development The World Bank Development Policy Forum InWEnt--Capacity Building International, Germany* Workshop EQUITY AND DEVELOPMENT in preparation of the WORLD DEVELOPMENT REPORT 2006 Grand Hyatt Hotel, Berlin 06­08 September 2004 PROGRAMME In cooperation with the German Federal Ministry for Economic Cooperation and Development (BMZ) *InWEnt (Capacity Building International) was established in 2002 through a merger of the German Foundation for International Development and Carl Duisberg Gesellschaft. 206 | APPENDIX 1: PROGRAMME Monday, 6 September 2004 06:00 .M. P Welcome and Opening Welcome Address: Gerhard Sabathil, Head, Representation, European Commission, Germany Address: Michael Hofmann, Director General, Federal Ministry for Economic Cooperation and Development (BMZ), Germany Introduction: Francisco Ferreira, Co-Director, World Development Report 2006, The World Bank 7:30 . . P M Dinner at the Restaurant "Weinlife" Tuesday, 7 September 2004 09:30 .M. A Keynote: Francois Bourguignon, Chief Economist and Senior Vice President, The World Bank 10:00 .M. A I: What is equity? What is the role for governments in the promotion of equity? Does the role differ between developed and developing countries? Chair: Arna Hartmann, Director, Development Policy Forum, Germany Gabriella Battaini-Dragoni, Director General, Council of Europe, Brussels Denis Cogneau, Researcher, Institut de recherche pour le développement (IRD), France Stephan Klasen, Chair of Development Economics, University of Goettingen, Germany Martin Ravallion, Senior Advisor, The World Bank 12:30 . . P M Lunch 02:00 .M. P II: Equity-enhancing social transformations: historical evidence from Europe and Transition Countries Chair: Carola Kaps, Business Correspondent, Frankfurter Allgemeine Zeitung, Germany Carles Boix, Associate Professor, University of Chicago, USA Evgeny Gavrilenkov, Managing Director and Chief Economist, Troika Dialog, Russian Federation Vladimir Gligorov, Economist, The Vienna Institute for International Economic Studies, Austria APPENDIX 1: PROGRAMME | 207 03:30 .M. P Coffee break 03:45 . . P M III: Building efficient welfare states: lessons learnt Chair: Francisco Ferreira, Co-Director World Development Report 2006, Washington Rolph van der Hoeven, Manager, Technical Secretariat, World Commission on the Social Dimension of Globalization, Geneva Fabrizio Zilibotti, Chair of Economics, Institute for International Economic Studies, Sweden 05:30 .M. P End of Session 07:00 .M. P Embarkment of the boat "Esplanade" Dinner and sightseeing--boat cruise on the river Spree Wednesday, 8 September 2004 09:00 .M. A IV: Global inequalities: What can be done to reduce international inequities? Chair: Boris Pleskovic, Research Manager, The World Bank Jeremy Hobbs, Executive Director, Oxfam International, United Kingdom Anthony Shorrocks, Director, WIDER Institute, Finland Andrew Warner, Chief Economist, Millennium Challenge Corporation, USA 10:30 .M. A Coffee break 11:00 .M. A V: What will greater integration mean for inequalities between and within the richer and poorer countries of the New Europe? Chair: Michael Hofmann, Director General, Federal Ministry for Economic Cooperation and Development (BMZ), Germany Kemal Dervis¸, Former Minister of Economic Affairs of the Republic of Turkey, Member, Task Force on Global Public Goods, Stockholm Joan Esteban, Director, Institute for Economic Analysis, Spain Jan Svejnar, Professor, University of Michigan, USA 01:00 . . P M Closing Statements 01:30 .M. P Lunch Appendix 2 Participants 210 | APPENDIX 2: PARTICIPANTS Final List of Participants Gabriella Battaini-Dragoni Kemal Dervis¸ Director General of Education, [as of January 2006] Culture and Heritage, Youth Administrator and Sport United Nations Development Directorate General IV Programme Council of Europe One UN Plaza 67075 Strasbourg New York, NY 10017 France Tel: +1-212-906-5791 Tel.: +33-3-8841 2178 Fax: +33-3-8841 2750 Joan Esteban E-mail: gabriella.battaini@coe.int Director Institute for Economic Analysis (CSIC) Carles Boix and University Pompeu Fabra Associate Professor Campus UAB Department of Political Science 08193 Bellaterra The University of Chicago Spain 5828 S. University Av. Tel.: +34-93-580 66 12 60637 Chicago, IL Fax: +34-93-580 14 52 USA E-mail: joan.esteban@uab.es Tel.: +1-773-702 34 03 Fax: +1-773 702 16 89 Francisco H.G. Ferreira E-mail: cboix@midway.uchicago.edu Co-Director World Development Report 2006 François Bourguignon The World Bank Chief Economist and Senior Vice 1818 H Street, NW President 20433 Washington, D.C. Development Economics USA The World Bank Tel.: +1-202-473 43 82 1818 H Street NW Fax: +1-202-522 11 51 20433 Washington, D.C. E-mail: fferreira@worldbank.org USA Tel.: +1-202-473 3774 Evgeny Gavrilenkov Fax: +1-202-522 0906 Managing Director and Chief E-mail: fbourguignon@worldbank.org Economist Troika Dialog Denis Cogneau Vice-Director Researcher Scientific Research Développement et Insertion Higher School of Economics Internationale (DIAL) 4 Romanov pereulok Institut de recherche pour le 125009 Moscow développement (IRD) Russian Federation 4, rue dEnghien Tel.: +70-95-258 0500 75010 Paris Fax: +70-95-258 0547 France E-mail: Tel.: +33-1-5324 1465 evgeny_gavrilenkov@troika.ru Fax: +33-1-5324 1451 E-mail: cogneau@dial.prd.fr APPENDIX 2: PARTICIPANTS | 211 Vladimir Gligorov Rolph van der Hoeven Economist Manager The Vienna Institute for International Technical Secretariat Economic Studies World Commission on the Social Oppolzergasse 6 Dimension of Globalisation 1010 Vienna International Labour Organization Austria (ILO) Tel.: +43-1-53366 1023 4, Route de Morillons Fax: +43-1-53366 1050 1211 Geneva, 12 E-mail: gligorov@wiiw.ac.at Switzerland Tel.: +41-22-799 78 55 Arna Hartmann Fax: +41-22-799 71 04 Director E-mail: hoeven@ilo.org Development Policy Forum InWEnt-Capacity Building Michael Hofmann International, Germany Director General Hallerstr. 3 Global and Sectoral Tasks; European 10587 Berlin and Multilateral Development Germany Cooperation Tel.: +49-30-43 99 63 41 Federal Ministry for Economic Fax: +49-30-43 99 62 50 Cooperation and Development E-mail: (BMZ) arntraud.hartmann@inwent.org Adenauer Allee 139-141 53113 Bonn Stefan Helming Germany Director General Tel.: +49-228-53 53 800 Planning and Development Fax: +49-228-53 53 804 Deutsche Gesellschaft für Technische E-mail: michael.hofmann@bmz. Zusammenarbeit (GTZ) bund.de Dag-Hammarskjöld-Weg 1-5 65726 Eschborn Carola Kaps Germany Economic Correspondent for Central- Tel.: +49-6196-79 14 01 East and South Eastern Europe Fax: +49-6196-80 14 01 Frankfurter Allgemeine Zeitung E-mail: stefan.helming@gtz.de Rue Belliard 197 1040 Brussels Jeremy Hobbs Belgium Executive Director Tel.: +32-2-2350 782 Oxfam International Fax: +32-2-2300 832 Suite 20, 266 Banbury Road E-mail: c.kaps@faz.de OX2 7DL Oxford United Kingdom Tel.: +44-1865-313 639 Fax: +44-1865-313 770 E-mail: jeremy. hobbs@oxfaminternational.org 212 | APPENDIX 2: PARTICIPANTS Stephan Klasen Giovanna Prennushi Professor of Economics Lead Economist Department of Economics Poverty Reduction Group University of Göttingen The World Bank Platz der Göttinger Sieben 3 1818 H Street NW 37073 Göttingen 20433 Washington, D.C. Germany USA Tel.: +49-551-397 303 Tel.: +1-202-473 26 41 Fax: +49-551-397 302 E-mail: gprennushi@worldbank.org E-mail: sklasen@uni-goettingen.de Martin Ravallion Tamar Manuelyan Atinc Senior Advisor Lead Economist Development Research Group The World Bank The World Bank 1818 H Street 1818 H Street, N.W. 20433 Washington, D.C. 20433 Washington, D.C. USA USA Tel.: +1-202-458 2460 Tel.: +1-202-473 68 59 E-mail: Fax: +1-202-522 11 53 tmanuelyanatinc@worldbank.org E-mail: mravallion@worldbank.org Berk Özler Gerhard Sabathil Economist Head Development Research Group EC-Representation in Germany (Policy Team) European Commission The World Bank Unter den Linden 78 1818 H Street 10117 Berlin 20433 Washington, D.C. Germany USA Tel.: +49-30-2280 20 00 Tel.: +1-202-458 5861 Fax: +49-30-2280 22 22 E-mail: bozler@worldbank.org E-mail: eu-de-kommission@cec.eu.int Boris Pleskovic Anthony Shorrocks Research Manager Director Development Economics United Nations University, World The World Bank Institute for Development 1818 H Street, N.W. Economics Research 20433 Washington, D.C. (UNU-WIDER) USA Katajanokanlaituri 6 B Tel.: +1-202-473 10 62 00160 Helsinki Fax: +1-202-522 03 04 Finland E-mail: bpleskovic@worldbank.org Tel.: +358-9-61 59 92 00 Fax: +358-9-61 59 93 33 E-mail: shorrocks@wider.unu.edu APPENDIX 2: PARTICIPANTS | 213 Jan Svejnar Observers: Professor of Economics Hans Gerhard Dembowski University of Michigan Business Editor School D+C Development and Cooperation 724 East University Frankenallee 71-81 48109-1234 Ann Arbor 60327 Frankfurt am Main USA Germany Tel.: +1-734-936 50 42 Tel.: +49-69-7501 4365 Fax: +1-734-763 58 50 Fax: +49-69-7501 4855 E-mail: svejnar@umich.edu E-mail: hans.dembowski@fsd.de Andrew Warner Dimo Iliev Chief Economist Economist Millennium Challenge Corporation Research and Analysis Department 1000 Wilson Blvd. Council of Europe Development Bank Suite 1411 55, av Kléber 22209 Arlington, VA 75016 Paris USA France Tel.: +1-703-875 72 00 Tel.: +33-1-47 55 55 62 E-mail: warneram@mcc.gov Fax: +33-1-47 55 37 52 E-mail: dimo.iliev@coebank.org Jürgen Zattler Head of Division Hartmut Janus World Bank, IMF and Debt Issues Economist, Planning Officer Division Department Economic Development Federal Ministry of Economic and Employment Cooperation and Development Deutsche Gesellschaft für Technische (BMZ) Zusammenarbeit (GTZ) Stresemannstr. 94 Dag-Hammarskjöld-Weg 1-5 10963 Berlin 65760 Eschborn Germany Germany Tel.: +49-30-2503 2709 Tel.: +49-6196-79 12 15 E-mail: zattler@bmz.bund.de Fax: +49-6196-79 80 12 15 E-mail: hartmut.janus@gtz.de Fabrizio Zilibotti Chair of Economics Claudia von Monbart Institute or International Economic Senior Counsellor Studies External Affairs Stockholm University Vice Presidency Europe Universitetsvägen 10, House A, The World Bank Office 8th Floor 66, avenue d'Iéna 10691 Stockholm 75116 Paris Sweden France Tel.: +46-816-22 25 Tel.: +33-1-40 69 30 14 E-mail: fabrizio.zilibotti@iies.su.se Fax: +33-1-47 23 74 36 E-mail: cvonmonbart@worldbank.org 214 | APPENDIX 2: PARTICIPANTS Jean David Naudet Walter Stechel Economist Head of Division Research Department Development Policy Agence Française de Développement Federal Foreign Office (AA) (AFD) Werderscher Markt 1 5 rue Roland Barthes 10117 Berlin 75598 Paris Germany France Tel.: +49-1888-172 477 Tel.: +33-1-53 44 35 55 Fax: +49-1888-1752 477 Fax: +33-1-53 44 36 08 E-mail: 401-RL@diplo.de E-mail: naudetjd@afd.de Thomas Weithöner Christoph Rauh Desk Officer World Bank, IMF and Debt Issues Development Policy Division Federal Foreign Office (AA) Federal Ministry of Economic Werderscher Markt 1 Cooperation and Development 10117 Berlin (BMZ) Germany Stresemannstr. 94 Tel.: +49-1888-173 256 10963 Berlin Fax: +49-1888-171 982 Germany E-mail: 401-1@diplo.de Tel.: +49-30-2503 2707 Fax: +49-30-2503 2632 E-mail: christoph.rauh@bmz.bund.de Uwe Strangmann Head of Sector and Policy Division Development Economics Development Cooperation KfW Banking Group Palmengartenstrae 5-9 60325 Frankfurt am Main Germany Tel.: +49-69-7431 2160 Fax: +49-69-7431 3363 E-mail: uwe.strangmann@kfw.de Themes for the 8TH ANNUAL BERLIN WORKSHOP SERIES Berlin, Germany "DEVELOPMENT AND THE NEXT GENERATION" September 12­13, 2005 Youthful Transition in a Changing World Schooling Work Migration and the Young Forming Families Becoming Citizens T he Berlin Workshop Series 2006 presents selected papers from meetings held September 6­8, 2004, at the seventh annual forum co-hosted by InWEnt and the World Bank in preparation for the Bank's annual World Development Report. The workshop drew key researchers and policy makers from academia, development institutions, governments, and nongovernmental organizations around the world. They discussed equity and development challenges and explored how an unequal distribution of wealth might affect long-term economic growth and efficiency--all themes examined in-depth in the World Development Report 2006. This volume presents papers from Berlin Workshop sessions on Equity and Development, covering issues relating to the role of governments in the promotion of equity, equity- enhancing social transformations, building efficient welfare states, reducing global inequalities, and integration and inequality in the new Europe. IN THIS VOLUME: Introduction by Gudrun Kochendörfer-Lucius and Boris Pleskovic; opening addresses by Michael Hoffman and by Francisco Ferreira and Michael Walton; a keynote address by François Bourguignon; and papers by Gabriella Battaini-Dragoni, Dennis Cogneau, Stephen Klasen, Martin Ravallion, Carles Boix, Evgeny Gavrilenkov, Vladimir Gligorov, Rolph van der Hoeven, Jeremy Hobbs, Andrew Warner, Kernel Dervis¸, and Jan Svejnar. ISBN 0-8213-6105-8