1ps . W POLICY RESEARCH WORKING PAPER 28 75 Does Liberte =Egafite'? A Survey of the Empirical Links between Democracy and Inequality with Some Evidence on the Transition Economies Mark Gradstein Branko Milanovic The World Bank Development Research Group Poverty Team August 2002 POLIcy RESEARCH WORKING PAPER 2875 Abstract The effect of the distribution of political rights on more emphatically compared with the ambiguous results income inequality has been studied both theoretically of earlier research, recent evidence indicates an inverse and empirically. Gradstein and Milanovic review the relationship between other measures of democracy, existing literature and, in particular, the available based on civil liberties and political rights, and empirical evidence. The literature suggests that formal inequality. exclusion from the political process through restrictions The transition experience of Eastern European on the voting franchise appears to have caused a high countries, however, seems to some extent go against degree of economic inequality. And democratization in these conclusions. This opens possible new vistas for the form of franchise expansion has typically led to an research, namely the need to incorporate the length of expansion in redistribution, at least in the small sample democratic experience and the role played by ideology of episodes studied. In a less pronounced way, albeit and social values. This paper-a product of the Poverty Team, Development Research Group-is part of a larger effort in the group to study the effects of inequality and poverty in the world. The study was funded by the Bank's Research Support Budget under the research project "Democracy and Redistribution" (RPO 683-01). Copies of this paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Patricia Sader, room MC3-556, telephone 202-473- 3902, fax 202-522-1153, email address psader@worldbank.org. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted atgrade@bgumail.bgu.ac.il or bmilanovic@worldbank.org. August 2002. (38 pages) The Policy Research Working Paper Series dissemrates the findings of work in progress to encourage the exchange of adeas about developinent issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polisbed. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Research Advisory Staff Does Liberte=Egalite? A survey of the empirical links between democracy and inequality with some evidence on the transition economies Mark Gradstein Department of Economics, Ben Gurion University (grade(ai)bqumaiI.bgu.ac.iI) Branko Milanovic Research Department, the World Bank (bmilanovic(dAworldbank. ora) Keywords: democracy, freedom, inequality, redistribution. JEL classification: D3 1, D72 We thank Natalia Pressman for very able research assistance. The research was financed by World Bank Research Grant RPO 683-01. 1. Introduction The relationship between political structure and economic performance has long been a subject of scrutiny for researchers. In particular, the links between the extent to which an economy is democratized and its income growth and income distribution have been examined. While the studies on the first of the above links has been recently surveyed by several authors,' the second link has received much less attention. The objective of this survey is to bridge this gap by focusing on the empirical evidence with regard to the second link, asking specifically, whether or not democratization should be expected to bring about a decrease in income inequality.2 That political factors play an important role in shaping inequality should not come as much of a surprise. While economic theory has much to say about efficiency in the allocation of resources, as described by the Pareto frontier, it is much less eloquent about the precise point(s) on the frontier likely to be reached in reality, claiming that this depends on normative/political considerations. The recent political economy literature tries to tackle this latter issue by specifying the institutional details of the political process, which resolve distributional conflicts (e.g., Meltzer and Richard, 1981, Alesina and Rodrik, 1994, Persson and Tabellini, 1994). In particular, a subset of this literature, surveyed in greater detail below, deals specifically with the relationship between the distribution of political power and the distribution of economic resources, which is the focus of this survey. It might be tempting to think that a more equal distribution of political power, that is more democracy, would necessarily imply less economic inequality. However, a cursory review of the recent development history in countries of East Asia and East Europe, reveals this intuition to be misleading. In fact, some East Asian countries, 3 such as Taiwan and South Korea, have achieved a relatively equal distribution of income when autocratic regimes prevailed, and, ironically, in the post communist countries of East Europe, inequality seems to have actually increased during the course of democratization. These examples call for a more formal empirical look at the co- evolution of democratization and inequality across countries. Indeed, the empirical research of the subject has been vibrant in recent years. The issue being of an inherently interdisciplinary nature, has been addressed by sociologists, political scientists, and economists. The early research conducted in the 60's and the 70's made evident the methodological issues that have to be faced by researchers in the field, such as measurement problems, sample composition, model specification and others. More recently, measurement of some key variables (particularly of inequality, but also of democracy) has been significantly refined, which has allowed for a better reassessment of the evidence. Our main purpose is to briefly review the theories that have been developed on the relationship between democracy and inequality and, in particular, to summarize the results of the empirical research. Since the early empirical studies of the 60's and the 70's, have been comprehensively surveyed elsewhere (see Sirowy and Inkeles, 1990) the present coverage focuses on the more recent evidence. We also make an attempt to more explicitly relate the findings to existing political economic models. The survey proceeds in the following order. The next section sketches some theories on the relationship between democracy and inequality. Section 3 then discusses some methodological issues associated with empirical tests of these theories. Section 4 reviews the empirical evidence, while section 5 pays closer attention to the recent transition process in the post communist countries of East Europe. Finally, concluding remarks are presented in section 6. 4 2. Outline of the theoretical arguments Economists and social thinkers have long held that by reducing inequalities in the distribution of political power, democracy helps to reduce inequalities of wealth and status.3 The view that political democracy leads to greater economic equality was commonly held in the late eighteenth- and early nineteenth-century period, and economic consequences of democracy have been at the center of debates concerning the right to vote and to organize during the first half of the nineteenth century. One of the main arguments has been that democracy, specifically universal suffrage and the freedom to form unions, threatens private property: endowed with political power in the form of universal suffrage, those who suffer as a consequence of private property will attempt to use this power to expropriate the rich. The British electoral reforms, in particular reduction of the suffrage in the wake of the French revolution, is generally the best-known, and the most researched, case. But the same problems were present hiroughout Europe: only some 10 percent of adult male Italians, Dutch and Belgians had the right of vote as late as 1880 (Flora, Kraus and Pfenning 1983, Lindert 1989); less than 2 percent of adult males were enfranchised in Hungary on the eve of World War I (Taylor, 1967, Polonsky 1975, p.46). The dilemma was eloquently summarized in 1871 by the Spanish stateman Canavas del Castillo in a rebuttal to those who complained about electoral fraud. He wrote: "to have to choose between the permanent falsification of universal suffrage and its abolition is not to have to choose between universal suffrage and preservation of property." (Ubieto et. al., 1972, p. 731). In modem times, the theoretical case on the link between democracy and inequality has been forcefully presented by Lipset (1959) and Lenski (1966). Lipset 5 argued that democratic political structures lead to elections that serve as the expression of the democratic class struggle. Citizens vote for parties that appeal primarily to either working-class or middle-class interests. The extension of the franchise since the last century to include those with little property and other assets has vastly increased polifical competition in industrial societies, and has served to move politics toward the left. Similarly, for Lenski, the "new democratic ideology" legitimates a. major redistribution of political power in favor of the majority - namely, the disadvantaged elements of society. This increased political equality has led to more social equality because the major electoral demand made on modem political elites has been for a more egalitarian distribution of material goods. The construct of Meltzer and Richard (1981, 1983) is the economists' counterpart of the above argument. They exhibit a formal pure redistribution model, whereby taxes, which are proportional to income, are used to provide a lump-sum payment. The tax rate, therefore, represents the extent of redistribution, and voters differ with regard to their political preferences: a higher endowment (of abilities, hence income) indicates the preference of a lower tax rate. The equilibrium tax rate is a decreasing function of the ratio between the income of the. decisive voter and the average income. Democratization, specifically, expansion of the voting franchise makes a poorer individual to become decisive, thereby leading to higher redistributive taxes and lower inequality. Most recently, this model has been recast in a growth framework by Alesina and Rodrik, 1994, and Persson and Tabellini, 1994, the tax being interpreted as an investment tax. Such an approach has dynamic implications that indicates ceteris paribus a decrease in inequality over time as the voting franchise is expanded. Saint-Paul and Verdier, 1993, reached a similar conclusion in a 6 framework wherein tax revenues are used to provide a uniform level of public education. Let us now take a few moments to review a simple variant of the basic model, that has become intellectually influential in shaping a concept of the relationship between democracy and inequality. Consider an overlapping generations economy with a continuum of households, indexed by i, each comprising two members, young and old, so that the total population remains constant over time. The households differ with respect to their income endowment and, in particular, in each period t, household i allocates post-fisc income yit between current consumption, cit, and investment, kit+, according to the budget constraint: Ylt = c,t + kit+, (1) The next-period pre-tax income equals At+jk,t+j, where At+j is an exogenously given productivity parameter increasing over time. Post-fisc income is determined as a function of the pre-fisc income as well as of a politically determined redistributive policy. Specifically, yit.i=At+j [(1-0t)kj,t+,+ Otk,+,] (2) where kt+l denotes the average amount of investment and 0 8 t 1 is the redistributive parameter. The specification in (2) implies that pre-fisc income is taxed at a flat rate, Ot, and the proceeds are redistributed as a lump-sum payment (Otkt+i). Note that the redistribution is of a progressive nature, more so the larger is 8 t; in particular, a higher value of 0 t implies a leftward shift in the Lorenz curve (hence, a decrease in inequality). The household utility derives from consumption and from anticipated next- period post fisc income: U(cWt, yit+i) (3) 7 where U is assumed to satisfy the standard assumptions. Let y * denote the income requirement for enfranchisement, so that only individuals whose income is higher than y* are allowed to vote.4 We envision in each period t a two-stage decision making process. In the first stage, the enfranchised old vote over the redistributive parameter 0 t. After that is determined by a majority vote, each old member of the household allocates her income between consumption and investment according to (1) so as to maximize her utility (3) and anticipating the determination of future income according to (2). The political-economic equilibrium consists of a redistributive parameter, which is undefeated by a majority vote in a pairwise comparison with any alternative and optimal consumption-investment choices. The equilibrium analysis proceeds backwards. In the second stage, each household makes its consumption-investment decisions taking the redistributive parameter as given; focusing on the internal solution this implies the following first order condition: -U1 + U2At+1(l-0t) = (4) In the first stage voting over redistribution takes place. Employing the envelope theorem, differentiation of (3) with respect to a t reveals that the preferred value of the redistributive parameter for household i is determined from the following equation: kt+, - kit+, +0 tdkt/d 0t = 0 (5) Further analysis of (5) shows that, under mild assumptions, the preferred level of redistribution, t, is a monotonically decreasing function of income. This implies (see Gans and Smart, 1996) that a political-economic equilibrium exists whereby the median-income household among all enfranchised households is decisive in determining Vt. Furthermore, a decrease in y *, that is expansion of the franchise, 8 results in a higher value of the redistributive parameter. Intuitively, the poorer a household, the higher the preferred level of redistribution. As political participation of the poor increases because of the franchise expansion, therefore, redistribution increases with the resulting decrease in inequality. This model of redistribution predicts, therefore, an inverse relationship between democracy and inequality. A somewhat related idea which leads to the same basic result is based on a different approach to political processes. Becker, 1983, and Pelzman, 1976, assume that these political processes guided by the struggle among interest groups. Enfranchisement empowers the poor to become more effective in exerting political influence thereby increasing their share of economic resources. Plotnick, 1986, and Kristov et al., 1992, elaborate on this idea and test it empirically.5 Justman and Gradstein, 1999a, reach a slightly different conclusion in a related framework, whereby initially, as long as the income of the decisive voter is higher than the average income, inequality increases due to the redistribution from the poor to the rich and only afterwards decreases. This indicates an inverse U-shape (or curvilinear) relationship between democracy and inequality, one which had also been stipulated in Huntington and Nelson, 1976. Although it is not the main focus of the present paper, one should mention that the above causality relationship has been challenged and, in particular, it has also been argued that extreme inequalities in wealth undermine democratic political structures, see Dahl, 1971. The general argument is that concentrated economic resources may leave the door open for the politically powerful rich to prevent political reforms that extend rights and liberties to the poor. Such actions may range from direct military suppression to more subtle controls such as limitations on access to education and the control of information through ownership and censorship of the mass media. An 9 authoritarian govemment can repress reactions against these inequalities, while a democracy cannot indefinitely do so and remain democratic. In sum, the effect of inequality on democracy is anticipated because concentrated economic rewards lead to similar concentrated political resources, all of which undermine political equality.6 Acemoglu and Robinson, 1999, present a different argument to reach a similar conclusion. They claim that inequality makes democracy more costly for the ruling rich elite due to the pressure for redistribution by the poor masses that it unavoidably creates. Therefore, inequality destabilizes democracy in the sense that the elite is likely to mount a coup to prevent democracy from functioning. While our coverage is not directly concemed with examining this causality relationship, we note in passing that the jury is still out to decide whether or not it holds; see e.g. the exchange in Muller, 1995a, 1995b, and Bollen and Jaclhnan, 1995. Finally, some researchers have argued for a positive relationship between democracy and inequality. Beitz, 1982, suggests that authoritarian regimes are more likely to pursue egalitarian development policies than are democratic regimes because they may be better able to protect the interests of the poor. Thus, although democracies are more receptive to claims made by voters, they fail to treat the voters equally as sources of redistributive claims. Acemoglu and Robinson, 1998, argue that when disenfranchised, the poor have access to a revolutionary technology, thus posing a threat to the enfranchised rich. This provides an incentive for the rich to co-opt the poor by expanding the voting franchise. This model seems to suggest that extreme inequality, though not moderate levels of inequality, can be associated with eventual democratization.7 10 3. Methodological issues Several important conceptual and methodological issues face researchers interested in studying democracy and inequality, which can broadly be classified into: measurement problems and selection bias. A. Measurement problems. (a) Income inequality The difficulties in accurately measuring income inequality are well known. Data on income distribution are many times heterogeneous in regard to the timing of the observations, the definition of income and income recipient, the duration over which income is recorded, the proportion of the population covered, and the nature of the data collection procedure, which makes international comparisons exceedingly difficult. Some of the sources of this heterogeneity include whether the income figures are net (after cash transfers and taxes), gross (after transfers but inclusive of taxes) or factor only (before either taxes or transfers); whether income recipients are households or individuals; and whether the income figures are representative of the nation as a whole or refer only to certain regions or cities. Also, the dates the income.inequality figures refer to in existing data banks vary across nations. Consequently, a good portion of the observed country differences in inequality may simply reflect measurement artifacts. These data problems are yet further complicated by the fact that there are a number of data sources available, which somewhat differ in terms of coverage and some of the figures reported; they force research designs to be cross-sectional, which gives one very little leverage in addressing questions posed in dynamic terms. A major improvement took place only recently with the release of the Deininger-Squire panel 11 data on Gini coefficients and income shares (covering the period 1960-96 and some 80 countries) which represents a major step on the road to full data standardization. The data set is reviewed in detail in Deininger and Squire, 1996, who also provide a detailed guide for its use. The data problems have particularly plagued the testing of the median voter hypotheses-a crucial hypothesis for our purposes-because the way the hypothesis is theoretically couched requires access to individual factor incomes. This is so because people vote on taxes and transfers based on their factor income levels, and the decisive voter is the one with the median factor (pre-fisc and pre-transfer) income. However, these data are typically very difficult to obtain for most countries and the researchers had therefore had to resort to using the post-tax and post-transfer income (net income) as the basis on which the median voter makes her decision (see Perotti, 1993 and 1996, Easterly and Rebelo 1993; Bassett, Burkett and Putterman 1999). Thus people's decisions about the optimal level of taxes and transfers were made to depend on the distribution which emerges as a consequence of that decision! Lack of data on the middle-class gain, furthermore, led to researchers approximating the gain by the share of transfers in GDP-a very imperfect proxy bearing in mind that many transfers may be captured by the rich. Only recently (Milanovic, 1999) have these two shortcomings, thanks to the rich data set provided by the Luxembourg Income Survey, been overcome, and the hypothesis tested in an empirically correct way. (b) Political democracy The second major area of interest for the present survey has also suffered from definitional and measurement problems. Some "hard" measures of democracy, have been based, for example, on the degree of national electoral participation. On the odier 12 hand, Bollen, 1980, found that a measure of the percentage of the adult population who voted was either not related to or inversely related to a number of other dimensions of a political democracy. A certain level of participation is necessary to democracy, but above some point the meaning of variations in participation is ambiguous. The interpretation of participation figures is further clouded by the existence in some countries of a legal obligation to vote. An additional problem is that a measure of level of democracy at a single point in time is insensitive to a country's experience of democracy over time. Research that focuses on the level of political democracy implicitly assumes that democracy can have a more or less immediate effect on inequality. All countries with a high level of democracy (controlled for other variables like income) are expected to have a relatively low level of income inequality, regardless of the length of time that democratic institutions have existed. But if the egalitarian influence of democracy is in reality a long-term incremental effect, relatively new democracies should not be expected to be as egalitarian as older ones, even if they have the same level of democracy in a given year. Thus, a reason for the failure to find a significant negative effect of the level of democracy on income inequality could be the confounding influence of new, inegalitarian democracies. The studies surveyed below have all used al least one of the following three proxies for political democracy. One group of studies identifies democracy with the voting franchise. This approach is useful in instances where an historic episode of massive expansion of the franchise takes place. In the prevalent situation, however, when no such expansion is being observed the literature has typically resorted to some subjective measures of political democracy. One early such measure was developed by Bollen, 1980, who defines political democracy as "the extent to which the political 13 power of the elite is minimized and that of the non-elite is maxiniized," suggests that democracy's two main components are popular sovereignty and political liberties. The first of these concerns the extent to which the elite of a country is accountable to the non-elite. The second dimension concerns institutions through which the non-elite can influence the decisions of the elite, including free speech, a free press, and freedom of opposition. Bollen then combines three indicators of popular sovereignty and three of political liberties into a single index. In parallel, other related measures have been developed by several authors. All these measures rely invariably on subjective rankings and are based on aspects of civil liberties and political rights. One commonly used such index is drawn from Freedom House publications and is derived from the work of Gastil and co-workers. Finally, there exists the Polity m index of democracy generated by Jaggers and Gurr, 1995. Despite the fact that all these indices are derived from subjective valuations and sometimes yield very different rankings for individual countries, they exhibit a relatively high correlation between them, of over 0.80. B. Selection bias This refers to the bias of including some countries in a cross section analysis while excluding others. Typically, those included are more advanced countries with much better developed data collection procedures than the average. Early work, in particular, was usually based on quite small samples of countries, because of data availability. It is unclear how much bias this introduces into the results; however, our survey of East European countries below - usually excluded from most analyses - suggests that these omissions might be too important to neglect. 14 4. Empirical evidence The existing empirical evidence can be divided into two broad groups of studies, which are distinguished by their proxies for democracy. One strand of the literature associates democracy with voting franchise. Papers in this group typically conduct analyses of historical episodes of expansion of the voting franchise and the subsequent trend in income inequality. Another strand of the literature conceives of democracy in terms of civil liberties and political rights (rights of expression, rights to organize, free press etc.) and are based on the above described indices to measure democracy. Most of the studies belonging to this group perform cross section analysis focusing on the link between these rights and inequality measures. 4.1. Votingfranchise Studies on the effects of the expansion of the voting franchise tend to focus on the documentation of historical episodes. While this methodology in itself introduces a selection bias and precludes any sweeping generalizations, it represents the most direct effort to test economic theories of redistribution described above. Pelzman, 1980, is probably one of the first papers addressing the issue. Reviewing the evolution of government spending in a small sample of rich countries, it does not discern any systematic effect of the expansion of voting franchise. In contrast, Lindert, 1994, investigates a bigger panel of 21 rich countries in the period 1880-1930, coming to the conclusion that expansion of franchise, in particular, women suffrage, is associated with an increase in redistributive spending. Justman and Gradstein, 1 999a, following the seminal work of Dicey, 1914/1962, document the gradual expansion of voting franchise in the 19gh century 15 Britain and similarly find that it was followed by a massive adoption of redistributive programs. Husted and Kenny, 1997, examine the effect of elimination of poll taxes and literacy tests on redistribution using a panel of 46 states in the US for 1950-88. They find that it resulted in a significant increase in welfare spending, but not in other government expenditures. Lott and Kenny, 1999, examine the evolution of women suffrage in the US in the period from 1870 to 1940. They find that giving women the right to vote has resulted in a significant increase in voter's turnout and ultimately, in a substantial increase in government spending and specifically, in transfers. Although they do not explicitly focus their analysis on the inequality effects, the fact that the most significant increase was in welfare spending suggests that inequality must have been reduced. Abrams and Settle, 1999, similarly use the Switzerland's extension of the franchise to women in 1971 to test the hypothesis that redistribution is affected by the relative income of the decisive voter, arguing that women suffrage must have lowered the income of the decisive voter. They find that this extension of voting rights increased Swiss social welfare spending by 28%. To sum up, it appears that expansion of franchise is associated with a subsequent increase in redistribution - this, at least, is the reading of the available evidence with regard to a (non-representative) selection of countries. Although inequality measures do not typically feature in the above surveyed analyses, the implication of this expansion of redistribution must have been a decrease in inequality. Extent of franchise, however, is of a limited value as a proxy for democracy. One reason is that the cross-country variation in this variable is expected to be quite small - most of the world has achieved a formal full franchise. Another problem is 16 that in some countries voting franchise may very little to do with actual political democracy - the former communist countries being the case in point. Thus, one is led to consider alternative, more subtle measures based on civil liberties and political rights. 4.2. Civil liberties and political rights Work in this direction uses subjective indices for political rights, which are based on personal impressions by panels of experts of the extent of those rights in different countries. Among the most commonly used are the indices of Bollen (1980), Gastil, and Polity mI (Jaggers and Gurr, 1995). Since most of the studies confirm a high degree of correlation between these indices and reach similar conclusions with regard to them, we will not make such distinctions either. Early work on the link between political rights and inequality is concisely reviewed in Sirowy and Inkeles, 1990. Sirowy and Inkeles themselves, covering 12 previous studies come to the conclusion that "political democracy does not widely exacerbate inequality," and that "the existing evidence suggests that the level of political democracy as measured at one point in time tends not to be widely associated with lower levels of income inequality." (p. 151) Specifically, six of the surveyed studies find a negative relationship between democracy and inequality and six other studies either find a positive relationship or fail to detect any significant association. Here we extend their survey, supplementing it with more recent studies. As we shall see, this leads to a somewhat different interpretation of the evidence. We begin by summarizing the results of some of the earlier papers identified as better dealing with some of the methodological problems reviewed above. 17 Bollen and Grandjean, 1981, properly controlling for the non-monotonic relationship between economic development (by including both the logarithm of energy consumption per capita and its square as a measures) and income inequality (measured by the Gini coefficient), did not find any significant evidence to support the view that political democracy contributes to a more egalitarian distribution of income in a sample of fifty countries, once economic development is controlled for. Bollen and Jackman, 1985, reexamine the argument presented in a previous study Rubinson and Quinlan, 1977, on the potential reciprocal relation between political democracy and income inequality, by constructing a simultaneous-equations model for income inequality and political democracy. Following Kuznets' insights, they specify inequality as a curvilinear function of the level of economic development and anticipate an inverted U-shaped curve, with a positively signed coefficient for economic development and a negatively signed coefficient for its square. The most important variables included in their investigation are: political democracy, measured by Bollen's index of political democracy; income inequality (the data are from World Bank sources); socioeconomic development, measured by the gross national product per capita; and population age structure, measured by the percentage of the population aged 0-14 years. The major result of this study is that, again, no significant effects between democracy and inequality in either direction is found. Some other studies have tested the effect of the length of democratic experience on inequality. Hewitt, 1977, for example, in a sample of 25 industrialized countries failed to find a significant relationship, after properly controlling for other variables. In an important study, Muller, 1988, extended the sample to 50 countries also using the number of years of democratic experience and controlling for the extent of economic development. The indicator of a country's level of democracy is Bollen's 18 index. Muller reports that at least approximately 20 years of democratic experience are required for the egalitarian effect to occur; countries with less than a generation of democracy are almost as egalitarian, on average, than non-democracies. In other words, democratic institutions, if maintained for a relatively long time, cause some gradual reduction of income inequality, independent of level of economic development. Also, a very strong inverse association is observed between income inequality and the likelihood of stability versus breakdown of democracy, even independent of a country's level of economic development. Thus, if a democratic regime is inaugurated in a country with an extremely inegalitarian distribution of income, high inequality is likely to undermine the legitimacy of the regime and cause democratic institutions to be replaced by authoritarian rule. Robustness of Muller's findings with respect to the inclusion of a control variable (literacy rates) and an observation (Taiwan) was subsequently questioned in Weede, 1989. Moreover, using an update inequality data set, Weede failed to reproduce these findings altogether. In his response, Muller (1989) reproduces his results using an altemative measure of the length of democratic experience, even controlling for literacy rates. Summarizing the results of early research on the subject, Sirowy and Inkeles conclude therefore that evidence on the positive effect of democracy on reducing inequality had been ambiguous, although the length of democratic experience (as opposed to the state level of democracy) may have contributed to reduction in inequality. Some more recent research seems to be more conclusive. A couple of papers done over the past several years have used an improved inequality data set created in the World Bank (the Deininger-Squire) to test the hypothesis that democracy reduces 19 inequality. Note that all the papers surveyed below use the Gastil index of political rights as opposed to Bollen's used in earlier studies; however, because of the high correlation between the two (typically above 0.80), this change is unlikely to matter much. Rodrik, 1998, provides empirical support for the assertion that controlling for labor productivity, income levels, and other possible determinants, there is a robust and statistically significant association between the extent of democratic rights and wages as a share of GDP received by workers. The association exists both across countries and over time within countries (i.e., in panel regressions with fixed effects as well as in cross-section regressions). Li, Squire and Zou, 1998, explore two channels, which are likely to generate inequality: the political economy channel and the imperfect capital markets channel. The key variables associated with the political economy channel (a measure of political freedom and initial secondary schooling) and those associated with the capital market imperfection (the initial degree of inequality in the distribution of assets as measured by the distribution of land and a measure of financial market development) are all shown to be significant determinants of current inequality. The results, in particular, show that expansion of political liberties reduces inequality.8 Simpson, 1990, finds that political democracy and i.c^me inequality are related by a curvilinear function: inequality tops at intermediate levels of democracy. Nielsen and Anderson, 1995, present further limited support to this finding, although the relationship ceases to hold at a statistically significant level when controlling for other additional variables. Justman and Gradstein, 1999b, likewise detect that democracy affects inequality in a curvilinear fashion and argue that it is a better predictor of inequality than level of development as stipulated by the Kuznets hypothesis. 20 The most recent paper Lundberg and Squire, 1999, which studies the simultaneous determination of growth and inequality within the framework of an even more expanded data set relatively to Deininger and Squire, 1996, does not discem a significant relationship between democracy and aggregate measures of inequality, but they do find that expansion of democracy benefits the lowest quintile of income distribution thus reducing poverty. For the purposes of better comparability between their results and those of other papers reviewed in this section, note that Lundberg and Squire, 1999, are the first to include a large sample of post communist countries of East Europe, whose transition experience may have differed demonstrably from that of other countries - see the discussion in the following section. To sum up, while the earlier research failed to detect any significant correlation between democracy and inequality, more recent studies based on improved data sets and bigger data samples typically cautiously suggest existence of a negative relationship between the two. Two hypotheses seem to be especially promising in the light of this recent research: one that defines democracy in terms of the length of democratic experience, and another that specifies a curvilinear relationship between democracy and inequality. 5. East European countries: An enigma? The transition economies' experience appears to run counter the strand of literature reviewed so far. Perhaps the two most dramatic changes which occurred in these countries since they abandoned Communism -noticed by economists and casual observers alike-were political liberalization and large increases in income inequality. 9Figure I shows on the horizontal axis, the change in the Freedom House combined index of political freedom and civil liberties for 21 transition countries between 1989 21 and 1997, and on the vertical axis, the change in the Gini coefficient between approximately the same two dates.10 All points fall in the NE quadrant (both democracy and inequality increase), whereas we would expect them to be in the SE quadrant (greater democracy, less inequality). 22 Figure 1. Change in democracy and income inequality between 1988-89 and 1997 Turkmen* -20- 0 ~~~~~~~~~ zkh O ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ --------- --------- ------- ------------------- - 200 .. -------- -bek-----------10-0 Croa4a Belarus I -2 2 4 6 * 10 1 2 Democracy change Note. Transition countries only. Change in democracy is calculated as the difference between the Freedom House (inverted) index of civil and political freedom in 1997 and 1989. The maximum value of the index is 14 (7 for civil freedom; 7 for political freedom); the minimum value is 2 (I for erach). The maximum gain would be 12. Change in the Gini index is calculated as the difference between the Ginis in 1996-98 (when available, otherwise 1993-95) and Gini in 1988-89. Sources: for democracy, see Freedom House; for Gini, Milanovic (1998) and DEC data base. However, Figure 1 shows also an interesting feature-namely that increases in inequality were less in countries that became more democratic. The regression line is negatively sloped if all countries are included,"1 and even more so if several outliers (three Central Asian republics, Belarus, and Croatia) are left out. In other words, while democracy and inequality increased practically in all transition economies, higher increases in democracy were associated with lower increases in inequality. Moreover, if we simply plot levels of cumulative democracy achieved between 1989 and 1997 against post-transition (1995-97) levels of inequality, the relationship is, as the 23 literature suggests, a negative one (see Figure 2).12 This suggests that the real relationship between democracy and inequality is a more complex one than most of the literature, based on cross-sectional analysis, has led us to believe. There may be non- linearities such that small increases in democracy are associated with increases in inequality, but that greater democracy chips away at these increases until the relationship becomes negative. Or it may be (as suggested above) that the current level of democracy is a very imperfect proxy for the real level of democracy, namely that in order for democracy to "work" on inequality through various redistributive mechanisms, sufficient "democratic time" needs to elapse. We mean by that that a sufficiently long period of democracy is needed in order for either institutions (like tax system, social transfers) or lobbying or countervailing groups (like trade unions, political parties) to be established. Reduction of inequality, we think, does not occur by a democratic fiat. It requires democratic time (experience) to build the channels of redistribution. Thus countries that have lacked democratic time, cannot be expected to exhibit the same (low) level of inequality than countries with sufficient democratic experience even if both exhibit the same current level of democracy. 24 Figure 2. Relationship between accumulated democracy (1989-97) and level of inequality 40.0 S 0 230.0 . . 0 0~~~~~~~~~~~~~ 25 .00 20.0< 0 20 40 60 80 100 120 Cumulative democracy Index Note: Transition economies only. Cumulative democracy index is the sum of (inverted) Freedom House democracy indexes between 1989 and 1997 (the maximum cumulative democracy index would be 126). Sources: see Figure 1. To fix the ideas, and since we are discussing transition economies, consider Azerbaijan and Poland. During the transition, both have experienced increased inequality and democracy. But while Azerbaijan's barely increased democracy was accompanied. by huge increases in inequality, the reverse happened in Poland: moderate increase in inequality, and significant expansion of democracy. Could not the much greater increase in democracy, and also the more democratic climate which existed in Poland before the transition, have something to do with a smaller increase in inequality? The accumulated democratic experience-the existence of free trade unions for some twenty years; tolerance of several, even if powerless, parties before the 25 transition; the strength and vibrancy of civil society throughout the 1980's-were all elements which did not exist in Azerbaijan. Moreover, transition in Azerbaijan occurred in the context of the break-up of the Soviet Union. The local elite was now free to appropriate the rents which, in a context of a unified country, it could not. For sure, Azerbaijan might represent an extreme example of rent-grabbing simply because there was a lot to grab thanks to large oil revenues. But substantially the same appropriation by the local elites happened in all the new countries created after the break up of the Soviet Union. This is, of course, an additional sui generis factor which might explain large inequality increases in the former Soviet republics. We thus run two additional specifications of the inequality equation (on the sample of transition economies only). The first, where we replace current level of political and civil democracy with the cumulative level over the last 9 years. In the second regression, we add a dummy variable for the former Soviet republics. In both cases, we control for income levels using the 1988 values of country's GDP per capita in PPP (equal purchasing power) dollars.. The results show that cumulative liberalization enters the regressions with the correct (negative) sign. However, once the CIS dummy variable is introduced, its effect, in view of large increases in inequality in CIS countries, swamps the effect of political variables, and in neither of several formulations, are democracy variables any longer significant nor do they have the predicted sign. 13 26 Table 1. Regressing inequality on democracy: transition countries Gini chag Gini post-tra nsition level Cumulative liberalization -0.106 0.131 -0.076 0.146 (-0.89) (0.38) (-0.59) (0.37) GDP 1988 (in $PPP) x -1.03 -0.80 -1.67 -1.45 1000 (-0.98) (0.40) (-1.49) (-1.4) CIS dummy 13.15** 12.29* (2.41) (2.0) Constant 28.81** 4.00 52.25** 29.07* z______________________ (3.60) (0.32) (6.13) (2.1) R2 adjusted (F) 0.07 0.27 0.12 0.25 Number of observations 21 21 21 21 Note: t-values between brackets. * = significant at 5 percent. **=significant at I percent. The analysis of transition economies allows us to make, we believe, some broader tentative conclusions about the relationship between democracy and inequality. First, the cross-sectional relationship between current democracy and current inequality might not tell the whole story: while the current levels of democracy and inequality are negatively correlated even for transition economies, the truly important story of the 1990's is their positive correlation (both increased). Their evolution through time may differ from the conclusion we would obtain from a snapshot picture (cross-section regression). Second, and related point: to explain current inequality, one may need to use a longer-term measure of democracy: either cumulative democracy as we have done here, or length of democratic experience. But our analysis still begs the crucial point-this is: why did socialist countries, characterized by a strong concentration of power, exhibit very low levels of inequality (see e.g. Atkinson and Micklewright, 1992)? For the entire discussion of the socialist enigma would have been redundant had these countries, in accordance with theory, exhibited very high inequality which "goes together" with high concentration of 27 political power. After all, one may wonder what purpose having a concentrated political power serves if it is not accompanied by a control over economic resources, that is easy living and high welfare of the elite? Socialist countries in the past, rather than the transitional countries now, do represent an anomaly. While we do not have an answer to this question, one hypothesis - which we propose to pursue in the forthcoming research - is that, in addition to all the variables which different researchers have considered, inequality is also a reflection of social values, or the values held by the elite. We believe that socialist values which frowned upon excessive wealth accumulation, explicitly banned private ownership of the means of production, made a sharp distinction between private and personal property (the latter being control over essential items of current consumption), did prevent increases in income inequality. It can be little doubted that, for example, the well-documented (Redor 1992; Phelps Brown 1988) low income differential between skilled and unskilled workers was ideologically motivated. These differences - in most societies labor income accounts for 2/3 or more of total disposable income - are the cause of most differences in personal incomes."4 But ideology-some of which is self-serving for the elite, but most of it derived from Marx and general left-wing ethos-must have played an important role in determining acceptable income differences. While Communist societies were ideological, they were not the only ideological societies. Ideology, broadly defined as religious or ethical beliefs of the ruling elite, plays a similar role in other societies. We plan to apply, in another paper, this insight derived from the analysis of trarisition economies to a broader set of countries. 28 6. Conclusion Our review of the recent literature reveals that there are some indications regarding a positive relation between democracy and equality. This conclusion emerges both when democracy is measured by the extent of the voting franchise on the basis of specific case studies and also in cross country regressions with democracy proxied by an index based on measures of civil liberties and political rights. As such it provides empirical support for the existing theories of income redistribution A la Meltzer and Richard, 1981, as well as others. An important caveat to the positive relationship is exemplified by the transition experience of post communist countries in East Europe. While the increase in inequality that took place despite continuous democratization may serve as a warning signal against broad generalizations and premature conclusions, it could also trigger additional theoretical as well as empirical research to explain this puzzle. Note, however, that none of the recent studies has used the length of democratic experience as an explanatory variable, which further confounds the comparison with some of earlier findings. This, of course, leaves open the question of whether the increase in inequality in post communist countries is really a puzzle that begs an explanation, or perhaps just a transitory phenomenon, that will vanish as democracies there stabilize. While recent research has addressed some of the methodological problems related to the measurement of inequality and democracy, more work is needed to improve the available databases in order to allow for better comparability across countries. In the light of the findings regarding the importance of the length of democratic experience, an analysis of longitudinal data may prove important in future research. Compilation of reliable panel data may enhance our understanding of the 29 interaction between the distribution of political rights and economic resources. In particular, in the light of previous findings, a curvilinear specification of the relationship between democracy and inequality deserves further attention. Another potentially interesting direction for future research could be incorporation of the prevailing ideology in different countries into the analysis; our preliminary discussion of the post communist experience in East Europe suggests that this may be a fruitful avenue to pursue. REFERENCES Abrams, A.B. and R.F. Settle, 1999, "Women's suffrage and the growth of the welfare state," Public Choice, 100, 289-300. Acemoglu, D. and J. Robinson, 1998, "Why did the West extend the franchise? Democracy, inequality and growth in historical perspective," mimeo. Acemoglu, D. and J. Robinson, 1999, "A theory of political transition," mimeo. Alesina, A. and R. Perotti, 1994, "The political economy of growth: A critical survey of the recent literature," World Bank Economic Review, 8, 351-371. Alesina, A. and D. Rodrik, 1994, "Distributive politics and economic growth," Quarterly Journal of Economics, 109, 465-490. Atkinson, Anthony B., and John Micklewright, 1992, Economic Transformation in Eastern Europe and the Distribution of Income, Cambridge and New York: Cambridge University Press. Becker, G., 1983, "A theory of competition among pressure groups for political influence," Quarterly Journal of Economics, 98, 371-400. 30 Beitz, C.R., 1982, "Democracy in developing societies," in Freedom in the World: Political Rights and Civil Liberties, ed. R. Gastil, 145-166, New York, Freedom House. Bassett, William F., John P. Burkett, and Louis Putterman, 1999, "Income Distribution, Government Transfers, and the Problem of Unequal Influence", European Journal of Political Economy, vol. 15, pp. 207-228. Bollen, K., 1980, "Issues in the comparative measurement of political democracy," American Sociological Review, 45, 370-390. Bollen, K. and B.D. Grandjean, 1981, "The dimension(s) of democracy: Further issues in the measurement and effects of political democracy," American Sociological Review, 46, 651-659. Bollen, K. and R.W. Jackman, 1985, "Political democracy and the size distribution of income," American Sociological Review, 50, 438-457. Bollen, K. and R.W. Jackman, 1995, "Income inequality and democratization revisited," American Sociological Review, 60, 983-989. Dahl, R., 1971, Polyarchy: participation and opposition, New Haven, Yale University Press. Deininger, K. and L. Squire, 1996, "A new data set measuring income inequality," World Bank Economic Review, 10, 565-591. Dicey, A.V., 1914/1962, Lectures on the relation between law and public opinion in England during the nineteenth century, 2nd edition, London, Macmillan. (Original work published in 1914.) Easterly, William and Sergio Rebelo, 1993, "Fiscal Policy and Economic Growth: an Empirical Investigation", Journal of Monetary Economics, vol. 32, No. 3, December, pp. 417-458. 31 Firebaugh, G., 2000, "The trend in between-nation income inequality," Annual Review of Sociology, forthcoming. Flora, Peter, Franz Kraus and Winfried Pfenning (1983), State, Economy and Society in Western Europe, 1815-1975, A Data Handbook in Two Volumes, Frankfurt: Campus Verlag. Gans, J. and M. Smart, 1996, "Political equilibrium with single-crossing preferences," Journal of Public Economics, 59, 219-237. Gouveia, M. and N.A. Masia, 1998, "Does the median voter model explain the size of government? Evidence from the states," Public Choice, 97, 159-177. Helliwell, J., 1994, "Empirical linkages between democracy and economic growth," British Journal of Political Science, 24, 225-248. Hewitt, C., 1977, "The effect of political democracy and social democracy on equality in industrial societies: A cross-national comparison," American Sociological Review, 42, 420-464. Huntington, S.P. and J. Nelson, 1976, No easy choice: Political participation in developing countries, Cambridge, MA, Harvard University Press. Husted, T.A. and L.W. Kenny, 1997, "The effect of the expansion of the votiilg franchise on the size of government," Journal of Political Economy, 105, 54- 82. Jaggers, K. and T. Gurr, 1995, "Tracking democracy's third wave with Polity m data," Journal of Peace Research, 32, 469-482. Justman, M. and M. Gradstein, 1999a, "The Industrial Revolution, political transition, and the subsequent decline in inequality in 19th century Britain," Explorations in Economic History, 36, 109-127. 32 Justman, M and M. Gradstein, M., 1999b, "The democratization of political elites and the decline in inequality in modem economic growth," in E.Brezis and P. Temin, eds., Elites, Minorities, and Economic Growth, Elsevier, Amsterdam. Lenski, G., 1966, Power and privilege: A theory of social stratification, New York: Mc-Graw Hill. Li, H., L. Squire and Heng-fu Zou, 1998, "Explaining international and intertemporal variations in income inequality," Economic Journal, 106, 26-43. Lindert, P. H. 1989, "Modem Fiscal Redistribution: A Preliminary Essay", Agricultural History Center University of California, Davis, Working paper series No. 55, June. Lindert, P.H., 1994, "The rise of social spending," Explorations in Economic History, 31, 1-37. Lipset, S.M., 1959, "Some social requisites of democracy: Economic development and political development," American Political Science Review, 53, 69-105. Lott, J.R. Jr. and L. Kenny, 1999, "How dramatically did women's suffrage change the size and scope of government?" John M. Olin Law & Economics Working Paper No. 60, http://www.law.uchicago.edu/Publications/Working/index.html Lundberg, M. and L. Squire, 1999, "Growth and inequality: Extracting the lessons for policy-makers," mimeo. Meltzer, A.H. and S.F. Richard, 1981, "A rational theory of the size of government," Journal of Political Economy, 89, 914-927. Meltzer, A.H. and S.F. Richard, 1983, "Tests of a rational theory of the size of government," Public Choice, 41, 403-418. Milanovic, B., 1998, Income, Inequality, and Poverty during the Transition from Planned to Market Economy, World Bank, Washington DC. 33 Milanovic, Branko, 1999, "Do more unequal countries redistribute more? Does the median voter hypothesis hold?" World Bank policy Research Working Paper No. 2264, December. Muller, E.N., 1988, "Democracy, economic development, and income inequality," American Sociological Review, 53, 50-68. Muller, E.N., 1989, "Democracy and inequality," American Sociological Review, 54, 868-871. Muller, E.N., 1995a, "The economic determinants of democracy," American Sociological Review, 60, 966-982. Muller, E.N., 1995b, "The economic determinants of democracy: Reply to Bollen and Jackman," American Sociological Review, 60, 990-996. Nielsen, F. and A. Anderson, 1995, "Income inequality, development, and dualism," American Sociological Review, 60, 674-701. Olson, M., 1982, The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities, New Haven, Conn. Yale University Press. Pelzman, S., 1976, "Towards a more general theory of regulation," Journal of Law and Economics, 19,211-240. Pelzman, S., 1980, "The growth of government," Journal of Law and Economics, 19, 211-240. Perotti, Roberto, 1993, "Political equilibrium, income distribution, and growth", Review of Economic Studies, vol. 60:755-76 October. Perotti, Roberto, 1996, "'Growth, income distribution, and democracy: What the data say," Journal of Economic Growth, vol. 1, June, pp. 149-187. Persson, T. and G. Tabellini, 1994, "Is inequality harmiful for growth?" American Economic Review, 84, 600-621. 34 Phelps-Brown, Henry (1988), Egalitarianism and the Generation of Inequality, Oxford: Clarendon Press. Polonsky, Antony, 1975, The Little Dictators: The History of Eastern Europe since 1918, London and Boston: Routledge & Kegan Paul. Przeworski, A. and F. Limongi, 1993, "Political regimes and economic growth," Journal of Economic Perspectives, 7, 51-69. Redor, Dominique (1992), Wage Inequalities in East and West, Cambridge: Cambridge University Press. Rodrik, D., 1999, "Democracies pay higher wages," Quarterly Journal of Economics, 114, 707-738. Rubinson, R. and D. Quinlan, 1977, "Democracy and social inequality: A reanalysis," American Sociological Review, 42, 611-623. Saint-Paul, G. and T. Verdier, 1993, "Education, democracy and growth," Journal of Development Economics, 42, 399-407. Simpson, M., 1990, "Political rights and income inequality: A cross-national test," American Sociological Review, 55, 682-693. Sirowy, L. and A. Inkeles, 1990, "The effects of democracy on economic growth and inequality: A review," Studies in Comparative International Development, 25, 126-157. Taylor, A.J. P., 1967, The Habsburg Monarchy, 1809-1918, Chicago: University of Chicago Press; first published in 1948. Ubieto, Antonia, Juan Regla, Jose Maria Jover and Carlos Seco, 1972, Introduccion a la Historia de Espana, Barcelona: Editorial Teide. Weede, E., 1989, "Democracy and income inequality reconsidered," American Sociological Review, 54, 865-868. 35 Weede, E., 1990, "Redistribution and income inequality in industrial democracies," Research in Social Movements, Conflict and Change, 12, 301-326. 36 1 For recent summaries of the growth evidence the reader is referred to Alesina and Perotti, 1994, Helliwell, 1994, and Przeworski and Limongi, 1993. 2 Note that we are concerned here with a within-nation inequality. The subject of the between-nations inequality (as examined in Firebaugh, 2000, for example) is outside the scope of this paper. In addition, this work focuses on the causal link from democracy to inequality only touching in passing the issue of the rverse causality. 3 Bollen and Jackman, 1985, claim that Aristotle should be credited with being the pioneer in developing the subject. In Justman and Gradstein, 1999a, the franchise varies across periods and is endogenously determined; but for the current presentation an exogenously given constant franchise is assumed. 5 The fact that the same conclusion is supported by an alternative political model is comforting in the light of the somewhat inconclusive evidence regarding the median voter theory, on which Meltzer-Richard's model rests; see Meltzer and Richard, 1983, for a supporting view, and Gouveia and Masia, 1998, for a dissenting evidence. 6 This line of argument has been elaborated upon and tested empirically in numerous studies, but lies outside the scope of the present paper. ' Building on Olson's theory of redistributive coalitions, Weede, 1990, does find that democracy is negatively associated with the income share of the poorest quintile. 8 Although the variables associated with the financial market imperfection argument were found to have an even greater effect on inequality. 9 A question can be legitimately asked whether the calculated changes in inequality in transition economies reflect true increases in inequality. This, for two reasons. Before the transition, many consumption items were rationed and there were sizeable benefits in kind received by the elite. The actual inequality, particularly in the countries of the former Soviet Union, was greater than the recorded inequality. On the other hand, after the transition, a number of income sources is not well covered in surveys, and some of those (e.g. home consumption) may be pro-poor. Thus, the current values may exaggerate true inequality. Both points are correct but they cannot account for more than a few Gini points. The massive extent of the change in recorded inequality leaves therefore little doubt that inequality did indeed increase; in addition, we notice that even in the countries where both the pre- and post-transition Ginis were measured with the same--rather high-level of accuracy (Poland, Hungary, Czechoslovakia), inequality went up. The biases in the comparison of inequality figures before and after the transition are discussed in some detail in Milanovic (1998, Appendix 1). 10 The Freedom House index is inverted so that its higher values show greater democracy and civil liberties. "The regression is: Change in Gini=24.03 - 0.281 change in democracy.- 0.001 GDP in 1988 (in PPP dollars) However, neither coefficient (except the constant) is statistically significant, and R2 is only 0.14. 12 The same result obtains if we use simply level of democracy in 1997 instead of cumulative democracy between 1989 and 1997. 13 We have tried several democracy formulations which are not reported here: simply change in level of democracy between 1989 and 1997, number of important 37 govemment changes both alone and interacted with cumulative liberalization. All the formulations give basically the same results. 14 Some of ideological limitations did pay a useful political role too: not allowing people to accumulate private wealth made it easier to control them politically, because people who had no economic back-up in the case of political disgrace were more likely to toe the political line. 38 Policy Research Working Paper Series Contact Title Author Date for paper WPS2854 Rich andPowerful? Subjective Michael Lokshin June 2002 C. Cunanan Power and Welfare in Russia Martin Ravallion 32301 WPS2855 Financial Crises, Financial Luc Laeven June 2002 R. Vo Dependence, and Industry Growth Daniela Klingebiel 33722 Randy Kroszner WPS2856 Banking Policy and Macroeconomic Gerard Caprio, Jr. June 2002 A. Yaptenco Stability: An Exploration Patrick Honohan 31823 WPS2857 Markups, Returns to Scale, and Hiau Looi Kee June 2002 M. Kasilag Productivity: A Case Study of 39081 Singapore's Manufacturing Sector WPS2858 The State of Corporate Governance: Olivier Fremond June 2002 G. Gorospe Experience from Country Mierta Capaul 32623 Assessments WPS2859 Ethnic and Gender Wage Mohamed Ihsan Ajwad June 2002 Z. Jetha Disparities in Sri Lanka Pradeep Kurukulasuriya 84321 WPS2860 Privatization in Competitive Sectors: Sunita Kikern June 2002 R. Bartolome The Record to Date John Nellis 35703 WPS2861 Trade-Related Technology Diffusion Maurice Schiff June 2002 M. Kasilag and the Dynamics of North-South Yanling Wang 39081 and South-South Integration Marcelo Olarreaga WPS2862 Tenure, Diversity, and Commitment: Somik V. Lall June 2002 Y. D'Souza Community Participation for Urban Uwe Deichmann 31449 Service Provision Mattias K. A. Lundberg Nazmul Chaudhury WPS2863 Getting Connected: Competition Frew Amare Gebreab June 2002 P. Sintim-Aboagye and Diffusion in African Mobile 38526 Telecommunications Markets WPS2864 Telecommunications Reform in Mary M. Shirley June 2002 P. Sintim-Aboagye Uganda F. F. Tusubira 38526 Frew Amare Gebreab Luke Haggarty WPS2865 Bankruptcy Around the World: Stijn Claessens July 2002 A. Yaptenco Explanations of its Relative Use Leora F. Klapper 31823 WPS2866 Transforming the Old into a David Ellerman July 2002 N. Jameson Foundation for the New: Lessons Vladimir Kreacic 30677 of the Moldova ARIA Project Policy Research Working Paper Series Contact Title Author Date for paper WPS2867 Cotton Sector Strategies in West Ousmane Badiane July 2002 A. Lodi and Central Africa Dhaneshwar Ghura 34478 Louis Goreux Paul Masson WPS2868 Universal(ly Bad) Service: George R. G. Clarke July 2002 P. Sintim-Aboagye Providing Infrastructure Services Scott J. Wallsten 38526 to Rural and Poor Urban Consumers WPS2869 Stabilizing Intergovernmental Christian Y. Gonzalez July 2002 B. Mekuria Transfers in Latin America: David Rosenblatt 82756 A Complement to National/ Steven B. Webb Subnational Fiscal Rules? WPS2870 Electronic Security: Risk Mitigation Thomas Glaessner July 2002 E. Mekhova In Financial Transactions-Public Tom Kellermann 85984 Policy Issues Valerie McNevin WPS2871 Pricing of Deposit Insurance Luc Laeven July 2002 R. Vo 33722 WPS2872 Regional Cooperation, and the Role Maurice Schiff July 2002 P. Flewitt of International Organizations and L. Alan Winters 32724 Regional Integration WPS2873 A Little Engine that Could ... Liesbet Steer August 2002 H. Sutrisna Domestic Private Companies and Markus Taussig 88032 Vietnam's Pressing Need for Wage Employment WPS2874 The Risks and Macroeconomic David A. Robalino August 2002 C. Fall Impact of HIV/AIDS in the Middle Carol Jenkins 30632 East and North Africa: Why Karim El Maroufi Waiting to Intervene Can Be Costly