STREAMS 9 Political Sociology Inequality, conflict and social cleavages in a Nordic and comparative perspective

STREAMS 9 Political Sociology Inequality, conflict and social cleavages in a Nordic and comparative perspective Barbra S. Frisvold & Håkon Leiulfsrud

Inequality, conflict and social cleavages in a Nordic and comparative perspective Abstract A common view is that Western societies during the last decades increasingly can be characterized by individualization, liberalization of the economy, loss of solidarity and rising inequality, and that the welfare state and the associated values of solidarity and equality are under great pressure. The purpose of this paper is to explore issues of inequality, conflict and social cleavages in the Nordic countries. The basis for the exploration is the 1999 module on Social Inequality from ISSP (International Social Survey Program). Our goal is to outline the ideologies of inequality and social cleavages in different regimes, based on Gösta Esping- Andersen s classification of liberal, conservative and social democratic welfare regimes. Sweden, Denmark and Norway are analyzed as typical examples of the social democratic regime, Great Britain, the USA and Canada as liberal and Germany and Spain as conservative. We ask if the Nordic countries today represent a specific and uniform model in terms of public perceptions of opportunity structures, ideologies of inequality and social conflicts and cleavages. Barbra S. Frisvold & Håkon Leiulfsrud Department of sociology and political science Norwegian University of Science and Technology N-7491 Trondheim, Norway Barbra.Frisvold@svt.ntnu.no Hakonl@svt.ntnu.no Paper to be presented at the sixth European Sociological Association Conference. Stream 9: Political Sociology. Murcia, Spain, September 23.-26. 2003.

1. Introduction In this paper we take a closer look at public perceptions of inequality, conflict and social cleavages in the Nordic countries, specifically Norway, Sweden and Denmark. We are in this respect mainly interested in exploring the question of whether it makes sense to speak of a Nordic value system. Our approach is to describe some differences within the Nordic countries and to place these in a comparative context with other nations and welfare regimes. Since Finland and Iceland unfortunately are not included in the survey we use, we have had to limit our analysis to Norway, Sweden and Denmark. Our data can not give a full picture of the entire Nordic model, but can give us valuable information about internal variations in this model. We claim that perceptions of conflict levels in society, opportunity structures, and reasons given by the public for inequality, can tell us an interesting story about discourses of inequality in different nations. In our study of social cleavages, we will give special attention to current perceptions of class conflicts, as this is an important element in many diagnoses of social change. According to Miller (1995, p. 99), people in advanced industrial societies share a basic grammar of justice. On a very general level, justice is a valued norm in all societies. Presented with a real problem, however, people in different places have different images of society, and therefore of what is just or fair. On this level, researchers have found substantial cross-national differences in values and attitudes (Gelissen 2002; Svallfors 1997; Listhaug 1990; Gundelach 1994). During the last two decades the interest in systematic variations in public attitudes and values between countries has been increasing. In the same period, attempts have been made to understand attitudinal differences in the light of theories of welfare state regimes. In a regime perspective, such differences are expected to follow a certain pattern, where groups of countries are internally similar, but different from other groups. In this way, cross-national differences are treated in a more coherent and systematic manner. In a comparative context, the Nordic countries are often described as egalitarian societies with comprehensive welfare-states. This understanding of a Nordic model is often built on institutional similarities. To take this a step further, one can ask whether it makes sense to speak of a Nordic model of welfare attitudes or a Nordic model of solidarity structures as well. A useful distinction goes between studies of social structure or institutional patterns on one hand, and value and attitude patterns on the other. It will be an empirical matter to decide how good the correspondence between the two is, and the relation might change over a period of time. We are here mainly concerned with attitudes and values related to inequality, conflict and social cleavages. We believe this has been a neglected perspective in comparative research, and that this focus can contribute to the discussion of social change related to individualization, welfare state crisis and solidarity structures. In the following paragraphs we will first give a short overview of some characteristics of regime theories. Here we also go further into the question of the relation between attitude patterns and institutional structures. Next, we focus on the Nordic model and some of its institutional features before we look at Nordic attitudes and values in earlier research. In the next part we present previous findings on attitudes of inequality, conflict and social cleavages in the Nordic countries, before we present our empirical data on Sweden, Norway, Denmark, Germany, Spain, Great Britain, the USA and Canada. Finally, we will discuss the results and interpretations in relation to the Nordic model and the wider question of institutional and value patterns. 1

2. Ideal types and real types in country classifications Until the 1980s, comparative studies of welfare state regimes and comparative analysis of values and attitudes were to a large degree separated strands of research. The relations between values and socio-economic variables had been researched at the level of individuals and groups, but cross-national variation was seldom explained in a thorough way (Gundelach 1994). Due to the development of several large-scale survey projects, there is now more suitable data material available for a coupling of these two traditions. The World Values Study (WVS), The European Values Study (EVS) and the International Social Survey Program (ISSP) are some central sources used extensively in comparative research on values to map cross-national attitude differences (see e.g. Aalberg 2001). Regime theories have been important in several fields of comparative research, and a standard reference is Gøsta Esping-Andersen s (1990) The three worlds of welfare capitalism. A welfare regime is here characterized by a specific logic of organization, stratification and societal integration, and can be defined as a complex of legal and organizational features that are systematically interwoven (Esping-Andersen 1990). Regimes are thus a kind of ideal types that represent something more than the sum total of social policies (for a discussion, see Gelissen 2002, p. 27). Esping-Andersen developed the idea that differences between welfare states clustered into three separate regimes; the social-democratic regime type, the liberal type and the conservative type. According to Esping-Andersen (1990), equality is the core issue of all welfare states. The definitions of equality vary, however, and because of that, the actual distributions and outcomes are different in different welfare states (Esping-Andersen 1990; Arts & Gelissen 2001). The social-democratic regime type is in short characterized by a high degree of decommodification (market independence), strong universalism and a marginalization of private welfare arrangements (Esping-Andersen 1990, 1999, pp 80). Regime classifications can be based on different key variables, and different political, social or economic institutions are given more or less weight depending on the purpose of the model. Some of Esping-Andersens critics have added one or two categories (Svallfors 1997), others have used different methods or criteria to differentiate between regimes (see Gelissen 2002 for an overview). This has resulted in a general criticism of regime models (e.g. Baldwin 1996). To what extent this criticism is related to Esping Andersen s model as empirically flawed or theoretically problematic, remains more unclear. This is after all a field that to a high extent is empirically and problem driven (c.f. Pierson 2000). Comparative research concerned with justice, welfare and solidarity is today not framed in one theoretical program (c.f. Lakatos 1978), but in several research programs and approaches: -Welfare state solutions and institutions (Esping-Andersen 1990; Palme 1999; Goodin et al 1999) -Welfare state support (Gelissen 2002; Svallfors 1997) -Strike level, unionization, labor market model (Kjellberg 1983; Crouch 1993; Stokke 1998) -General value patterns, culture (Inglehart 1990, 1997) -Justice principles (Aalberg 2001, Gelissen 2002) This list is far from complete, but these examples illustrate how some topics are approached only from certain angles. For example are national differences in perceived conflict patterns rarely examined (but see Allardt 1975), compared to institutional conflict variables, e.g. unionization and class politics. The same applies to questions of how people assess opportunity structures and possibilities for getting ahead in society. These are topics that have 2

received little attention compared to for example, questions of intergenerational mobility patterns in different welfare regimes. What we can derive from these examples is the usefulness of differentiating between studies where the emphasis is on institutionalized welfare configurations and outcomes in different nations on one hand, and studies that emphasize the values and attitudes of the population in a nation on the other. Further, it will for our purpose be helpful to separate studies of the welfare state as such from studies of inequality, justice and conflict, even though the two are partly merging. In each area there are institutional studies and attitude studies. The resulting different fields of research are intertwined, and clearly build on each other, but the main focus varies. Attitude and value studies are often used as a way to test institutional regime models, and the institutional approach is an interesting and necessary background for our focus on values and attitudes. Attitude regimes Institutional traits are important in Esping Andersen s regime classification, and an institution s approach has also contributed to the explanation of cross-national variance in values and attitudes (Gundelach 1994). Both Esping-Andersen and other authors have developed the theory of how institutional welfare regimes influence public opinion. Several authors have concluded that the type of welfare regime actually does matter for people s preferred level of solidarity and inequality and for their view on the role of government in securing social rights (Arts & Gelissen 2001; Svallfors 1997). In research, as well as in politics and media, the question of changing public support for the welfare state has been a hot topic. The so-called welfare state crisis has actuated a debate on the popular foundation of the welfare state (see Kautto et al 1999; Gelissen 2002). Several authors find that public support for the welfare state is relatively stable in the Nordic countries, and that the crisis hypothesis is hard to substantiate (e.g. Andersen et al 2001). Attitudes towards the welfare state are hard to measure in a straightforward way, however. Some topics, like attitudes towards the functions of the welfare state (security and equality), and the sub-question of whether people see the government as responsible for fulfilling these functions, are better covered in the large surveys, and have therefore been more thoroughly analyzed (Andress & Heien 2001). Another well-covered topic is variation in the support for different principles of justice. The principles of equality, need and equity have been found to receive varying support in different countries (Aalberg 2001; Arts & Gelissen 2001). The results may be hard to summarize though, as patterns observed depend on the choice of items and the wording of questions (Haller 2002; Aalberg 2001). Citizens of immature welfare states are in line with Arts & Gelisen (2001) the ones who want more equality, while the people in mature, social democratic welfare states and hybrid countries like Great Britain and The Netherlands demand less equality. The general public in these countries favor leveling, but seem to be more willing to accept income inequality - some degree of inequality is legitimate if it is justified through education, skills or otherwise. Factors like the actual level of inequality and poverty, or the level of social security, are often mentioned only in a subordinate clause, recognizing their existence, but not going further into the matter. Questions of conflict and consensus are a more neglected part of the wider inequality field. Most attention has been given to the claim that new social cleavages are replacing class-based cleavages in the post-industrial society (Crompton 1993; Scott 1996; Pakulski & Waters 1996). According to Svallfors (1997), a comparative perspective has been lacking in this 3

debate. Even though cross-national differences in the aggregated levels of attitudes may be large, Svallfors (1997) found strikingly similar group differences and cleavage structures in nations belonging to different regimes. Both class and gender have substantial effect on attitudes in all the nations he studied (ibid). Several authors have tested attitude differences between groups using additional background variables, while related subjects, like national differences in conflict level and group composition, have been given less attention. There are exceptions, however, as we will come back to below. 3. The Nordic model Researchers anchored within the regime paradigm may disagree about the total number of regimes and about the placing of certain countries (c.f. Goodin et al 1999). As a Nordic model is our main focus here, we do not go into all aspects of this discussion. However, according to most of these regime models, the Nordic countries stand out as different. Not only are most European states set apart from countries like the USA and Australia in the way the middle class is included in the social security system (Palme 1999). The Nordic countries are also believed to represent a distinct welfare system within Europe (Esping-Andersen 1990). Sweden, Norway and Denmark are labeled Social Democratic or Scandinavian in most of these models, typically with Sweden as a prototype. Finland is a more contradictory case that is sometimes grouped together with other European countries, and Iceland is seldom included at all (for an overview see Gelissen 2002, p. 46-48). The Nordic countries are not similar in all respects, and Mjøset et al (1986) are for example treating them as five separate models in their study of the Nordic economic and political systems. Neither is it given that social democratic characteristics necessarily are exclusively Nordic or Scandinavian: some authors classify Austria, The Netherlands and Belgium as hybrid types of the social democratic regime, as they display some of these characteristics (Arts & Gelissen 2001). The picture of a Nordic model is therefore not without contradictions. Institutions and outcomes A Nordic model of welfare solutions and social policy has become an established concept in the social policy field (Kautto et al 1999, 2001; Palme 1999). This is usually associated with universalism, generous benefits, comparatively high labor participation rates for men and women, a dual-earner model, active labor market policies and extensive social services, to mention some key elements. The Nordic countries further have comparatively egalitarian income distributions and low levels of poverty and inequality (Fritzell 2001). The Nordic model in this sense is heavily based on traditions and institutionalized practices. There are also historical and institutional differences. Historically, Denmark and Sweden were in a Nordic context for a long time more powerful and hegemonic, the others less powerful and more peripheral. The social democratic political hegemony after World War 2 was strongest in Sweden and more unstable in Norway and Denmark. Finland has been marked by a larger number of internal conflicts and political disputes, while the situation in Sweden and partly in Norway and Denmark was more stable in this period (see Allardt 1975; Mjøset et al 1986; Gudmundson 1993; Kautto et al 1999). After 1973, the Nordic countries have followed partly diverging tracks. Denmark went through a period of recession already from the 1970s with high unemployment and budget deficits. Sweden and Finland were deeply affected by the economic crisis in the early 1990s. Unemployment rose sharply, and the governments introduced cuts in public spending on social security (Kautto et al 1999). The structural changes of the Swedish welfare state have been deeper than in Denmark and Norway (see e.g. SOU 2001:79). Denmark had already started restructuring the welfare state at this time, and Norway escaped the worst effects 4

thanks to their oil and gas revenues. The Swedish welfare state is despite of cuts and restructurings still extensive with high social expenditure, high tax levels and high coverage (Kautto et al 1999). European integration is deepening, also in the Nordic countries. Denmark has been a member of the European Union since the mid-1970s and Sweden and Finland became members in 1995. Norway is not a member of the union, but is de facto associated through the membership in the European Economic Area (EEA). With some exceptions (e.g. Mjøset et al 1986), it is commonly argued that it still makes sense to speak of a Nordic model. This is however ultimately a question of what kind of issues we are looking at. Attitudes and values On the attitude dimension, there are similarities and differences between the Nordic countries. While Allardt (1975), despite variations found a striking uniformity of Nordic values in his study from the 1970s, Halman (1994) concluded that there was no exclusively Scandinavian value pattern in orientations regarding religion, family, work and politics. Some of these values may better be described as European rather than Nordic. The ideal typical Nordic model includes images of a predomination of socialism, solidarity and egalitarianism. Equality and egalitarianism are important parts of the national self-image in the Nordic countries, but the hypothesis that people here are exceptionally egalitarian has actually been refuted by several researchers (Martinussen 1993; Gelissen 2002; Aalberg 2001). Scandinavians are moreover claimed to prefer freedom above equality, but this applies to the rest of Northern Europe as well (Halman 1994). Does it then make sense to speak of a uniquely Nordic model of justice, inequality and solidarity attitudes? In Esping-Andersen s approach (1990), it was an important point that the social democratic welfare states had managed to incorporate the new middle classes by making social rights universal. This was not accomplished in the liberal Anglo-Saxon countries. In continental European countries, occupationally segregated social insurance programs ensured the loyalty of the middle classes. Based on this, the hypothesis has been that the level of support will be higher in universal, social-democratic regimes than in corporatist and liberal (Gelissen 2002, p. 66). This conclusion, however, has not received clear support in empirical studies (Gelissen 2002; Svallfors 1997; Gundelach 1994). Welfare state support has not decreased in the Nordic countries, however, and the level is still comparatively high (Andersen et al 1999). According to Svallfors (1997), the populations in the social democratic countries both support state intervention and want an egalitarian income distribution. None of these traits are unique for the Nordic countries, but the combination of the two can be a key to further explore the Nordic model. The support for state intervention is not undisputed however, partly because the answer may depend on the level of generality of the questions asked. Aalberg (2001) finds that the stereotypical image of the egalitarian Scandinavian citizen fits in regard to support for ideals, but that Scandinavians in fact are less enthusiastic about egalitarian policies. Thus, policies to achieve an equal society are probably not as popular in these countries as one might expect. On the other hand, most research suggest that people in the Nordic countries favor small occupational differences in pay when asked what different groups ought to earn (Aalberg 2001; Svallfors 1997). Here, the match seems to be quite good between actual situation (egalitarian income distribution) and attitudes (ideal income distribution). Policies are more controversial than ideals, and policy attitudes are to a larger degree influenced by the different realities that people in different countries relate to (Aalberg 2001, p. 218). State intervention to achieve more redistribution could already have reached a satisfactory level in the Nordic countries. On the other hand, high levels of taxation and transfers do not seem to alter the ideals of economic equality as an ultimate goal. 5

The relations between ideals of equality, welfare state support and values like individualism, liberalism or socialism are complex. According to Martinussen (1993, p. 49): the Norwegian population has simultaneously become more individualistic and more in favor of welfare-state arrangements. At present egalitarianism and welfare-state support is only weakly associated. This apparently contradicts the common understanding that socialist governments and an egalitarian culture have brought about the strong support for redistribution and transfers in Norway and Scandinavia, and therefore explain the growth of the welfare state (Martinussen 1993). A preliminary conclusion must be that the Nordic passion for equality (Graubard 1986) is not as uncomplicated as some of the stereotypical images will have it. 4. A closer look at inequality, conflict and social cleavages The Nordic model appears as less clear-cut in some of the attitude and value questions than in the social policy area. Still, most authors agree that people in the Nordic countries are among the most egalitarian when asked about how big occupational wage differences ought to be. Other aspects of a Nordic attitude model are far more complex, and here we wish to look into some of the less developed questions, namely those of inequality, conflict and consensus. In a 1972 Nordic survey respondents were asked to evaluate the strength of conflicts between different groups based on among other factors class, age, religion, and education (c.f Allardt 1975). Oppositions based on class cleavages turned out to be among the conflicts most often perceived as strong in all the Nordic countries. Quite many also evaluated the conflicts between people and politicians to be strong. Allardt (1975) interprets this as an indication that the Nordic countries are generally politicized meaning that people here interpret society in terms of conflict and justice. The opposite of politicized is termed privatized by Allardt (1975). Within the Nordic countries, he classifies Denmark and Norway as more privatized societies than Sweden and Finland. Norwegians and Danes experience conflicts more seldom than the Swedes and Fins. According to Allardt (1975), this is an indication that Sweden and Finland in the 1970s could be defined as more politicized than its neighbors. Class consciousness and interest conflicts are interpreted as much closer to the surface here than in Norway and Denmark, and people more often experience the relations between different groups as characterized by conflict. This could perhaps be translated into a division between societies with an individualistic or collectivistic inclination. Sweden in the 1970s would in this case be classified as an example of a more collectivistic society than Denmark and Norway, which were more individualistic and less conflict oriented societies. Allardt described an era in the Nordic countries marked by growth (Mjøset et al 1986). This was the heyday of fordism, of welfare state expansion and social democracy. The 1970s became a turning point, often dated to the oil crisis in 1973. The line between privatized and politicized, or individualist and collectivist might have changed since then. It is a recurring claim that modern societies are becoming more individualistic (Giddens 1991; Castells 1998; Bauman 2000). A fresh light on conflict perception and the development of conflict level in the Nordic societies will add interesting knowledge to this debate. Conflict levels can be studied in different ways. Andersen et al (1999) studied conflicts in terms of the level of disagreement between groups about welfare state arrangements, and found that disagreement was most common in Sweden, followed by Finland and Denmark, while Norwegians displayed the highest consensus on these issues. They claim that class related factors generally have the greatest impact in Sweden, and that the Norwegian pattern is the most consensual in the Nordic countries (Andersen et al 1999, p. 254; Knutsen 1998). Our approach is more in line with Allardt s (1975) focus on people s perceptions of social 6

conflicts. Questions partly corresponding to the ones he used in the 1972 survey were asked in the ISSP 1999 Social Inequality module, but this material has not been thoroughly researched. Whether society is perceived as fair or not, and people s views on their own position and on social mobility and awards, are very interesting questions indeed. In line with both Allardt (1975) and Andersen et al (1999) we can expect some variation within the Nordic model in this question. There is also reason to believe that the relatively equal distributions in these countries will be reflected in people s images of a fair society and their assessment of conflicting interests. 5. Nordic inequality values in a comparative perspective The term inequality is less frequently used than equality in all of the Nordic countries today. Equality is generally perceived as an integral part of the ideology and goals of the welfare state supporting social reforms and social engineering (integration perspective). Social inequality, on the other hand, is to a higher extent associated with conflicts and an inherent critique of the economic and political system (conflict perspective). It is however not self evident how and to what extent equality is ideologically coded and perceived in the Nordic countries and elsewhere. To be able to address these questions empirically it is necessary to distinguish between different aspects of inequality. Data and methods We will in this paper make use of data from the International Social Survey Program (ISSP; www.issp.org). The ISSP is an annual and cross-national program inaugurated in the mid- 1980s, which now involves more than 30 nations. We build our analysis on the 1999 replication of the Social Inequality module, but have restricted the analysis to eight countries. Of the Nordic countries, unfortunately only Denmark, Sweden and Norway participated in the 1999 survey. Building on regime theory, we have further chosen Great Britain, the USA and Canada as representatives of the liberal regime type, while Germany and Spain represent the conservative regime type. The last two nations are certainly different in many respects, and Spain is e.g. classified as a Southern country by Svallfors (1997). Since our main aim is to explore the Nordic model, such differentiations will be of secondary importance to us. In any case, we do not intend to generalize within these models without caution. The choice of nations is also motivated by practical reasons. France was a clear candidate, but had to be excluded due to an exceptionally low response rate of only 17%. The response rates have been a cause for worry in several cases, and we have therefore limited our analysis to aggregated measures, not splitting up the results for different groups within the nations. 1 This is not as accentuated in the Scandinavian countries, but with an eye to possible methodological problems of this kind, we will interpret our findings with caution. On the other hand, the ISSP is today one of the best sources for quantitative and comparative data on attitudes and values. The empirical results will be presented using mainly percentages, means and standard deviations. In all kinds of comparative research, special caution is demanded (e.g Jowell 1998), not less so when analyzing the ways institutional regime types are related to patterns of values and attitudes. Values have different levels of generality, and it seems that the lower the level, the more controversies can be detected (Haller 2002). Question wording can therefore make large 1 1. Sample sizes and response rates: Sweden: N=1150 (61,2%), Norway: N= 1268, 51,7%, Denmark: N=1823 (59,0%), Spain: N=1211 (98,5%), Great Britain: N=804 (44,6%), USA: N=1272 (32,3%), Germany: N=1432 (22,4%) and Canada: N=984 (29,4%). The actual response rate for Germany is probably higher, as this module was done in split with the 2000 module. 7

impact, and abstract questions might generate other results than more specific questions (Aalberg 2001). Despite obvious difficulties, attitude comparisons can tell us important things about the similarities and differences between advanced welfare states that are hard to illuminate through institutional analysis alone. It is not our main aim to perform a test of the regime theory as such, but our results will hopefully contribute to the understanding of variation both within and between regime types, especially related to the coherence of the Nordic or Scandinavian model. The 1992 replication of the Social Inequality module includes a number of the same questions we analyse here, but not for all eight countries in our study. For this reason, and because our main aim is to give an updated condition report, we have restricted our analysis to the 1999 survey, and will only in a few instances refer to changes. The ISSP 1999 survey includes a multitude of questions measuring people s perceptions, attitudes and values related to social inequality. For our purpose, the survey deals with factors that affect perceptions of life chances, explanations for inequality and people s assessments of the strength of different kinds of social conflicts. Here, we present three main dimensions that are of special interest: 1) Perceptions of opportunity structures: how family and social network are assessed as factors affecting life chances. 2) Inequality as ideology: how inequality is explained. 3) Inequality in terms of conflict patterns: assessments of conflicts between different classes or interest groups. In order to deepen the understanding of the ways conflict and consensus are perceived in different nations, we also explore questions of social status and of people s perceptions and ideals when it comes to the social structure and stratification pattern in the country they live in. This is of interest because it may reveal some differences in the way people see their societies as more or less egalitarian in different nations or regimes. Inequality and perceptions of opportunity structures Social mobility is clearly a phenomenon related to inequality, in the sense that people regard opportunity structures as more or less dependent on social background, capital or status, and therefore more or less equal. The crucial importance of family origin and background is well documented in comparative mobility research suggesting surprisingly similar mobility patterns between father and sons over time and across different industrialized societies and welfare regimes (Erikson & Goldthorpe 1992; Marshall et al 1997). The number of questions about perceptions of opportunity structures in society is unfortunately somewhat limited in the ISSP 1999 material. The available questions are however of great interest as they highlight the relative importance of family, social network and corruption for the opportunities of the individual. People are asked how important each of these factors is for getting ahead in society. The data presented in table I below, suggests that most citizens in highly developed market societies actually believe that coming from a rich family is not crucial for people s life chances today. This belief is not unique for the Scandinavian countries. It is widely accepted both in social democratic welfare regimes (particularly in Norway and Denmark) and in liberal welfare regimes (Canada, the USA and the Great Britain), but less often supported in conservative welfare regimes (Spain and to a lesser extent Germany). Spain is interesting as a case where the public more often highlight the importance of family capital as opposed to more meritocratic principles gained via the educational systems or work experience. The second item in table I refers to the perceived importance of social capital (knowing the right people). This factor is generally given much more weight than the family in all of the countries reported, but we see the same pattern where the importance of social networks is less often stressed in the social democratic and liberal regimes and more often highlighted in the conservative regime. Interestingly, if not 8

surprisingly, the same pattern also emerges if we look at the perception of corruption for getting ahead in society. Corruption is, as least when it comes to the public, widely believed to play a more crucial role for people s life chances and opportunity structures in conservative welfare regimes than in social democratic and liberal regimes, once again underscoring the importance of family, network and informal contacts for getting ahead. Table I. Opportunity structures in eight countries. Percentage who agree, means and standard deviations SW NO DK GE SP GB USA CA 1) Coming Agree 19,0 10,8 9,4 26,6 54,4 19,2 20,5 14,8 from a wealthy family is Mean 3,26 3,48 3,66 3,12 2,61 3,45 3,43 3,51 important for getting ahead St.dev. 0,94 0,85 0,86 0,99 0,98 1,06 1,04 1,00 2) Knowing Agree 56,5 33,3 42,3 60,9 66,1 41,7 47,2 51,0 the right people is important for Mean 2,45 2,83 2,67 2,33 2,27 2,66 2,58 2,52 getting ahead St.dev. 0,86 0,85 0,90 0,88 0,81 0,95 0,97 0,86 3) To get to the top, you have to be corrupt Agree 20,7 10,8 15,0 42,9 35,1 19,0 17,1 17,6 Mean 3,40 3,86 4,10 2,98 3,15 3,46 3,61 3,59 St.dev. 1,09 1,01 1,24 1,22 1,13 1,05 1,10 1,07 Source: ISSP 1999 Inequality III. Notes: Question wording: For getting ahead in life 1) How important is coming from a wealthy family? and 2) Knowing the right people? And 3): Do you agree or disagree with the following statement: To get to the top in <respondent s country> today, you have to be corrupt. Values on 1) and 2) are 1=Essential, 2= Very important, 3=Fairly important, 4=Not very important and 5=Not important at all. Entries are percentages that answer essential very important. Values on 3) are 1=Strongly agree, 2=Agree, 3=Neither agree nor disagree, 4=Disagree and 5=Strongly disagree. Entries are percentages that strongly agree or agree. Can t choose is coded as missing on both items. Mean values close to one indicate that many respondents answer affirmative. The results are with one crucial exception, Sweden, supporting the overall logic of Esping- Andersen s regime classification. Citizens in social democratic and liberal regimes are more often expressing a trust in meritocratic principles as opposed to people in conservative regimes, where the importance of family background and social network is more often emphasised. Contrary to what might be expected, Sweden is scoring comparatively high when it comes to family and social network as institutional mediators of social inequality. This may, at least partly, be explained by the recession in the early 1990 s, but also by a sense of growing class differences during the recovery period after 1995, benefiting well connected and powerful groups at the expense of ordinary people. The status as one of the most developed welfare states and economies in Northern Europe is in 1999 no longer self evident or true, generating a growing discontent with the invisible and informal opportunity structures and growing inequalities (SOU 2001:79; Nilsson 2000). 9

A first look at the Scandinavian countries in the ISSP 1999 survey suggests in sum that citizens in Norway and Denmark more often are putting trust in to meritocratic principles of stratification than in Sweden. People in Norway and Denmark seem to believe in more egalitarian opportunity structures than people in the remaining countries, including Sweden. At the same time, we also find support for a general hypothesis that the functions of family and network are more highlighted in conservative welfare regimes than social democratic and liberal. To what extent this is a good explanation for how social inequality is actually reproduced in different welfare regimes (c.f. the above references to intergenerational mobility research) is another matter. It seems reasonable to assume that family and social network in the Scandinavian countries, and elsewhere, are more significant as mediators of social inequality and life chances than both the public and many sociologists tend to believe. Ideologies of inequality Economic inequality is found in all countries, and the explanations people give for the existence of economic inequality informs us about the way differences between people are legitimized. In the 1999 ISSP survey, respondents are asked whether they agree or disagree with different statements about social inequality. A majority of citizens in all of the eight countries included in our investigation fully agree or agree that social inequality continues to exist because it is in the interest of the rich and powerful. The results presented in table II suggests a picture where this statement is most often referred to in the conservative welfare regimes (an average of 76% agree that inequality exists because it benefits the rich), and less often in the liberal regime (56%). The Scandinavian countries are in this respect placing themselves closer to the liberal regimes than the conservative, but there are however, pronounced differences between the Scandinavian cases. More than two thirds of the Norwegians (71%) agree that inequality exists because it benefits the rich, as opposed to 55% of the Danes and 61% of the Swedes. Norwegians are in this respect closer to the conservative countries than to their Scandinavian neighbours. Contrary to what might be expected do we not find strong public support for market liberal ideas, at least not as it is operationalized it in question 3 in table II. The average support in the three welfare regimes for the statement that large income differences are necessary for prosperity is with the exception of Denmark surprisingly similar (2-3 out of 10 on average support the idea that inequality is necessary for prosperity). Once again we find substantial differences between the Scandinavian countries: Danes show the strongest support for the idea that income differences are necessary (37%), while only 16% of Norwegians and 21% of the Swedes agree with this statement. Denmark and Norway are actually representing the extremes, showing the highest and the lowest support, respectively, for market liberal ideas among these eight countries. Question 2 taps the same issue in a less abstract way. The statement that no one would study for years unless they expect to earn a lot, generally receives much more support in all nations reported than the general statement that income differences are necessary for prosperity. A majority in all nations agree, but the support is particularly large among Germans. Norwegians show less support than Swedes and Danes also on this item, confirming the impression that market liberal ideas have less impact here. The difference between Sweden and Denmark is negligible, and people in the USA actually least often agree with this statement. 10

Table II. Ideologies of inequality in eight countries. Percentages that strongly agree or agree, means and standard deviations SW NO DK GE SP GB USA CA 1. Inequality benefits the rich and powerful Agree 61,1 72,4 57,7 77,1 79,7 61,9 50,1 62,0 Mean 2,42 2,29 2,60 2,14 2,09 2,41 2,64 2,43 St.dev 1,03 0,93 1,41 0,92 0,85 0,92 1,01 1,07 2. No one Agree 71,1 64,4 73,1 89,7 72,9 69,2 62,0 64,5 would study for years unless paid more Mean St.dev 2,24 0,95 2,46 1,01 2,17 1,30 1,75 0,78 2,16 0,92 2,39 0,97 2,43 1,14 2,38 1,12 3. Large Agree 20,8 16,8 37,2 25,4 27,9 19,5 27,4 18,8 differences in income necessary for prosperity Mean St.dev 3,41 1,05 3,50 0,94 3,20 1,37 3,32 1,08 3,33 1,12 3,45 0,95 3,19 1,04 3,57 1,02 Source: ISSP 1999 Social Inequality III. Note: Respondents are asked if they agree or disagree with the following statements: 1) Inequality continues to exist because it benefits the rich and powerful, 2) No one would study for years to become a lawyer or doctor unless they expected to earn a lot more than ordinary workers, and 3) Large differences in income are necessary for <respondents country> s prosperity. Entries under agree are percentages that answer 1= strongly agree or 2= agree. Values go from 1=Strongly agree to 5=strongly disagree. Mean scores close to one indicate that many respondents agree with the statement. Based on the ISSP data we may conclude that the general public in the Scandinavian countries believe that social inequality exists because it benefits the rich and powerful, and that inequality is not a priori a premise for increased material prosperity. These are ideas in line with the official social policy in the post-war period and contradicting a political discourse where market ideology in many cases is taken for granted in politics, in social policy and in the media. The Norwegian public, and partly the Swedish, are more often expressing a critique against the market liberal ideology of increasing inequality compared to the Danes. 2 An interpretation of this is that citizens in Denmark tend to legitimize inequality differently than people in the other Scandinavian countries. The data presented in table II is not lending full support to the idea that Scandinavians are expressing a distinct ideology of inequality. On the contrary, we find substantial differences between Norwegians on one hand (followed by the Swedes) and more liberal minded and market oriented Danes on the other hand. 2 The Danes are also less inclined (46%) than the Norwegians (61%) and the Swedes (59, 5%) to support that their tax pay is used by the government to reduce social inequality (c.f. the 1999 ISSP Study). The ISSP data suggests a general sentiment where the Scandinavian welfare states still have a solid public mandate to actively prevent social inequality. The scepticism expressed by the Danes, in terms of how much the welfare state in the future can or should do to conquer social inequality, is less often expressed in Norway and Sweden. If we compare the 1999 ISSP data with ISSP data for Norway and Sweden from 1992 we find overlapping support for the active welfare state in Norway in 1999 (61% in 1999 and 60% in 1992) and an increased support in Sweden in 1999 (59,5%) compared to 1992 (53%). 11

Social cleavages and conflicts Following Esping-Andersen (1990, 1999) and a whole industry of research on income distribution in the OECD area it is reasonable to assume that the Scandinavian cases are more equal in terms of the distribution of economic resources than most other countries (Fritzell 2001; Smeeding 2002). In line with this argument, and consistent with long traditions of heavily institutionalized and organized collaboration between labour and capital (Mjøset et al 1986; Stokke 1998), we would also expect more consensus and less articulated cleavages between interest groups than in most other welfare regimes. In table III we present a number of possible cleavages between poor and rich, the working class and the middle class, management and workers, top and bottom of society and between generations. In table IV we present two indexes, one summing up the average scores on the five areas of potential social cleavages (b), and one summing up the two class related items (a). Only four out of five areas are included in Canada. As a consequence of this Canada is unfortunately not included in the two indexes in table IV. Table III. Perceptions of conflict and social cleavages in eight countries. Percentages that regards conflicts as very strong or strong, means and standard deviations Conflicts are SW NO DK GE SP GB USA CA* very strong or strong between: 1. Poor and Agree 35,0 15,2 13,8 41,7 32,2 49,0 57,8 34,7 rich Mean 1,37 1,08 0,98 1,40 1,08 1,52 1,69 1,35 St.dev 0,65 0,56 0,65 0,77 0,87 0,69 0,75 0,69 2. Middle Agree 12,6 5,3 5,1 15,8 22,1 26,7 28,0 14,2 class and Mean 1,00 0,81 0,69 0,93 0,88 1,19 1,24 0,97 working class St.dev 0,55 0,54 0,59 0,68 0,82 0,65 0,70 0,60 3. Managers Agree 33,2 20,6 16,5 46,5 39,7 48,9 55,0 -- and workers Mean 1,34 1,18 1,06 1,47 1,27 1,51 1,65 -- St.dev 0,60 0,53 0,60 0,71 0,87 0,68 0,74 -- 4. Top and Agree 64,6 38,1 33,7 46,6 28,6 63,0 72,0 52,1 bottom in Mean 1,82 1,41 1,30 1,48 0,99 1,77 1,99 1,66 society St.dev 0,78 0,70 0,81 0,81 0,87 0,77 0,82 0,84 5. Young Agree 19,5 14,6 13,6 32,3 20,1 34,8 43,1 27,1 and older Mean 1,08 1,05 0,88 1,22 0,86 1,29 1,47 1,17 St.dev 0,65 0,55 0,69 0,75 0,80 0,73 0,78 0,74 Source: ISSP 1999 Social Inequality III. Notes: Question wording is: In all countries, there are differences or even conflicts between social groups. In your opinion, in <respondent s country> how much conflict is there between 1) Poor people and rich people, 2) The working class and the middle class, 3) Management and workers, 4) People at the top of society and people at the bottom, and 4) Young people and older people. Values are recoded, so that 0=There are no conflicts, 1=Not very strong conflicts, 2=Strong conflicts and 3=Very strong conflicts. Entries under agree are percentages that have answered Very strong or Strong. Can t choose are coded as missing. The higher the mean value, the more people see the type of conflict as strong. *The question about managers and workers is not asked in the Canadian questionnaire. A first and superficial reading of table III confirms a picture where citizens in the social democratic welfare regime less often experience articulate interest conflicts. With the exception of generational cleavages, we find an interesting distinction between the nations belonging to the social democratic regime on the one hand, and the more conflict ridden 12

nations in the liberal welfare regime (Great Britain, the USA and Canada) on the other hand. The conservative welfare regimes (Germany and Spain) are locating themselves in between the social democratic and liberal welfare regimes when it comes to the cleavages between poor and rich, top and bottom, young and older and the working and middle classes, but closer to the liberal regimes as far as conflicts between management and workers are concerned. Contrary to what is commonly claimed by German sociologists (c.f. Offe 1984, Beck 1992, 2001; Luhmann 1996) and British sociologists (Saunders 1991; Lash & Urry 1987), we still find a strong perception of class conflicts rooted in production in both Germany and Great Britain. The percentages highlighting conflicts between management and workers are substantially higher in both of these countries compared to any of the Scandinavian cases. Great Britain is also scoring comparatively high on the class conflict index presented in table IV. A closer examination of the figures presented in tables III-IV reveals several interesting findings. Table IV. Conflict indexes for seven countries*. Means and standard deviations Country a) Class conflict index (2 items) b) Total conflict index (5 items) Mean St.dev. Mean St.dev. SW 2,34 0,96 6,60 2,29 NO 1,98 0,89 5,52 2,08 DK 1,74 0,98 4,88 2,40 GE 2,39 1,21 6,49 2,89 SP 2,15 1,52 5,10 3,51 GB 2,70 1,14 7,31 2,51 US 2,89 1,20 8,08 2,67 Source: ISSP 1999 Inequality III. Notes: Question wording, see table 3. The conflict indexes are constructed in the following way: a) Class conflict index = working-middle class + management-workers. 2 items, recoded as in table 3. Values range from 0-6, 6 being the highest score, indicating a strong conflict perception. b) General conflict index= poor-rich + working-middle + management-workers + top-bottom +young-old. 5 items, recoded as in table 3. Values range from 0-15. * Index values are not computed for Canada, as one of the conflict items is missing. First of all, we notice that Sweden by no mean is synonymous with the average of the Scandinavian countries. On the contrary, Sweden at the end of the 1990 s represents a more conflict driven society than both Norway and Denmark. Swedes more often express conflicts and cleavages in society between poor and rich, workers and middle class, management and workers and top and bottom of society, the only exception being generational conflicts. We have corresponding data from 1992 for Norway and Sweden, but not for Denmark. The Norwegian picture is more or less stable from 1992 to 1999. The Swedes are both in 1992 and in 1999 experiencing social cleavages more often than the Norwegians. The 1999 figures for Sweden are significantly higher than in 1992, but the differences can not be interpreted as supporting a scenario suggesting a break-down of the Swedish welfare model. 3 Given the relative stability in both Norway and Sweden during the 1990 s, we find it reasonable to interpret these results as systemic rather than as expressions of temporary discontent with the welfare state or the overall economic development. This interpretation is in line with a general 3 The 1992 ISSP percentages for Sweden expressing substantial conflicts are: a) poor - rich (30%), b) working class - middle class (8,5%), c)management workers (28%) and d) young - older (14,5%). The 1992 figures for Norway are: a) 18%, b) 5%, c) 23% and d) 15%. 13