Social Justice and Democracy: Investigating the Link WOLFGANG MERKEL/ MIRKO KRÜCK O ver the last 25 years few terms in philosophy, sociology and political science have been as widely discussed as»democracy« and»justice«. The abundance and analytical precision of theoretical reflections and empirical research on the topic are impressive. This impression fades, however, if theoretical and empirical analysis of the relationship between the two phenomena is considered: that is, what links or separates, advances or hinders democracy and(social) justice. The present study deals with these ligatures. Even a cursory overview of the research literature reveals a peculiar difference between Anglo-Saxon and German or Latin American publications. With a few exceptions, studies of transformation and democracy from the latter which deal with issues of social justice, often neglect or ignore contemporary theories of social justice. 1 Occasionally, the term »social justice« appears in the title of an article which in fact deals with inequality; the term»social inequality« is used where what is meant is»inequality of income«; differences between equality of outcome and equality of opportunity are not dealt with at all. Social justice, justicia social, social equality, inequality of income, social development, the a-priori justice of equal opportunity and the ex post equality of outcome or just distribution are cited, but they disappear in a fog of implicit synonymy. One look at the authors’ cited literature shows that they have consulted neither John Rawls nor F. A. von Hayek, Michael Walzer nor Amartya Sen, Brian Barry nor Ronald Dworkin for terminological pointers. Anglo-Saxon research into transformation and democracy is usually more cautious and precise. The terms»social justice« and»democracy« are hardly ever linked; if they are, the more precise terms»distributive inequality« or even»income inequality and democracy« are used. The 1. Guillermo O’Donnell(1993; 1998) represents one of the few exceptions, providing lucid accounts of the connections between poverty, the dysfunctional state of law and»low intensity citizenship«. 134 Merkel/Krück, Social Justice and Democracy ipg 1/2004

broad question of justice is dissected and concrete indicators such as income distribution, poverty rates, education quotas or employment rates are tested in terms of their interrelations and correlative significance. This is a comprehensible limitation and clear definition of the research topic, which is legitimate, heuristically sound and analytically lucid. But can we obtain additional insights into the development of democracy and society if we try to introduce a complex, controversial and normatively charged term such as social justice into the discussion of democracy? Do we not thereby surrender scientific modesty and analytical clarity? Since the term»social justice« has figured in democracy research for some time but without a sustainable theoretical foundation and analytical differentiation, it is legitimate to investigate the opportunities, limits and possible aporias related to the connection between democracy and social justice. Much of this has an exploratory character, also with the aim of defining the field of research. We depart from the thesis, which has been advanced in many empirical studies, that social justice and the quality of democracy in a country mutually reinforce one another. Democracy: Defining the Concept Our purpose is not to engage in a general discussion of the real, authentic and true understanding of the term»democracy«. 2 Initially, we shall be agnostic in the normative debate. We shall adopt the view that»democracy« conjoined with different adjectives describes so many different political regimes. From a wide variety of such concepts of democracy – Collier and Levitsky(1997) count more than 500 conjoined adjectives – we will initially introduce three concepts which presumably will deliver different analytical insights: 1. Electoral democracy: The necessary and sufficient criterion for this kind of democracy is elections which are general, free, secret and fair. The use of this term in political research can be justified if we have a high number of cases and analyze correlations. They can be used to recognize global patterns and regional trends, and to filter out special explanatory variables and generate hypotheses for more in-depth analysis. It is the mission of»Freedom House« to collect data for every 2. For an overview and discussion of the different»theories« and»models« of democracy see, amongst others, Held 1996; Schmidt 2000; Waschkuhn 1998. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 135

country and to evaluate them according to this minimalist conception of democracy. 2. The concept of democracy can also be dissolved in the general conception of political regime. We do not use precisely defined criteria or threshold-values 3 for autocracies and democracies, but rather use scales to evaluate the political system’s degree of polyarchy. The array of available data ranges from Polity I–IV, through Vanhanen(1990) to Freedom House(compare Schmidt 2000). Despite various methodological critiques(e.g. Lauth 2000) we have chosen the Freedom House data set because it helps us to avoid problems with threshold levels(democracy–autocracy). In addition, Freedom House itself indicates by its evaluation of the political-rights dimension those countries that can be grouped together under the heading electoral democracies. 3. Even an ambitious concept of democracy can be chosen for qualitative research. In our research on defective democracies we distinguished between»constitutional liberal democracy« in the normative case and »embedded democracy« for the purpose of analysis(Merkel 1999; Merkel and Croissant 2000; Merkel and Puhle et al. 2003). The use of such a concept of democracy limits the number of cases and defines the type of analysis. Instead of measuring correlations, causalities are researched. It is an analysis disciplined by the selected variables that goes beyond a thick description but nevertheless considers singular contexts, draws up country-specific path dependencies and investigates causal relations. All three concepts of democracy can claim legitimacy as analytical categories. But they have to be clearly defined and should not simply be used in connection with the undefined term»democracy«. Ideally, an investigation of the relationship between democracy and social justice should incorporate two steps. First, a correlational analysis incorporating a high number of cases has to be undertaken in order to filter out trends and hypotheses. Second, these hypotheses can subsequently be examined in depth in a detailed causal analysis with a small n -sample. A»most dissimilar cases« design seems to be most fruitful. However, we want to limit ourselves here to the first step, though even for this purpose the second term, social justice, needs to be defined. 3. For a discussion of the threshold-values between democracy and autocracy see, amongst others, Bendel, Croissant and Krennerich 2002; Lauth 2000; Merkel, Puhle et al. 2003, vol. 1. 136 Merkel/Krück, Social Justice and Democracy ipg 1/2004

Social Justice: the Normative Discussion It has become difficult to take in the range of theories of justice in a single glance. 4 We will therefore limit ourselves to a cursory investigation of four major contemporary theories of justice, those of F. A. von Hayek, John Rawls, Michael Walzer and Amartya Sen. We will proceed selectively, eclectically and solely oriented by the aim of our research. Our goal is a definition of social justice which is informed by the theoretical discourse, but which, beyond the purely theoretical debate, at the same time helps us to make a reasonable selection of indicators of justice suitable for an empirical–comparative analysis. Selection of the above-mentioned four theories of justice was not coincidental, but due rather to the influence of the four theorists on the justice debate and, more importantly, their contemporary character. They can be distributed in three quadrants of a four-cell matrix(Figure 1), which is vertically divided by the»distribution-sensitive versus distribution-averse« axis and horizontally by the »individual versus community« axis. Only the upper right quadrant»distribution-averse« and»community-oriented« remains empty because otherwise the matrix would be logically inconsistent. The Libertarian Position: Hayek 5 For Hayek(1971; 1996), as for all liberal and libertarian philosophers who concern themselves with questions of justice, individual autonomy takes normative precedence over the public sphere of political decision making. Limitation of this autonomy, for example by welfare state interventions, should therefore be subject to stringent examination as regards its justification. According to Hayek, institutionalized redistribution by the welfare state in order to correct market outcomes is not acceptable, for at least three reasons: The logical argument: The manifested outcomes of market exchange in society are the unintended results of the actions of individuals. Since intentionality and thus responsibility for their results are not given, they are, by definition, outside the evaluation of any theory of justice. 4. Höffe 2002; Kersting 1997 and 2000, amongst others, provide a good overview of the discussion. 5. This section is based mainly on Merkel 2001. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 137

Figure 1: Theories of Justice Distribution-averse Individual Hayek Rawls Sen Community Walzer and the communitarians Distribution-sensitive The cognitive argument: The market gives rise to a»spontaneous order in society«. From this voluntary cooperation traditions and institutions arise which construct their own»evolutionary morality«. These moral rules exceed the capabilities of reason(Hayek 1996: 6). Therefore they should be corrected neither by political majorities nor by the abstract principles of rationality. The efficiency argument: The market is a sphere of unequalled efficiency. It is cumulative and is not produced by rationalist design. Additionally, man owes some of his greatest achievements to the circumstance that he was unable to consciously plan social life(Hayek 1971: 48). The logical, cognitive and efficiency arguments all lead Hayek to an outright rejection of welfare-state intervention in markets and property, income and welfare. Hayek proposes a society of legal equality plus maximum freedom of contract(in the market), supplemented by transfer-supported minimum social protection. Any further limits on market freedom reduce liberty and are therefore illegitimate. The»meritocratic« principle of distribution therefore dominates via the market. 138 Merkel/Krück, Social Justice and Democracy ipg 1/2004

The Social-Liberal Position: John Rawls Rawls(1975; 1993) does not consider the market an appropriate arbiter of social justice. Although its efficiency in allocating resources cannot be rivalled, the market is not known for its ability to create socially just conditions. The main reason for this ethical blindness lies in the unequal and unjust conditions of access to the market. Therefore it is Rawls’s intention to equip individuals with an equal set of basic goods which will correct the uneven social starting conditions. Thus institutions need to be inscribed in the political, economic and social constitution of a society which distribute basic goods fairly in order to guarantee equal opportunities. The most important of these basic goods, according to Rawls, are basic rights, liberties and opportunities, income and wealth, and, above all, the social conditions of self-respect. The distribution of basic goods must accord with two determining principles of justice. The first, lexically prior principle demands an absolutely equal distribution of basic liberties and political rights. This is not controversial and has long been guaranteed in democracies based on the rule of law. This is different, however, for»electoral polyarchies« where the rule of law has at best only particular – territorial or class-specific – validity. 6 The second principle of distribution, which is based on socio-economic justice, has been subject to much debate. According to this principle social and economic inequalities can be permitted only if they also benefit the least well off in society. Rawls’s goal is to free individual lifechances of coincidental inequalities caused by social origin, gender or natural talent. The Communitarian Position: Michael Walzer Michael Walzer(1983; 1988; 1998), one of the most influential communitarians, 7 attempts to avoid the universalism of one particular principle of justice. His credo is: There are numerous arenas and criteria for distribution, as well as an abundance of goods and resources. There cannot and 6. Guillermo O’Donnell(1993; 1998) depicts this convincingly in theory and with reference to Latin America. Compare also Merkel, Puhle et al. 2003. 7. Walzer is not only the most influential but also the most moderate communitarian because he increasingly considers»liberal« arguments in his concept of justice. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 139

must not be an all-encompassing logic of distribution for different spheres such as citizenship, social security and welfare, money and goods, upbringing and political power. For Walzer, every sphere of life and goods has its own rules of distribution; in other words: no rule of distribution may infringe upon another sphere(1998: 161). This is especially the case for the monetary sphere. For reasons of justice the distribution of some goods, primarily health and education, should not depend on money. Their distribution has to be oriented in terms of the principles of equality and need. Even if Walzer does not attain Rawls’s theoretical stringency, in some relevant spheres he derives rules of distribution which can be compared to Rawls’s. However, he prefers context sensitivity to a generally applicable principle of distribution. In the last instance, it is a given community which determines the rules of distribution. The»Activating« Position: Amartya Sen’s»Capabilities« Since the mid-1970s Amartya Sen has been developing a concept which should open up appropriate and fair paths to democracy and solidarity for everyone in a market economy(Sen 2000). For Sen, individual, selfdetermined action is the key element in abolishing personal misery and establishing social justice. Incorporating Aristotelian thought into his theory he deals with the active side of individuals, their economic, social and political participation. Humans have to be made capable of being »agents« of their own interests. Individual action is subject to fundamental social, political and economic restrictions, according to Sen. Only the abolition of serious restrictions on action and of lack of freedom can provide the basic conditions underlying the capability of a person to recognize and realize their life-chances(Sen 1999a: 10). Capabilities are an individual’s opportunities to realize different»beings and doings«:»Capability is a set of vectors of functionings(beings and doings), reflecting the person’s freedom to lead one type of life or another«(Sen 1992: 40). Individual capability is the core of Sen’s concept of justice. To the classical liberal»negative liberty« from something(force or intervention by the state or a third party) it adds the»positive liberty« 8 of being able to do something. Sen defines social 8. The difference between»negative« and»positive« liberty was derived from Isaiah Berlin(1995): together they determine to a large extent Amartya Sen’s concept of justice. 140 Merkel/Krück, Social Justice and Democracy ipg 1/2004

justice, consequently, as»equality of capabilities … or the elimination of unambiguous inequalities in capabilities, since capability comparisons are typically incomplete«(ibid: 7). In the area of basic capabilities, Sen(2000: 50) makes a fundamental distinction between two basic liberties: the»constitutive« and the»instrumental«. The constitutive liberties include elementary liberties which have a value»in themselves«(Sen 1999a: 37). Their extension, however, also improves a human being’s opportunities to develop life plans and to extend life options, and to realize them according to their own choice (Sen 2000: 30). This self-determined choice, based on the capability of the person, is of fundamental importance for Sen’s conception of justice. Amongst the constitutive or substantive liberties Sen considers the possibility of avoiding hunger, malnutrition, curable diseases and premature death, as well as such liberties as the ability to read and write and to participate in the political process, being able to express oneself freely, and so on(ibid.: 50). While the instrumental liberties in their abstract formulation apply to societies at all levels of development, the constitutive liberties aim especially at the societies of the Third World. 9 Attaining these constitutive liberties, however, largely depends on the instrumental freedoms. In Rawls’s words: the latter must have lexical priority. The intrinsic meaning of substantive freedom as a constitutive part of social justice has to be differentiated from the instrumental effectiveness of freedom for the advancement of social justice, but both are still functionally intertwined. Political freedom, economic facilities, social opportunities, transparency guarantees and protective security(Sen 2000: 30 and 52ff.) are the institutional core of the development of capabilities and only they will offer the individual self-determined life-chances. Freedom has two different meanings for Sen: on the one hand, procedures need to be established which make possible equal formal freedom of action and decision. These are closely related to Rawls’s primary goods; on the other hand, going beyond Rawls, the real opportunities which humans enjoy due to their personal and social circumstances should not only be adjusted in a fair manner but should also be developed. This serves to prevent a situation in which, despite the constitutional existence of abstractly just norms and institutions, individuals are 9. Sen(2000), however, shows by way of concrete examples that even in the richest societies in the world constitutive liberties are by no means adequately realized for certain segments(poor, women, African-Americans) of the population. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 141

faced with uneven and insufficient opportunities to improve their lot. Whereas Rawls somehow neglects the way in which individuals can use the evenly distributed primary goods, Sen’s primary concern is that people can use these goods and transform them into new life-chances. Sen does not dissolve Rawls’s general principles of justice in the contingency of particular communities, like the communitarians, but preserves and embeds them in a context-sensitive manner according to given concrete circumstances. He develops a concept of justice which is universalistic, saturated with content but not culture-bound. This conception of justice is therefore particularly suitable for an intercultural comparison. Reconceptualizing Social Justice Our concept of justice is based on John Rawls and Amartya Sen. Although Rawls’s abstract principles of justice are convincing we see two gaps in his theory which we aim to fill via connections with Sen, namely (i) the activating component which is included in Sen’s concept of the capabilities of a person; and(ii) the particular context, which allows us to apply Rawls’s original theory, mainly applicable to developed societies, to societies at any other stage of social and economic development. The following central principles form the basis of our conception of social justice: ̈ Equal access to necessary basic goods is central in order to make possible individual choice for the free development of life-chances. ̈ Social justice requires the strengthening of individual capabilities and protecting, securing and extending personal autonomy, dignity, freedom of choice, life-chances and a variety of options. They are an important guarantee of the full protection of and unhindered access to »negative« and»positive« liberties. ̈ Our concept of social justice is very much a priori, that is, focused on equal opportunities. Ex post redistribution by passive, welfare state measures is subordinate to the former since it is less suitable for breaking up structures of class, extending life chances and avoiding poverty traps. ̈ If poverty occurs in spite of equal opportunities, which surely can be attained only in theory, it must be made a top political priority to fight it with ex post redistribution, since poverty damages individual autonomy and human dignity and can trap successive generations in indigence. 142 Merkel/Krück, Social Justice and Democracy ipg 1/2004

̈ Any redistribution beyond this is no longer subordinate to the principle of social justice, but rather to the normatively weaker principle of solidarity(Höffe 2002: 119). But even the solidarity that is organized subsidiarily must be guided by the principle»help people to help themselves« and may only fall back on paternalistic care in the last instance. ̈ Inequalities in ex post redistribution of material goods can be accepted if equal opportunities are guaranteed and poverty as well as»low intensity citizenship« are avoided. ̈ Sen’s»constitutive« and»instrumental« liberties must be measured by means of different indicators. Operationalizing Social Justice We derived five dimensions from the above-mentioned general principles of our reformulated conception of justice, which, in turn, we translated into indicators for empirical research. Figure 2 depicts the dimensions and their respective indicators in context. In the first dimension, avoidance of poverty, the percentage of infants with low birth weight and the percentage of malnourished people out of the total population are suitable indicators for determining whether the avoidance of hunger in a society is successful or not. Quotas, which massively exceed the average values of other societies, point towards problems in this dimension. This is also the case for infant mortality and life expectancy. Low life expectancy and high rates of mortality of children under five years of age indicate that hygiene, the provision of medical care, the number of doctors and nutrition are below average. The second dimension, which captures the social opportunities conditional upon education, is first measured in general using public education expenditure as a percentage of gross domestic product( gdp ). The number of students in a society indicates the extent to which continuing education is possible. On top of that, the United Nations Development Programme( undp ) uses the»Education-Index« 10 as an indicator to measure the general standard of education based on different quotas of participation, as well as indicators of outcome. The three indicators taken together enable us to depict the state of education in a country. As in the 10. For the exact composition of the un -Education Index see undp 2002: 252. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 143

Figure 2: Dimensions and Indicators of Social Justice Dimension 1. Avoidance of poverty in the substantive sense (especially hunger, malnutrition and curable diseases) 2. Social opportunities conditional upon education 3. Social opportunities conditional upon an inclusive labor market 4. Consideration of the special role of women (gender equality) 5. Social security Indicators Infants with low birth weight(%) Malnourished people as% of total population Under-five mortality rate per 1000 Life expectancy at birth, in years, both sexes Public education expenditure as% of gdp Students per 100,000 inhabitants UN Education Index Gini Index Labor force participation rate(%) Number of economically inactive per 100 active Female economic activity rate Adult literacy rate(female) Tertiary education, gross enrolment ratio (%) Public health expenditure as a% of gdp Total social security expenditure as% of gdp first dimension, below average outcomes point towards distribution problems in an area which is of fundamental importance for equal opportunities in a society. The third dimension is concerned with social opportunities conditional upon a national market’s capacity for integration. The labor force participation rate and the number of economically inactive indicate the extent to which labor markets are inclusive. Inclusivity – that is, a high rate of employment – indicates that the labor market does not exclude but rather integrates and therefore distributes life-chances and purchasing power more fairly than labor markets with lower employment rates. On top of that, we expect that a narrow distribution of income as measured 144 Merkel/Krück, Social Justice and Democracy ipg 1/2004

by the Gini Index, 11 permits us to comment on the distribution mechanisms of markets and states. A high rate of employment, a low number of economically inactive in the population and a relatively even distribution of income allow us to postulate the existence of a fair distribution of life-chances, which is to be valued positively from a social justice perspective. Social justice requires the strengthening of individual capabilities and protecting, securing and extending personal autonomy, dignity, freedom of choice, life-chances and a variety of options. The fourth dimension concerns the role of women and concentrates mainly on opening up opportunities conditional upon education and integration into the market mechanism. The»female economic activity rate« 12 , the rate of female adult literacy and the percentage of women in tertiary education indicate the state of gender equality. If the results of these indicators are below average, female life-chances in comparison to other societies are lower. Social security is dealt with in the fifth dimension. Public health expenditure and total social security expenditure, as percentages of gdp , are supposed to show the quantity of resources utilized for this purpose in a society. Again, results below average are problematic from the perspective of social justice. The overall index of social justice for a country is calculated using the average values of the respective dimensions of justice. The values of the different dimensions are themselves average values of the set of indicators that represent them. Problems that arise due to a comparison of the particular dimensions are eased with the help of a Z-value transformation. 13 11. The Gini Index shows how far the income distribution, measured on the Lorenz curve, deviates from an absolutely equal distribution of income in a society. Earned and transfer income are included in the measurement. 12. undp definition:»The proportion of the specified group supplying labour for the production of economic goods and services during the specified period«( undp 2002: 234). 13. Z-values allow for a categorical comparison of data from different dimensions. They are adjusted in such a way that the mean value of each of the included distributions is zero and the standard deviation is one. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 145

Social Justice and Democracy: Investigating the Link In democratic theory it is widely believed that, as a result of the free competition for public office characteristic of democracies, the rulers act according to the needs and wishes of the population, much more than is the case in autocratic regimes(Merkel 1999). For democratically elected politicians the provision of just and equal opportunities conditional upon the avoidance of poverty, the provision of education but also increased market integration, gender equality or the provision of social security are important sources of legitimacy, which they cannot do without. A democratically elected and, accordingly, responsive state authority is after all the only means by which a society ready to learn may act upon itself and regulate distributive ratios and life-chances according to criteria of social and political justice(Offe 1996: 143). Sen also states that the establishment of a liberal, constitutional democracy is demanded on the grounds of social justice. The selection and control of elected officials are part of this, as much as freedom of the press and freedom of expression. For Sen, the central, instrumental importance of this aspect of social justice is the connection between political freedom and civil rights on the one hand and the prevention of hunger, poverty and illiteracy on the other. Knowledge and»elementary cultural capabilities« are prerequisites of political participation. Therefore, it is against the fundamental premises of positive liberty to withhold from individuals education and a minimum of material requirements for political participation and freedom of choice. Transparency guarantees are imperative in order to protect just distribution against corruption and elite clientism which may otherwise limit and distort free access to the market and social opportunities for broad segments of the population, entrench dependency relationships and become an essential cause of encrustation and defects in the political system. In what follows we would like to examine empirically whether the type of political regime influences the degree of social justice in a society. We shall examine the hypothesis that the higher the democratic quality of a political system(measured using Freedom House data 14 ) the more just is the structure of opportunities and outcomes of a society. The influence of the democratic quality of a political regime on social justice will be 14. For methodological details on Freedom House data see http://www.freedomhouse.org/research/freeworld/2000/methodology.htm. 146 Merkel/Krück, Social Justice and Democracy ipg 1/2004

tested on several levels. 15 Initially an analysis of correlation will be used, which will test the general connection between the variables. This will be followed by a regression analysis, which will make it clear whether the democratic quality of a regime as an independent variable correlates significantly with the dependent variable social justice. Thirdly, regionally differentiated analyses will shed more light on the respective specific connections between the two variables. 16 Correlations 17 The first hypothesis postulates a positive correlation between social justice and the degree of democratization for the 124 countries we examined. Table 1: Correlation between Social Justice and Democracy Correlation Coefficient (Pearson’s r) Ranking of Social Justice Civil Liberties 0.397 Political Rights 0.393 Combined Rating 0.405 15. For a detailed description of the statistical procedures used in this study see Wagschal(1999). 16. The statistical analysis comprises 124 countries of all regime types from 16 regions (Appendix 1). Apart from the states for which no data were available for the different indicators, the classic oecd countries were removed from the statistical analysis because the indicators used are much better suited to the problems of the developing countries. The developed welfare systems of the oecd states, for example, render a poverty analysis that recurs to hunger and malnutrition senseless. The data come from sources from which one can expect a sufficiently comparable set of values. The survey period of the data is the»last available year« since it was impossible to derive a continuous timescale for all the indicators together. A social justice ranking of the included states can be found in the German version of this study(Merkel/Krück 2003), and in the online-edition of international politics and society . 17. The correlation coefficient r according to Pearson is the value most frequently used to measure the relation between two variables. The correlation coefficient r increases the closer the pairs of scores are to a straight line in a distribution diagram. This straight line is called the regression line. By definition Pearson’s correlation coefficient r can vary between –1 and+1, whereas –1 describes a perfect negative linear correlation und+1 a perfect positive correlation. If r is equal to 0 no correlation can be observed. Almost all correlations in the following tables were found to be significant or highly significant. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 147

The analysis shows that a highly significant correlation exists between the Freedom House scores and the index of social justice. This initially supports our assumption that the degree of democratization of a political regime can influence the just distribution of opportunities in a society and vice versa. A division of this connection into the single dimensions of social justice reveals that the relation between democratization and the structure of opportunities is especially due to the strong positive link between democracy and the avoidance of poverty and the provision of opportunities for education, as well as gender equality and the construction of a social safety net. The degree of democratization and market integration show no significant correlation. Table 2: Correlation between Dimensions of Social Justice and Democracy Poverty Civil Liberties Political Rights Combined Rating 0.233 0.237 0.242 Education 0.339 0.345 0.351 Market 0.121 0.115 0.120 Gender Equality 0.265 0.236 0.255 Social Security 0.394 0.396 0.405 Social Justice as Dependent Variable Whereas correlation analysis can reveal only a statistical link, regression analysis examines the direction of the relationship between the two variables. 18 We will test the extent to which the variance of social justice as a dependent variable can be explained by variations in degree of democra18. The strength of the correlation between two variables is calculated using the Pearson r. Reasonable correlations can only be calculated, however, if causality between the two variables exists. Causality means the determination of a cause–consequence relation or a cause–effect relation. Regression analysis deals with this functional correlation of one or more independent variables X and a dependent variable Y . The determination coefficient R 2 is a measure which allows us to assess the quality of the assumed causal relationship. An R 2 of 1 is equal to a perfect correlation. No correlation exists between the variables if R 2 is equal to 0. 148 Merkel/Krück, Social Justice and Democracy ipg 1/2004

tization as an independent variable. Bivariate regression analysis 19 shows that the degree of democratization has a decisive influence on the distribution of life-chances: social justice increases with the extension of political rights and civil liberties. The regression is highly significant. 15.8 percent of the variance in the social justice ranking can be explained by the extension of civil liberties; 15.4 percent of the variance of the index of social justice can be explained by the extension of political rights. Also, for the combined rating, which consists of civil liberties and political rights, the influence of the degree of democratization on the degree of social justice can be statistically proven(16.4 percent explained variance). The data therefore seem to confirm that liberal democracy, because of the free and open competition for public office, induces elected representatives to pay more attention to citizens’ opportunities to realize their life-chances. For politicians in democracies the provision of socially just structures by avoiding poverty, providing opportunities for education, gender equality and the provision of social security systems are important sources of legitimacy which they cannot forgo within the framework of democratic competition. Regional Comparison We will now test to see whether regional differences in the degree of social justice can be explained with the help of the different degrees of democratization in political regimes. To answer this question the 124 states are divided into 16 regions. 20 Our correlation analysis demonstrates that the link between democracy and social justice is further reinforced across the regions. A region with a high social justice index also exhibits a good score in the Freedom House rankings, and vice versa. Thus it also holds for a comparison between the regions: the more socially just, the more democratic, and the more democratic, the more socially just. 19. Due to the strong correlation of the two Freedom House dimensions a multivariate regression analysis is not meaningful. Civil liberties and political rights show a correlation coefficient r of 0.900**. 20.For the composition of the particular regions and an overview of the included states see Appendix 1. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 149

Table 3: Correlation between Social Justice and Democracy(Regional Mean Values) Correlation Coefficient (Pearson’s r) Social Justice Civil Liberties 0.606 Political Rights 0.677 Combined Rating 0.650 One look at the particular dimensions shows that the link between social security and the degree of democratization is further reinforced. Education and gender equality also show a significant statistical link with democratization, whereas market integration remains unaffected by it. Table 4: Correlation between Social Justice and Democracy (Regional Mean Values and Dimensions) Poverty Civil Liberties Political Rights Combined Rating 0.464 0.517 0.497 Education 0.522 0.573 0.554 Market 0.231 0.292 0.267 Gender Equality 0.498 0.551 0.531 Social Security 0.675 0.738 0.714 A comparison of the regional mean values for social justice and democracy(Appendix 2) shows the following: ̈ All European sub-regions show a high degree of democratization corresponding to a high index-score of social justice. The sample of the young post-communist countries in Europe impressively confirms our central research hypothesis. By uncritically adopting Jon Elster’s and Claus Offe’s theorem of the»dilemma of simultaneity« research on the transition to democracy 21 has drastically overestimated the difficulties of the post-communist democratization process. Empirically it has also long been repudiated by the successful consolidation of democracies in Northeast and Central Europe. The high standard of education and training and the comparatively high level of economic and 21. We also include ourselves under this(compare Merkel 1999). 150 Merkel/Krück, Social Justice and Democracy ipg 1/2004

technological development in the former communist countries are not the only important factors. The relative equality of living conditions and the more socially just texture of post-communist societies in comparison to Latin America, South- and Southeast Asia or Sub-Saharan Africa are equally important»prerequisites«(Lipset) for the successful democratization of these countries. ̈ This correlation also shows up on the African continent, where the average degree of democratization is very low. This also holds for the index of social justice: education, gender equality and social security levels are all very low. Southern Africa is an exception: although it is the most democratic and socially just part of Africa, the correlation found above can only be confirmed indirectly. If one compares Southern African democracy scores with those of other regions at the same level, one finds a degree of democracy comparable to Eastern Europe; however, as regards social justice it lags far behind. Furthermore, North Africa is clearly the least democratic region but not the most socially unjust. Here the otherwise striking link between social justice and democracy does not hold. ̈ The average level of democratisation in Asia is even below that of Africa. Justice values, however, are situated on a much higher level. 22 Although Asia as a whole is the most undemocratic in comparison to other regions the countries on this continent feature slightly above average scores of social justice. The East Asian region in particular is highly socially just despite relatively low democratic values(see Merkel 2003: 117ff.). Exactly the opposite holds for South Central Asia(especially India). East Asia should be considered a separate case from South Central and South East Asia in this respect. But even for the entire heterogeneous continent of Asia, with the much-cited exception of India, our central hypothesis proves to work well: where injustice prevails, we have extremely low levels of democracy. Where social justice prevails instead, impressive progress for democratization has been made in the last few years. ̈ If Asia is a positive exception to the rule in an interregional justice comparison, Latin America and the Caribbean are negative exceptions. Latin America features relatively high democracy scores in international comparison, but it does not manage to translate this»democratic advantage« into social justice. Although South and Central 22. For more on democracy and justice in Asia compare Merkel 2003. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 151

American indices of democracy are similar to Southern Europe, Eastern Europe or even East Asia, according to Freedom House, they are far from reaching a comparable level of social justice. Even within this region a link between social justice and democracy is hardly visible: the Caribbean is comparatively undemocratic, but just, whereas Central America is comparatively democratic but socially unjust. The main Achilles heel of sustainable democratization in Latin America is the extremely unjust distribution of primary goods(Rawls), capabilities (Sen) and life-chances(Dahrendorf 1979) in society. As long as this remains the same, democracy on the Latin American continent will remain fragile and at the same time exposed to destabilization through exogenous shocks, which it cannot counter endogenously. The interregional perspective clearly shows that a correlation exists between democracy and social justice. However, it cannot account for Latin America and the Caribbean, Southern Africa, East Asia or North Africa. The first two regions fail to translate their»democratic advantage« into social justice; East Asia and North Africa apparently do not need progressive democratization in order to create comparable socially just conditions. Conclusion The correlation analysis shows that a significant statistical correlation exists between social justice and Freedom House’s democracy scores. The general rule for the 124 countries is: the more democratic, the more socially just, and the more just, the more democratic. With reference to the particular dimensions of social justice the analysis shows that increased spending on social institutions is the main effect of democratization, while at the same time education, gender equality and the fight against poverty are advanced. Market integration remains unaffected by democratization. The regression analysis exposed a causality which in turn confirmed the postulated connection across the 124 countries: the degree of democratization as an independent variable can explain a relatively large variance in social justice in our research sample. Our analyses prove that a society grows more socially just with an increase in civil liberties and political rights. Lee Kwan Yew, the autocratic patriarch of the city-state of Singapore, presented two intertwined theses which initiated a massive political de152 Merkel/Krück, Social Justice and Democracy ipg 1/2004

bate. First, the»non-democratic« regimes in the Third World and the countries in transition can advance economic development more effectively than a democratic political order. This thesis, anecdotally underpinned by the exemplary development of the four East Asian»tigers«, cannot withstand systematic empirical examination: see Przeworski (1995), Barro(1996) and, last but not least, Amartya Sen(1999a; 1999b). Second, like other advocates of authoritarian regimes, Lee considers the traditional values of certain societies, such as those in East Asia, as largely incompatible with the norms of liberal»Western« democracy. Sen also rejects this thesis vehemently(1999b). In his essay»Democracy as a Universal Value« he lists three paramount»merits« of democracy which fortify its validity across cultures(Sen 1999b: 10ff.). First, the intrinsic value of rights to political participation and civil liberties for a self-determined »humane life«; second, the instrumental function of democracy, rendering politicians more accountable and responsible; third, the constructive role of democracy in the generation of values and the interpretation of social needs. All three reasons are convincing. Every reason in itself is legitimately superior to the self-interested relativization of»Western« democracy. Our small-scale examination of the correlation between democracy and justice reinforces Sen’s»instrumentalist« and»constructivist« arguments, showing that the degree of social justice increases with an increasing degree of democracy. Democratic political regimes put the maxims of social justice on the political agenda more swiftly and realize them more effectively than autocratic regimes want to, or are able to. Literature Barro, Robert J. 1996: Getting it Right: Markets and Choices in a Free Society. Cambridge. Bendel, Petra, Aurel Croissant and Michael Krennerich 2002: Zwischen Demokratie und Diktatur. Zur Konzeption und Empirie demokratischer Grauzonen . Opladen. Berlin, Isaiah 1995: Freiheit – Vier Versuche . Frankfurt a. M. Beyme, Klaus von 1991: Theorie der Politik im 20. Jahrhundert. Von der Moderne zur Postmoderne . Frankfurt a. M. Collier, David, and Steven Levitsky 1997:»Democracy with Adjectives: Conceptual Innovation in Comparative Research«, in: World Politics (49) 3, 430–51. Croissant, Aurel 2002:»Majoritarian and Consensus Democracy and Democratic Consolidation in Asia«, in: Asian Perspective (26), 2, 5–41. Dahrendorf, Ralf 1979: Lebenschancen. Anläufe zur sozialen und politischen Theorie . Frankfurt. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 153

Freedom House 2002: Freedom of the World – The Annual Survey of Political Rights and Liberties 2001–2002 . New Haven& London. Hayek, F.A. von 1971: Die Verfassung der Freiheit . Tübingen. Hayek, F. A. von 1996: Die Anmaßung von Wissen . Tübingen. Held, David 1996: Models of Democracy . Stanford. Höffe, Otfried 2002: Politische Gerechtigkeit . Frankfurt(3. Aufl.). Kersting, Wolfgang 1997: Recht, Gerechtigkeit und demokratische Tugend . Frankfurt. Kersting, Wolfgang(ed.) 2000: Politische Philosophie des Sozialstaats . Weilerswist. Lauth, Hans-Joachim 2000:»Die Kontrolldimension in der empirischen Demokratiemessung«, in: Hans-Joachim Lauth, Gert Pickel and Christian Welzel(eds.): Demokratiemessung. Konzepte und Befunde im internationalen Vergleich . Wiesbaden, 49–72. Merkel, Wolfgang 1999: Systemtransformation. Eine Einführung in die Theorie und Empirie der Transformationsforschung . Opladen. Merkel, Wolfgang 2001:»Soziale Gerechtigkeit und die drei Welten des Wohlfahrtsstaats«, in: Berliner Journal für Soziologie ,(11) 2, 135–57. Merkel, Wolfgang, Hans Jürgen Puhle et al. 2003: Defekte Demokratien , 2 vols. Opladen. Merkel, Wolfgang 2003: Demokratie in Asien. Ein Kontinent zwischen Diktatur und Demokratie . Bonn. Merkel, Wolfgang, and Aurel Croissant 2000:»Formale und informale Institutionen in defekten Demokratien«, in: Politische Vierteljahresschrift (41) 1, 3–30. Merkel, Wolfgang, and Mirko Krück 2003: Soziale Gerechtigkeit und Demokratie: Auf der Suche nach dem Zusammenhang, Friedrich-Ebert-Stiftung, Reihe Globalisierung und Gerechtigkeit. Bonn. O’Donnell, Guillermo 1993:»On the State, Democratization and Some Conceptual Problems. A Latin American View with Glances at Some Post-Communist Countries«, in: World Development (21), 1355–69. O’Donnell, Guillermo 1998:»Horizontal Accountability in New Democracies«, in: Journal of Democracy (9) 3, 112–26. Offe, Claus 1996:»Bewährungsproben – Über einige Beweislasten bei der Verteidigung der liberalen Demokratie«, in: Werner Weidenfeld(ed.): Demokratie am Wendepunkt. Die demokratische Frage als Projekt des 21. Jahrhunderts , Berlin, 141–57. Przeworski, Adam et al. 1995: Sustainable Democracy , Cambridge. Rawls, John 1975: Eine Theorie der Gerechtigkeit . Frankfurt a.M. Rawls, John 1993: Political Liberalism . New York. Schmidt, Manfred G. 2000: Demokratietheorien . Opladen. Sen, Amartya 1992: Inequality Reexamined . New York et al. Sen, Amartya 1999a: Development as Freedom . Oxford. Sen, Amartya 1999b:»Democracy as a Universal Value«, in: Journal of Democracy (10) 3: 3–17. Sen, Amartya 2000: Ökonomie für den Menschen . München. Sen, Amartya 2002: Rationality and Freedom . Cambridge. undp 2002: Human Development Report 2002. New York. www.undp.org/hdr2002/ english/HDR2002.html; November 2002. 154 Merkel/Krück, Social Justice and Democracy ipg 1/2004

Vanhanen, Tatu 1990: The Process of Democratization. A Comparative Study of 147 States 1980–88 . New York. Vanhanen, Tatu(ed.) 1992: Strategies of Democratization . Washington. Wagschal, Uwe 1999: Statistik für Politikwissenschaftler . Oldenburg. Walzer, Michael 1983: Spheres of Justice. A Defense of Pluralism and Equality . New York. Walzer, Michael 1988: Die Vergesellschaftung des Wohlfahrtsstaates . Frankfurt. Walzer, Michael 1998: Über Toleranz . Hamburg. Waschkuhn, Arno 1998: Demokratietheorien . München. Continent Africa Asia Europe Appendix 1: Sampled Countries Region North Africa West Africa Central Africa East Africa Southern Africa West Asia East Asia South Central Asia South East Asia South Europe East Europe North Europe Countries Included Algeria, Egypt, Libya, Morocco, Tunisia Benin, Burkina Faso, Cape Verde, Gambia, Ghana, Guinea, Ivory Coast, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra Leone, Togo Cameroon, Central African Republic, Chad, dr Congo, Gabon Burundi, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Mozambique, Rwanda, Tanzania, Uganda, Zambia, Zimbabwe Botswana, Namibia, South Africa, Swaziland Armenia, Azerbaijan, Bahrain, Georgia, Israel, Jordan, Kuwait, Lebanon, Oman, Syria, Turkey, United Arab Emirates China, Mongolia, South Korea, Taiwan Bangladesh, Bhutan, India, Iran, Kazakhstan, Kyrgyzstan, Maldives, Nepal, Pakistan, Sri Lanka, Tajikistan, Turkmenistan, Uzbekistan Burma, Cambodia, Indonesia, Laos, Malaysia, Philippines, Singapore, Thailand, Vietnam Albania, Croatia, Macedonia, Slovenia, Yugoslavia Belarus, Bulgaria, Czech Republic, Hungary, Moldova, Poland, Romania, Russia, Slovak Republic, Ukraine Estonia, Latvia, Lithuania ipg 1/2004 Merkel/Krück, Social Justice and Democracy 155

Appendix 1: Continued Continent Region Countries Included Latin America and the Caribbean Caribbean South America Bahamas, Barbados, Cuba, Dominican Republic, Haiti, Jamaica, Trinidad and Tobago Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Guyana, Paraguay, Peru, Uruguay, Venezuela Central America Belize, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama Melanesia Fiji, Papua New Guinea Note: Melanesia is not considered in the statements about particular regions, since data are available only for two countries. Appendix 2: Mean Value Comparison of Social Justice and Democracy(Regions) Pov. Edu. Market Gen. S.S. S.J. C.L. P.R. C.R. Africa –0.17 –0.63 –0.30 –0.35 –0.44 –0.38 4.43 4.50 4.47 North Africa 0.62 –0.15 –1.06 –0.77 –0.39 –0.35 5.60 6.00 5.80 West Africa –0.44 –1.29 0.05 –0.53 –0.72 –0.59 4.00 3.79 3.89 East Africa –0.70 –0.87 –0.27 –0.31 –0.56 –0.54 4.80 5.40 5.10 Central –0.64 –0.81 0.57 –0.06 –0.43 –0.28 4.64 4.45 4.55 Africa Southern 0.31 –0.02 –0.78 –0.08 –0.09 –0.13 3.00 2.75 2.88 Africa Asia 0.12 0.35 0.16 0.28 –0.24 0.13 4.53 4.48 4.51 West Asia 0.52 0.42 –0.82 –0.11 –0.12 –0.02 4.91 5.00 4.95 East Asia 0.26 1.01 0.91 1.00 0.26 0.69 3.25 3.00 3.13 156 Merkel/Krück, Social Justice and Democracy ipg 1/2004

Appendix 2: Continued Pov. Edu. Market Gen. S.S. S.J. C.L. P.R. C.R. South –0.30 –0.10 0.04 –0.05 –0.10 –0.10 5.08 5.15 5.12 Central Asia South –0.00 0.07 0.50 0.28 –1.00 –0.03 4.89 4.78 4.83 East Asia Europe 0.77 0.81 0.69 0.66 1.38 0.86 2.73 1.87 2.30 South Europe 0.73 0.53 0.34 0.47 1.56 0.73 3.00 2.20 2.60 East Europe 0.57 0.86 0.84 0.87 1.28 0.88 3.20 2.40 2.80 North Europe 1.00 1.04 0.89 0.64 1.31 0.97 2.00 1.00 1.50 Latin America 0.18 0.38 –0.48 –0.02 0.15 0.04 2.98 2.53 2.76 Central America 0.28 0.29 –0.96 –0.25 0.37 –0.06 2.75 2.00 2.38 Caribbean 0.09 0.17 0.20 0.27 –0.03 0.14 3.29 3.14 3.21 South America 0.19 0.69 –0.68 –0.06 0.12 0.05 2.91 2.45 2.68 Melanesia –0.01 –0.41 –0.21 –0.21 –0.11 –0.19 3.00 3.00 3.00 Note: Pov.: avoidance of poverty; Edu.: provision of education; Market.: market integration; Gen.: gender equality; S.S.: social security; S.J.: social justice; C.L.: civil liberties; P.R.: political rights; C.R.: combined rating: civil liberties+ political rights. The values depicted beneath the single dimensions of social justice are averages of the Z-values, referring to the particular regions. The values depicted beneath the Freedom House category are the average values of the original evaluation, referring to the particular regions: 1 is the best possible and 7 the worst evaluation. ipg 1/2004 Merkel/Krück, Social Justice and Democracy 157