Understanding of democracy in unified Germany. Is there convergence of democratic values between east and west through institutional learning?

Transcription

1 Understanding of democracy in unified Germany. Is there convergence of democratic values between east and west through institutional learning? Benjamin C. Sack Department of Political Science University of Mainz Paper prepared for the 5th ECPR Graduate Student Conference University of Innsbruck, Innsbruck 3-5 July 2014 Draft - June 2014 Abstract The stability of democratic political systems depends on the congruence of the actual institutional design and the normative model of democracy favoured by the public. In the late 1990s, research on political culture in Germany had shown that after 40 years of separation people were divided by different understandings of democracy. Given that political culture is primarily determined by political socialisation, I use the natural experiment of german reunification to analyze the democratic attitudes of the first generation completely socialised in unified Germany. Empirically I draw from the sixth round of the ESS covered in 2012, which provides a wide range of questions concerning the understanding of democracy. The results support the thesis that differences between people from East and West Germany are considerably smaller within the youngest generation. Surprisingly the convergence of attitudes is bidirectional: young Germans combine positions to a new unified understanding of democracy: while there is a convergence in the expected direction in a socialist model of democratic understanding, convergence in a liberal model runs against theoretical expectations. Introduction One of the major findings of political culture research in unified Germany was presented by Fuchs (1999). He showed, few years after german reunification, that the normative models of democracy prefered by the public of East 1

2 and West Germany differed tremendously. While principle and generalized support for democracy was rather high in both parts, different considerations existed about what the basic characteristics of a democracy should be. While in West Germany a model of liberal democracy was favoured, which was in line with the institutionally implemented structure, a model of socialist democracy was mostly prefered in East Germany (Fuchs, 1997a, 1999). Starting from the central postulate of political culture theory (Almond and Verba, 1963) this finding represents a large challenge. Because an ongoing incongruity of the prefered model of democracy by the public and the institutionalized model of democracy on the structure level can lead to a growing instability of the political system. Given the assumption that value orientations play a central role in shaping political attitudes and behaviour it is not surprising that German social scientists still find large differencens between East and West Germans in these attitudes and behaviour. Recent research brought up differences in the satisfaction with democracy (Niedermayer, 2009), voting behaviour (Rohrschneider et al., 2013), orientations to the political community (Westle, 2013), political trust (Campbell, 2011) and attitudes towards the welfare state (Arzheimer, 2013; Roller, 2014). These differences between the East and West-German population is mainly explained by underlying differences in values and value orientations towards democracy and thus in different understandings of democracy. 1 Besides these findings there is also additional and therefore contrary evidence for an increasing convergence of East and West Germans in their political value orientations. Studies investigating the differences between generations find lower support for socialism in the younger East generations compared to the older East Germans (Fuchs and Roller, 2013). Not only against this background, it appears relevant and necessary to ask whether 1 For the conceptualization and concept specification of "normative understanding of democracy" and "democratic value orientations" see the next section. 2

3 the yielded findings by Fuchs (1999) still hold, or whether there is a convergence of generations in the understanding of democracy happening and thus the "wall in the heads" is slowly degraded by a common socialisation in unified Germany and a related institutional learning (Rohrschneider, 1999). In short, do democratic value orientations converge in the youngest generation, which was the first one socialised in a reunited Germany? The relevance for investigate this question is given by the before expressed basic assumptions of political culture theory in the first place. Secondly, it must come as a surprise that such a serious finding was not reexamined in the past for almost twenty years. One possible reason for this, however, could be the lack of available data sources. The latest wave of the European Social Survey (2014) provides for the first time a module (European Social Survey, 2013), which measures the concept of democracy in a comprehensive manner. This data set allows for the first time to set a cohort classification, in which a) the adoption of a pan-german socialised generation is plausible and b) offers approximately the same amount of number of cases within the separate generations and high enough number of cases which is needed for inferential analyzes. The paper is organized as follows. First, the theoretical framework is described. Here a necessary conceptualization of the terms "understanding of democracy" and "democratic value orientations" is made. In addition, the theoretical model of institutional learning is presented. This will be followed by a brief overview of the state of research and based on that, as well as on the theoretical model, hypotheses are derived which will guide the subsequent empirical analysis. That includes the representation of the data base and the operationalization of the variables used. The results are then presented and discussed. A conclusion summarizes the results and gives an outlook. 3

4 Clarifying concepts: understanding of democracy and value orientations towards democracy The central dependent variable in the following analysis will be the understanding of democracy of the German citizens. This is understood as a pattern of different value orientations towards democracy. Conceptualized as such it is defined as the normative model of democracy favoured by the public. More generally we understand values as "conceptions of the desirable" (Kluckhohn, 1951) and value orientations as attitudes towards values. By this conceptualization one major difference between values and value orientations is, that the latter can be measured just like attitudes (van Deth and Scarbrough, 1995, 25-28). We distinguish between societal, political and democratic value orientations. While societal value orientations represent conceptions of an ideal society (Arzheimer, 2005) and political value orientations represent conceptions of how a political system should be shaped (Fuchs and Rohrschneider, 2001) democratic value orientations represent conceptions about the desirable type of a democratic system. These could be seen as equivalent to a normative understanding of democracy (Fuchs, 1999). Thus, essential for the individual understanding of democracy are both, number and shape of the normative principles which are linked to the term democracy. Defined and conceptualized as democratic value orientations the normative understanding of democracy plays a crucial role in political culture theory (Almond and Verba, 1963; Almond, 1980) and in the concept of political support (Easton, 1965, 1975). Fuchs (1999, 2007) combines these theories to a model of democracy divided in hierarchically levels. In this model a (democratic) political system is differentiated into three levels: 1) the value level or level of political culture, 2) the level of political structure and 3) 4

5 the level of performance and political processes (Fuchs and Klingemann, 2002, 2006). When it comes to stability and persistence of a democratic political system the congruence of the internalised values of the public on the one hand, and the structure of the political system on the other hand is primarily important. Discrepancies of those two levels can lead to serious losses of legitimacy. To put things easy: when the public supports a type of democracy which is not implemented in the institutional design in which they live, the public will detract support for this implemented structure in the long run. From an individual perspective, values and value orientations play a crucial role when it comes to stability and persistence of a democratic system. This is also found in the definition of diffuse support given by Easton (1975, 444): "The briefest way of describing the primary meaning of diffuse support is to say that it refers to evaluations of what an object is or represents - to the general meaning it has for a person - not of what it does". For differentiate diverse normative models of democracy we refer to the conceptualization provided by Fuchs (1999) and Fuchs and Klingemann (2002, 2006). They differentiate four theoretic models of democracy: the liberal, the socialist, the republican and the libertarian model of democracy. The comparison of the liberal and the socialist model builds the focal point of the following analysis. Basically, the difference between the two models can be simplified to the conflict of two basic democratic values: freedom and equality (Thomassen, 1995, 2007). More concrete, Fuchs (1999) sees the main difference between the two models in the emphasis of the organisation principle of relationships between individuals. In the liberal model this principle is competition while in the socialist model it is solidarity. Furthermore in the socialist model we find a widespread welfare state while in the liberal model the welfare state is limited to the preparation of primary goods. Moreover in the liberal model individual freedom has the primacy 5

6 over distributive justice. Other constitutive differences are found in the implemented functions of the socialist model: basic social rights, broad and direct citizen participation, a strong notion of equality, a regulated market and public property. Within political culture theory the development of such a normative model of democracy, or as we conceptualized it, democratic value orientations is seen as the result of the process of socialisation (Almond and Verba, 1963, 14; Almond, 1980, 29; Inglehart, 1988, 1205). One theoretical model arguing in this line is expressed and presented in the next step. The model of institutional learning To explain how value orientations are internalised models of socialisation are most striking. As a convincing and widely established model (Thomassen, 2007, 430) the model of "institutional learning" (Rohrschneider, 1999) can be counted. As in other models (Dalton, 1994; Roller, 1994) in this one there is a division of two modes of institutional learning: system-internal learning and system-external learning. It is assumed that the most important and basic value orientations are built through system-internal learning. The main assumption is, that value orientations towards democracy and civic virtues are transmitted through the institutional design of the regime one lives in. For this causal relationship it is argued as follows: only if there is the possibility to practise specific behaviour which is connected to specific values the fundamental values can be internalised. For democratic values to be transmitted only democratic regimes provide these possibilities (Rohrschneider, 1999, 2; 18-19). Beside this model of system-internal learning the model of system-external learning was proposed to explain the high values of support for basic democratic principles shortly after german reunification (Fuchs et al., 1997). System-external learning means the diffussion of values from one (demo- 6

7 cratic) political system to another (non-democratic) political system (Roller, 1994). The assumption is, that system-internal learning does only occur if the revision of actual internalised (non-democratic) values does not require too much effort. For example, democratic rights which are compatible with socialistic values could be learned easily through diffusion mechanisms like the consumption of information from a political system which is not the own (through TV-programs for example (Dalton, 1994)). On the contrary, ideals like market economy are not learned easily because they are not compatible with prior internalised socialistic values (Rohrschneider, 1999, 24-25). One implication of the model of institutional learning and the most profound one for our purpose is, that changes in the normative model of democracy which is prefered by the public can only occur through generational replacement (Rohrschneider, 1999, 20). Thus, individuals born after a regime change should " be more receptive to the citizenship-qualities of a new political process than individuals whose political views are shaped by a prior regime" (Rohrschneider, 1999, 21). But it is not impossible that value orientations could change over the individual life cycle. Mainly situative factors, like the economic performance of a recently changed (and so new) political regime can contribute to a rising acceptance and support of the basic values of that new regime. West Germany represents a good example for such a development. Continuous experience with a prosperous market economy led to a rising acceptance of a democratic structure in the first place, which in turn led to a rising acceptance of democratic values because these were seen as the main cause for the positive economic development (Fuchs, 1999; Conradt, 1974). We will assume that value orientations show high stability and that they are not changed through one-off experiences but through longterm enduring experiences. So, models, which explain change of value orientations through processes of socialisation will be the main theoretical resource of explanation 7

8 in the following considerations about the understanding of democracy in unified Germany. Development of democratic value orientations in East and West Germany Since reunification the research of political attitudes and political behaviour in both Germanies have gained a large interest. Therefore findings are highly heterogenous. While in many attitudes there still exist tremendous differences between East and West Germans, in some attitudes there is convergence observed. Recently, differences where found concerning collective identity, the emotional connection with the Federal Republic and the general ideological orientation (Roßteutscher and Scherer, 2013). While there was some kind of backlash observed with a growing connection of East Germans with the former GDR (Neller, 2006) this development seems to change to more convergence and a growing connection with the Federal Republic (Westle, 2013). More specific value orientations as well as social and economic attitudes are still different in the two Germanies. But here, too, partial convergence is observed (Arzheimer, 2005, 2013; Roller, 1994, 2012). Davidov and Siegers (2010) found no difference between East and West Germans in general value orientations measured by the Schwartz Value Inventory. But differences in participation and especially in voting behaviour are striking. The former SED party, now called "Die Linke" (The Left Party) is getting nearly 4 or 5 times more votes in the East than in the West. These differences are partly explained by differences in political value orientations (Arzheimer and Rudi, 2007; Campbell, 2011; Rohrschneider et al., 2013). To sum up, when it comes to political value orientations there are meaningful East-West differences, but these differences seem to dissapper on and on. 8

9 Studies of political support in unified Germany show two stable results over several points of time: First, support for democracy on the value level is generally high in both parts, with significant lower values in the east, and second, both for support on the structure level and for support on the performance level considerable high differences exist between East and West Germans (Arzheimer and Klein, 2000; Campbell, 2012; Fuchs, 1997a,b, 1999; Fuchs and Roller, 2006a, 2008, 2013; Fuchs et al., 1997; Gabriel, 1999, 2000; Niedermayer, 2009; Westle and Niedermayer, 2009; Völkl, 2005). Most of these differences, especially those in political support and political behaviour are explained by the impact of democratic value orientations in both Germanies (Fuchs, 1999; Rohrschneider, 1999; Westle, 1994). While in the West an understanding of democracy near to the liberal model of democracy dominates, in the East the dominating understanding of democracy is one that corresponds to the socialist model of democracy. These differences in democratic value orientations are attributed mainly to a varying socialisation in the former West and the former East. In international comparative studies this result was confirmed for the West-, Central- and Eastern-European states (Neundorf, 2010; Fuchs and Roller, 2006b; Pop- Eleches and Tucker, 2014; Voicu and Peral, 2013): According to them, differences in political value orientations between the western and the former socialist states in a great part can be explained by diverging socialisation experiences in these states. With these results in mind we still expect significant differences of democratic value orientations between the West and East German public (H 1 ). More concretly in the East we expect an understanding of democracy that corresponds to the socialist model of democracy while in the West we expect an understanding of democracy that corresponds to the liberal model of democracy (H 2 ). Based on the assumptions of the model of institutional learning we expect that these postulated differences could not be found in 9

10 the youngest generation of the German public - the generation first born, raised and so socialised in unified Germany. Technically we expect convergence of democratic value orientations in the youngest cohort in direction from the East to the West (H 3 ). Research design, used data and measures The following analysis is based on data of the sixth wave of the European Social Survey (2014). This survey was conducted mainly in 2012 and contains a large battery of items concerning the understanding of democracy (European Social Survey, 2013). For an overview see Table 1. With these items, attitudes and value orientations towards various specific characterics of normative principles of democracy are measured. It is assumed that, whith every single item the acceptance of a characteristic of a normative principle is measured which could be related to democracy. These characteristics could be assigned to "normative principles" (Fuchs, 1999) or "dimensions" European Social Survey (2013) which in turn can be used to describe different normative models of democracy (Fuchs, 1999, 2007; Fuchs and Klingemann, 2002, 2006). To simplify the following analysis and the analytic model the used items are aggregated to such normative models of democracy. For this purpose exploratory factor analysis has been used. The results of the factor analysis basically leave us with two factors. The first one is represented through the following items: 1) Free and fair elections; 2) Viable opposition; 3) Reliable information provision; 4) Freedom of press and 5) Protection of minority rights. Both items which represent some kind of a theoretical "rule of law" dimension (Equality of the judicial system and Horizontal accountability) also have high loadings on this factor, but 10

11 those are weaker than the ones of the before mentioned items. 2 Table 1: Chosen items from the European Social Survey (2014) to measure different understandings of democracy Characteristic Item: Using this card, please tell me how important you think it is for democracy in general Free and Fair elections... that national elections are free and fair? 2. Viable opposition... that opposition parties are free to criticise the government? 3. Reliable information provision...that the media provide citizens with reliable information to judge the government? 4. Freedom of press... that the media are free to criticise the government? 5. Protection of minority rights... that the rights of minority groups are protected? 6. Equality of the judicial system... that the courts treat everyone the same? 7. Horizontal accountability... that the courts are able to stop the government acting beyond its authority? 8. Social equality... that the government takes measures to reduce differences in income levels? 9. Welfare... that the government protects all citizens against poverty? 10. Transparency of political decisions... that the government explains its decisions to voters? 11. Direct democracy... that citizens have the final say on the most important political issues by voting on them directly in referendums? 12. Retrospective accountability... that governing parties are punished in elections when they have done a bad job? 13. Deliberation... that voters discuss politics with people they know before deciding how to vote? 14. Competition of parties... that different political parties offer clear alternatives to one another?? Answers were made on a 11-point scale where 0 means "Not at all important for democracy in general" and 10 means "Extremely important for democracy in general" We will call this first factor in the following analysis the liberal model 2 These conceptions of democracy agree broadly with those of electoral democracy proposed by Dahl (1989) and the broader model of liberal democracy, presented by Diamond (1999). In addition to the exploratory factor analysis (EFA), confirmatory factor analysis (CFA) on the basis of these two thereotical models was carried out. The measuring structure was largely confirmed. However, the following analysis is using the results of the EFA since the interpretation can be made more intuitive. After all, calculations with structural equation models and measurements strongly based on theoretical grounds instead of OLSregressions with factor scores as dependent variables did not lead to significant changes in the main results which a presented in the following. All calculations were done with STATA 13. Replication data for the following results as well as alternative specifications with SEM can be obtained from the author. 11

12 of democracy. The second factor is that of the socialist model of democracy with the items of Social equality and Welfare. For the following regression analysis the factor values were used as dependet variables with a standardized value range from 0 to 10. Table 2: Dimensions of democratic value orientations Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Free and Fair elections 0,74-0,10-0,01 0,24-0,16 Viable opposition 0,81-0,09 0,15-0,13 0,06 Reliable information provision 0,73 0,06-0,11-0,14-0,18 Freedom of press 0,67-0,06 0,05-0,42 0,06 Protection of minority rights 0,62 0,08-0,07 0,01-0,20 Equality of the judicial system 0,54 0,15-0,26 0,19 0,02 Horizontal accountability 0,56 0,17-0,25 0,10 0,24 Social equality -0,04 0,71 0,06-0,08-0,07 Welfare -0,06 0,80 0,02-0,05-0,11 Transparency of political decisions 0,31 0,50-0,08 0,05-0,00 Direct democracy 0,14 0,42 0,06 0,00 0,04 Retrospective accountability 0,33 0,33-0,05-0,02 0,25 Deliberation 0,48 0,04 0,34 0,20 0,02 Competition of parties 0,33 0,19 0,37 0,12 0,07 European Social Survey (2014). Factor analysis with oblimin-rotation and the assumption of a correlated factor structure. Central independent variables are the belonging to the East or the West and the cassification of cohorts for representing differently socialised generations. With additional items of the German partial study of the European Social Survey (2014) we are able to measure the East-West affiliation very accurately. Respondents were asked for the place of residence before 1990 and the place of residence before 1990 of the parents if the respondent was born in or after 1990, respectively. Using these items we can operationalize wether the primary socialisation has happenend in the East or the West with a dummy variable that takes 1 if we can assume a socialisation in the East for the respondent or a 0 if we can not. Political socialisation is measured through generational effects, for this purpose cohorts are used. This requires a sophisticated approach since there are a lot of possibilities depending on the assumptions of socialisation con- 12

13 cepts. Within the concept of political generations (Gensicke, 1998) the assumption is, that certain political or societal events shape the political orientations of a whole generation. The insertation of shaping power of such events is in this connection assumed with the beginning of the 10th year of age. But recent empirical research shows that the development of political attitudes could easily assumed to begin before this time (Neundorf et al., 2013; Healy and Malhotra, 2013; van Deth et al., 2011). So, it is still a open question when political socialisation starts in the life cycle. Concerning the scope of the following analysis we are focussing on the youngest generation, the first one socialised in unified Germany after 40 years of separation. Because of this purpose a search for shaping events is not relevant. Instead and in line with recent research results we set the first year of birth for this youngest generation to So the oldest members of this generation where 5 years old when the Iron Curtain fell and 27 when the survey was conducted. From this generation on the backward classification of older generations follows pragmatic reasons in the first place, especially a comparable number of cases between generations was a relevant criterion. 3. For the number of cases in the East and the West for the Cohorts see Table 3. Table 3: Number of cases by cohorts Number of cases Cohorts West East Total Total Source: European Social Survey (2014). Number of cases are unweighted. As control variables we include the education level, political interest as a proxy for political involvement, political ideology placement on the left 3 A second important criterion was approximately equal high time span between the starting end the endpoint of a cohort, which range from years. The exception builds the oldest generation, here the effective number of cases would have been to small if it was split up. Different generation classifications did not change the main results. 13

14 right scale, gender and situative factors like the assessment of the national economic situation, unemployment in the last 5 years and if there are living kids in the household of the respondent. All of these control variables were coded as dummy variables. Empirical findings First we will look at the descriptive findings. Watching Figure 1, we see large differences at the mean differences of the specific items and the factor variables between respondents with a socialisation happend either in the East or in the West. In particular we can observe significant mean differences, where western socialised respondents score high at the items to Free and fair elections (t = 4.85), Viable opposition (t = 3.37), Freedom of press (t = 2.96), Protection of minority rights (t = 5.18) and the model of liberal democracy (t = 3.63). Reversely, we observe significant higher values for eastern socialised respondents at the questions to Social equality (t = 10.11), Welfare (t = 5.25), Direct democracy (t = 7.24), Retrospective accountability (t = 2.83) and the model of socialist democracy (t = 8.05). Though, with regard to this first results of the analysis it becomes obvious that the findings presented by Fuchs (1999) still stand on empirical grounds. Obviously, people socialised in different parts of Germany still differ in their understanding of democracy (as expressed in H 1 ). While people socialised in West Germany favour a model in terms of liberal democracy, people socialised in the East favour a model in terms of socialist democracy (as espressed in H 2 ). Nearly 25 years after reunification and 15 years after the findings presented by Fuchs (1999) we still do not find unity on the value level regarding orientations towards democratic norms and principles 14

15 in reunified Germany. Figure 1: Mean differences in separate democracy principles and normative models of democracy in east-west comparison Free and Fair elections Viable opposition Reliable information provision Freedom of press Protection of minority rights Equality of the judicial system Horizontal accountability Social equality Welfare Transparency of political decisions Direct democracy Retrospective accountability Deliberation Competition of parties Liberal Democracy Socialist Democracy Approval scale west east Source: European Social Survey (2014). Own calculations, show are separate, weighted means for east- and west-socialization plus the associated 95% confidence intervals. Answers were made on a 11-point scale where 0 means "Not at all important for democracy in general" and 11 means "Extremely important for democracy in general" But what is striking too, is that the mentioned differences, although statistical significant, seem to be very small in absolute terms reaching the size of at least one scaling point not only once. With this said the next interesting question is how this differences developed between different generations. For this purpose multivariate OLS-regressions were estimated to explain the approval of the two different models of democracy. Results of the first models, explaining approval to a liberal model, are presented in Table 4. Model 1 shows negative significant effects of being socialised in the East as well as being part of the youngest cohort. Women and respondents with a right ideological self-classification also show less approval to a liberal model of democracy. Positive effects show up with a view to an increasing level 15

16 of education and increasing political involvement. In addition, respondents born between (oldest cohorts combined) and being ideologically left rather tend to for a liberal democratic model. In Model 2 additionally specified interaction effects with the East-socialisation are included. None of these interaction effects is significant. Most of the effects of Model 1 retain their direction and significance, only the effect of the East-socialisation becomes insignificant. Model 3 additionally takes into account situational explanation factors, of which only the assessment of the national economic situation has a significant, positive effect. At the direction and significance of the previously described effects, this does not change anything. Finally, the fully specified Model 4 additionally comprises from interaction effects of these situational factors with the East-socialisation. Here, the effect of gender and leftist ideology lose their significance in addition to the effect of the East socialisation. Of the interaction effects still none shows significance. The focus of interpretation lies in accordance with the theoretical expectations in the generational effects in both parts of the country. The following can be concluded: the effect of belonging to a generation is only significant for respondents, who were socialised in the West, the effect of the two older cohorts ( born) is positive, the effect of the most recent cohort ( born) is negative. Interaction effects with the East-membership are not significant. This finding concerning the recent and youngest generation can be interpreted as follows: this generation approaches in comparison to the older generations to converge each other (best seen through the positive sign of the effects of the generations of and , as well as the negative sign of the effect of the youngest generation). This approach, however, is one sided and based on a movement of the West. This finding is confirmed in the graphical representation of the estimated values for certain 16

17 groups on the basis of this regression model, which can be seen in Figure 2. Table 4: Determinants of approval of the model of liberal democracy Model 1 Model 2 Model 3 Model 4 East Born Born Born Born Female Mid-level education High education Low political interest Some political interest High political interest Ideological left Ideological right East x born East x born East x born East x born East x female East x mid-level education East x high education East x low political interest East x some political interest East x high political interest East x ideological left East x ideological right Economic situation good Child in household Unemployed (last 5 years) East x Economic situation good 0.11 East x Child in household 0.05 East x Unemployed (last 5 years) 0.16 Constant N R adj. R Source: European Social Survey (2014), own calculations. Unstandardized regression coefficients, robust standard errors (not shown). : p 0.05, : p 0.01, : p We observe a significant difference between East and West in the birth cohort, which is the cohort were we can assume a full socialisation through the given political regime as the born range is a long time after the 17

18 second world war and before the fall of the Iron Curtain. This difference levels in the following generations, to the point where effect and confidence interval for the youngest generation are almost identical. This finding is directly the opposite from the prior articulated theoretical expectation. On the basis of the model of institutional learning we would expect an approximation of the East to the West (H 3 ). What we observe however is an convergence of the West to the East. A significant change in the attitude to liberal democracy of individuals socialised in the East across generations can not be stated: All confidence intervals overlap. The latest generation of West Germans, however, differs significantly from that born between (both cohorts combined). Figure 2: Estimated values: liberal understanding of democracy on the basis of the regression model Born Born Born Born Born Male Female Low education Mid-level education High education No political interest Low political interest Some political interest High political interest Ideologically lefts Ideologically centered Ideologically right Economic situation bad Economic situation good No child in household Child in household Not unemployed (last 5 years) Unemployed (last 5 years) Liberal democracy (factor scores) west east Source: European Social Survey (2014), own calculations. Shown are the estimated values for the approval of the normative model of liberal democracy and the associated 95% confidence intervals for certain population groups and attitudes on the basis of a linear regression (Model 4 in Table 4). This finding is surprising and can not be explained by any given theoret- 18

19 ical model. Instead only ad hoc explanations could be used. In total, explanations must be found for two questions raised by this finding: First Why is a movement of the West German population oberserved? Second Why is there no movement in the East German population, as it is predicted by the theoretical model? Before explanations occur to these questions, however, first the results are presented for the analysis of socialist model of democracy as the dependent variable to check whether the theoretical expectations can be confirmed in this case. When asked to consent to a socialist model of democracy, a similar, but here theory compliant picture shows up. Looking at the regression results, we observe strong generational effects in Table 5 both for the West and for the East. Except for the level of education, gender and a left ideological self-classification, none of the control variables show any a significant effect. The estimated coefficients are consistent in direction over all four models. As with the estimation of the approval of the liberal democratic model, the effect of East-socialisation in the models with interaction effects (Model 6 and 8) loses its significance. In Models 5 and 7, however, it is significant and positive. So it can be said, that an East-socialisation leads to higher approval of the normative model of a socialist democracy. From the situational factors, it is again the assessment of the national economic situation, which has a significant effect. In accordance to the model of socialist democracy, however, it decreases with a positive assessment. In regard to the generation effects, consistently negative effects can be oberserved. Those are significant for the West in the recent generation and for the East in the recent three generations (Model 8). The results of the regression models indicate a close convergence of the East to the West. At the same time, the agreement to a socialist model of democracy declines in the West within the youngest generation. 19

20 Table 5: Determinants of approval to the model of social democracy Modell 5 Modell 6 Modell 7 Modell 8 East Born Born Born Born Female Mid-level education High educaiton Low political interest Some political interest High political interest Ideologically left Ideologically left East x born East x born East x born East x born East x female East x mid-level education East x high education East x low political interest East x some political interest East x high political interest East x ideologically left East x ideologically right Economic situation good Child in household Unemployed (last 5 years) East x Economic situation good 0.07 East x Child in household 0.01 East x Unemployed (last 5 years) 0.36 Constant N R adj. R Source: European Social Survey (2014), own calculations. Unstandardized regression coefficients, robust standard errors (not shown). : p 0.05, : p 0.01, : p This finding is confirmed with a look on the predicted values on the basis of the regression estimates. In Figure 3 we see a significant movement away from a socialist democracy and a significant convergence between the two parts of the country in recent cohorts with an approach of the East to the 20

21 West. This expectation is consistent with our theoretical expectation (H 3 ) and thus corresponds to the predictions that meet the model of institutional learning. As a limiting factor should be noted, however, that an approach (or no significant difference) does already show up in the birth cohort with the overlapping of the confidence intervals. But as expected, this should be the generation whose socialisation was strongest influenced by the individual regime and its institutional setting. Figure 3: Estimated values: socialist understanding of democracy on the basis of the regression model Born Born Born Born Born Male Female Low education Mid-level education High education No political interest Low political interest Some political interest High political interest Ideologically lefts Ideologically centered Ideologically right Economic situation bad Economic situation good No child in household Child in household Not unemployed (last 5 years) Unemployed (last 5 years) Socialist democracy (factor scores)) west east Source: European Social Survey (2014), own calculations. Shown are the estimated values for the approval of the normative model of liberal democracy and the associated 95% confidence intervals for certain population groups and attitudes on the basis of a linear regression (Model 8 in Table 5). In summary, some expectations can be confirmed. We still find significant differences in the democratic value orientations between East and West Germans (H 1 ). So, stronger endorsement of the model of liberal democracy can be explained by a socialisation in the West and stronger endorsement 21

22 of the model of a socialist democracy can be attributed to socialisation in the East (H 2 ). Looking at the approval of the model of socialist democracy, our expectations are also confirmed. There is a strong convergence of the recent East-generation to the youngest generation in the West (H 3 ). An effect which is largely due to the socialisation in unified Germany. As for the preference of a liberal democratic model, our theoretical expectations are not confirmed. For a possible explanation of this finding at least two questions should be answered: 1) Why ist the West moving to the East? 2) Why is there completely no movement in the East? One possible explanation for the movement of the West could be that a kind of class enemy no longer existed after the fall of the Iron Curtain on both sides of the wall and so it did not in the birth cohorts after 1985, too. In the generation of their parents and grandparents the definition of the own system (Western democracy) always went hand in hand with the rejection of the opposing system (socialism). This distinction no longer exists for the today year olds. Fundamental normative principles such as freedom of the press, effective opposition rights, free and fair elections are therefore no longer in enduring danger through a "communist threat" and therefore are no longer the focus of one s own perception and self-description. With this, the emphasis of these normative principles in the youngest generation will automatically loose importance. However, this decreasing importance due to a lack of emphasis must not take place solely in the West. Through this lack of emphasis, ultimately, in the public debate, in family circles, as well as in school these normative principles are less salient within the recent East German generation which they got through the socialisation process of the older Western generations. Thus, omitted enemy stereotypes could also give an answer to the second question. Of course, it can be argued in a reverse direction, too. A possible approach of the West to the East has recently been in the public debate even accompanied with the call to East German citizens to be more confident. However, both explanations remain only tentative since they can not be 22

23 tested with the available data. In addition, however, it may also involve life cycle instead of generation effects. Again, this is also a question that can not be resolved with the available data. In conclusion, therefore no final answer can be given to the questions of why the West moves to the East, when approving the model of liberal democracy, and why in the East there is no motion. So only the need for further research to can be stated here. Concluding remarks The aim of this paper was to re-examine the findings presented by Fuchs (1999) in the late 90 s. Doing so, the question was, if the differences in political value orientations between East and West Germans, precisely differences in the normative understanding of democracy, converges in the recent generations as postulated by socialisation theories, or whether differences could still be found. To answer this question, first the concepts of understanding of democracy and democratic value orientations were conceptualized. Then, a theoretical model was presented that is able to explain possible convergence effects between East and West German citizens, especially in the recent, first jointly socialised cohort. Based on this model and the previous state of research unique expectations were expressed: We still expected differences in the democratic value orientations between East and West Germans (H 1 ); It was expected that East Germans favour a model of socialist democracy while West Germans favour a model of liberal democracy (H 2 ); These differences were expected to decrease in the most recent cohort, through approaching by the East to the West (H 3 ). The empirical results given through descriptive findings and regression analysis only give partial support to the theoretical expectations. In many normative principles and thus in many democratic value orientations significant differences between East and West Germans continue to exist (H 1 ). We found a still more socialist understanding of democracy in the East German population and a more liberal understanding of democracy dominating in 23

24 the West German population (H 2 ). Also, a convergence within the youngest cohort (and to some extent even in the cohort before) concerning democratic value orientations could be found (H 3 ). This expectation was confirmed especially for the approval of a normative model of socialist democracy. For the model of liberal democracy our findings are contrary to the theoretical expectation. Looking at support for a liberal model of democracy (operationalized by factor scores on the basis of exploratory factor analysis), the latest generation of the West approaches to the approval ratings of the East. This finding is surprising in two ways: first it contradicts the expectations of the theoretical model of "institutional learning", after which we would expect the opposite direction. Second, in the youngest generation of the West German population compared to the generation of their parents and grandparents a negative attitude towards the normative principles that characterize a liberal democracy seem to develop. Although these differences appear in rather small dimensions, this finding seems questionable from a normative perspectiv taken by democratic theory. This finding could only be explained with ad hoc statements, which can not be tested with the available data. Further limitations mitigate the validity of the overall findings: using the available data it is not possible to diver between generation- and life cycle effects. Theoretically, it is possible that the democratic value orientations of their parents and grandparents have been at the same level as the ones of the children and grandchildren generation, when the former were in the age of the latter. So, it is possible that the now determined rather high level of approval of older generations has developed at a later age. This question may only be tested with longitudinal data that take a time perspective into account by repeated measurements of the same individuals (so-called panel surveys). Such data is not available with the items used in this study. Thus, there are profound reasons for further research: first, the lack of theoretical explanatory power of the socialisation model, when approving the model of liberal democracy; second the alarmingly result from a point of 24

25 view taken by democratic theory when finding decreasing democratic value orientations and the lack of theoretical explanation for this; and third, the need for panel data to answer the question whether the results presented here are caused by life cycle or generation effects. References Almond, Gabriel A. (1980): The Intellectual History of the Civic Culture Concept, in: Almond, Gabriel A./Verba, Sidney (Eds.): The Civic Culture Revisited, Newbury Park: Sage, pp Almond, Gabriel A./Verba, Sidney (1963): The Civic Culture. Political Attitudes and Democracy in Five Nations, Princeton: Princeton University Press. Arzheimer, Kai (2005): Freiheit oder Sozialismus? Gesellschaftliche Wertorientierungen, Staatszielvorstellungen und Ideologien im Ost-West- Vergleich, in: Gabriel, Oscar W./Falter, Jürgen W./Rattinger, Hans (Eds.): Wächst zusammen, was zusammengehört? Stabilität und Wandel politischer Einstellungen im wiedervereinigten Deutschland, Baden- Baden: Nomos, pp Arzheimer, Kai (2013): Twenty Years After: Sozial- und wirtschaftspolitische Einstellungen von Ost- und Westdeutschen im Vergleich, in: Keil, Silke I./Thaidigsmann, S. Isabell (Eds.): Zivile Bürgergesellschaft und Demokratie: Aktuelle Ergebnisse der empirischen Politikforschung. Festschrift für Oscar Gabriel, Wiesbaden: Springer VS, pp Arzheimer, Kai/Klein, Markus (2000): Gesellschaftspolitische Wertorientierungen und Staatszielvorstellungen im Ost-West-Vergleich, in: Falter, Jürgen/Gabriel, Oscar W./Rattinger, Hans (Eds.): Wirklich ein Volk? Die politischen Orientierungen von Ost- und Westdeutschen im Vergleich, Opladen: Leske + Budrich, pp Arzheimer, Kai/Rudi, Tatjana (2007): Wertorientierungen und ideologische Einstellungen, in: Rattinger, Hans/Gabriel, Oscar W./Falter, Jürgen W. 25

26 (Eds.): Der gesamtdeutsche Wähler. Stabilität und Wandel des Wählerverhaltens im wiedervereinigten Deutschland, Baden-Baden: Nomos, pp Campbell, Ross (2011): Socialist Values and Political Participation in Germany: A Barrier to Inner Unity?, West European Politics, 34, pp Campbell, Ross (2012): Values, Trust and Democracy in Germany: Still in Search of Inner Unity?, European Journal of Political Research, 51, pp Conradt, David P. (1974): West Germany: A Remade Political Culture?: Some Evidence from Survey Archives, Comparative Political Studies, 7, pp Dahl, Robert A. (1989): Democracy and its Critics, Yale: Yale University Press. Dalton, Russell J. (1994): Communists and Democrats: Democratic Attitudes in the Two Germanies, British Journal of Political Science, 24, pp Davidov, Eldad/Siegers, Pascal (2010): Comparing Basic Human Values in East and West Germany, in: Beckers, Tilo/Birkelbach, Klaus/Hagenah, Jörg/Rosar, Ulrich (Eds.): Komparative empirische Sozialforschung, Wiesbaden: VS Verlag für Sozialwissenschaften, pp Diamond, Larry J. (1999): Developing Democracy. Toward Consolidation, Baltimore: Johns Hopkins University Press. Easton, David (1965): A Systems Analysis of Political Life, New York u.a.: Wiley. Easton, David (1975): A Re-Assessment of the Concept of Political Support, British Journal of Political Science, 5, pp European Social Survey (2013): Round 6 Module on Europeans Unterstandings and Evaluations of Democracy - Final Module in Template, London: Centre for Comparative Social Surveys, City University London. 26