6 Hungary Structure and dynamics of democratic consolidation Christian W. Haerpfer Introduction The end of communism started with an event at the Hungarian-Austrian border in May 1989. Hungarian politicians arguably dismantled the Iron Curtain - assisted by the Austrian Foreign Minister and observed by Hungarian and Austrian media - by providing their fellow citizens with international passports. In September 1989, the Hungarians finally permitted citizens of the German Democratic Republic to pass the same Hungarian-Austrian border to leave towards West Germany. According to some scholars (e.g. Swain 1993) these political events in Hungary at the dusk of communism were of significant importance for accelerating the speed of the central and eastern European revolution at the end of 1989. Hence, it is not an historical exaggeration to postulate that the first cracks in the Iron Curtain in the annus mimbilis of 1989 occurred in the Hungarian part of the communist bloc. Over 15 years after the events of 1989, the democratic transformation of the Hungarian political system is almost complete. This chapter focuses on an analysis of political attitudes of Hungarian citizens. Particular attention is given to support of democracy and related attitudes and values. The first section deals with support for democracy. In addition to the database especially generated for this volume, we also make use of data collected in 1994, 1996 and 1998 by the New Democracy Barometers and related surveys. This enables us to put support for democracy in Hungary in a cross-time and cross-country perspective. The second section analyzes the acceptance of political tolerance, of the principle of non-violence and support of solidarity with the poor in Hungary. The third section focuses on political motivation and protest behavior. The topic of section four is civil society. The impact of selected variables on attitudes toward democracy and autocracy is measured by multivariate regression analyses. A multivariate regression model that integrates the whole set of democratic attitudes is presented in the conclusions. Hungary 149 The dynamics of democratic consolidation in Hungary between 1989 and 1999 Political scientists generally tend to agree that Hungary, together with Poland and the Czech Republic, belongs to the leading troika of democratic consolidation in post-communist Europe (see Derleth 2000; Haerpfer 2002) There are different types of transformation processes. Hungary is the best example of a gradual and slow transformation. Its success was promoted through proto-forms of pluralism and market economy that existed in Hungary as early as in the 1980s. In addition, the Hungarian transformation was a non-violent one (in contrast to, say, Romania), and based upon elite consensus (Derleth 2000). Rudolf L. Tokes (1996) and Laslzo Bruszt (1990) termed it a "negotiated revolution." The only country in central Europe in which an absolute majority evaluated the communist one-party system in a positive way is Hungary. This has to do with the special nature of the so-called "Goulash Communism" that was implemented under Secretary General Janos Kadar during the 1980s. Many Hungarians remember this decade as the "golden era" in modern Hungarian history. The share of citizens who value the ancien regime increased sharply from 51 percent in 1991 to 68 percent in 1992, decreased again to 58 percent in 1994 and remained at that comparatively high level until 1998. In an earlier study I have reported changing levels of public support of democracy in Hungary. This index is conceptually similar but differently operationalized than the "democracy-autocracy index" which is used throughout this volume.1 However, it also allows us to identify citizens who can be characterized as "democrats," i.e. respondents who support liberal democracy as an ideal without necessarily evaluating the performance of the political regime in a positive way. That is to say, the index of democracy does not measure the evaluation of the performance of the current government at a given point in time but the level of support for democracy as a form of government in contrast to any other non-democratic regime. Table 6.1 includes only those respondents who can be categorized as "democratic" (scores 8-10, see endnote 3). The "Haerpfer-index of democracy" encompasses the category of "democrats" (^values 9-10) as well as the group of "weak democrats" (=value 8). The "democracy-autocracy index" on the other hand, distinguishes "strong democrats," weak democrats, undecided citizens and autocrats. Its rationale and construction has been thoroughly described in the introductory chapter.2 In Table 6.1, strong and weak democrats have been collapsed into the category "democrats" (WVS 1999). It should be stressed thai because of the differences in index construction, a direct comparison of proportions of democrats in the period of 1994 to 1998 and 1999 is not possible. More than 60 percent of Hungarian respondents can be characterized as democrats by the more restrictive Haerpfer index. Collapsing "strong" 150 Christian W. Haerpfer Table 6.1 Proportion of democrat* as measured by Haerpfer's democracy index 1994-8 and the democracy-autocracy index in post-communisi societies, 1999 Country NDBß NUB 4 NDB 5 wvs 1994 1996 1998 1999 Central Europe - mean 57 52 61 89 1. Poland 47 52 66 84 2. Czech Republic 77 69 65 95 3- Hungary 50 38 62 82 4. Slovenia - 47 57 92 5. Slovakia 55 53 55 94 South-eastern Europe - mean 56 52 55 88 1. Romania 59 60 56 86 2. Croatia 65 51 55 97 3. Bulgaria 44 45 54 79 4. FRY - - 37 91 Northern Europe - mean 28 30 - 85 1. Estonia 43 46 - 90 2. Lithuania 18 27 - 80 3. Latvia 22 18 - 86 Eastern Europe - mean 21 14 30 70 1. Belarus 23 15 41 77 2. Ukraine 25 12 19 74 3. Russia 15 - - 58 Source: NDB 3 (1994), NDB 4 (1996), NDB 5 (1998), New Baltic Barometer 1994, New Baltic Barometer 1996, WVS - World Values Survey. (42 percent) and "weak" democrats (51 percent) the democracy-autocracy index brings the figure up to 82 percent in 1999. Thus we may well speak about an almost unanimous acceptance of democracy as a form of government in Hungary, In our earlier work, we have classified a country in which 40 percent of citizens were democrats (as measured by the Haerpfer index) in an "emerging democracy." Observing a rise to more than 60 percent democrats in 1998 justifies a classification of Hungary as a "consolidated democracy" on the micro-level, i.e. on the level of the Hungarian citizenry.3 In the first years of political transformation we observed a high degree of skepticism with regard to democracy in general and the new Hungarian regime in particular. Only 38 percent of the respondents qualified as democrats in 1996. This share increased considerably until 1998. Together with Poland and the Czech Republic, Hungary clearly belongs to the group of stable and consolidated democracies in central and eastern Europe as far as the populace is concerned. The Haerpfer index of democracy clearly shows the different degrees of democratization on the micro-level in post-communist Europe. In 1998, Hungary 151 an average of 61 percent of the respondents can be characterized as democrats. With respect to public support, the countries of the central European region are already consolidated democracies. The country with the highest share of democrats is Poland (66 percent in 1998). In 1994, only 47 percent of Polish respondents could be characterized as democrats. Their share increased to 52 percent in 1996. Thus, Poland crossed the border from an emerging to a consolidated democracy some time between 1996 and 1998, and since then has been well prepared for European integration. The Czech Republic, with a share of 65 percent of democrats in 1998, occupies the second rank. Compared to the Polish trend, the development of public acceptance of democracy in the Czech Republic was reversed. In 1994, 77 percent of the Czech citizenry indicated democratic attitudes; there seemed to have been "democratic euphoria." The political development and its public perception in the period of the governments of Prime Ministers Vaclav Klaus and Milos Zeman acted as a cold shower for this initial enthusiasm and for the self-ascribed role of the Czech Republic as an "ideal-type of democratic transformation." Nevertheless it is justified to label the current Czech political system as a consolidated democracy. According to the criteria mentioned above, three out of 15 post-communist countries have reached the status of consolidated democracy after ten years of political transformation: Poland, the Czech Republic and Hungary. In Slovenia, the Haerpfer democracy index identified 57 percent democrats in 1998, which puts Slovenian democracy just below the threshold of 60 percent. In Slovakia, a stable majority of well above 50 percent of the respondents score as democrats; in 1998 their share reached 55 percent. This development was validated by the general election in autumn 1998, when the government of Prime Minister Dzurinda replaced the government of Prime Minister Meciar, who was not known to be a strict defender of democratic values. Summarizing results of the Haerpfer democracy index which allowed a comparison over time one can safely say that, in 1998, there is no new democratic regime in central and eastern Europe that can rely on less than 50 percent of their citizens to support democracy as an ideal. The general preference of democracy over autocracy finds an even greater expression in the results generated by the democracy-autocracy index. We want to reiterate, however, that the operationalization of the two indices is different and the proportions of democrats cannot be compared directly. To conclude this section we will analyze the social structural composition of the four groups as defined by the democracy-autocracy index. As mentioned above, strong democrats were distinguished from weak democrats, and undecided citizens from autocrats. Who are the strong democrats in terms of the Hungarian social structure? They are primarily male (38 percent), they belong to the young generation (40 percent), 152 Christian W. Haerpfer received tertiary education (45 percent), live predominandy in big cities (41 percent) and belong to the middle class (46 percent). The weak democrats have their strongholds also among men (49 percent), live in the countryside (49 percent) and belong to the working class (51 percent). What is the social location of the undecided citizens? They are found primarily among women (11 percent), in the older generation (aged 55 years and above), among respondents who received primary education only (16 percent), in the lower class as well as the working class (13 percent each). Finally, who are the autocrats? They are found primarily among women (10 percent), in the lowest educational stratum (15 percent), in small villages (11 percent) and first of all in the lower class (16 percent). Impact of public support of political principles on support of democracy The analysis of the level of public support for political principles and values in Hungary indicates that the Hungarian democrats reach far beyond a Schumpeterian concept of democracy. They embrace a class of political ori- Table 6.2 Democracy-autocracy index, 1999 Hungary 153 entations which have an obvious and direct impact on the quality of the democratic process. With regard to democratic consolidation, the most important political principle is tolerance, followed by the acknowledgement of the illegitimacy of the use of violence in politics and solidarity with the poor. Political tolerance is operationalized as tolerance toward homosexuality (Table 6.3). A total of 37 percent of the respondents are tolerant toward homosexuality and homosexuals. The level of tolerance toward homosexuals is higher among men, the young, in bigger cities, in the middle class and in the higher educational strata. Intolerance toward homosexuality is more widespread in the working class, in rural Hungary, in the lowest educational groups, in the old generation and among women. The second most significant influence on the level of public support for basic democratic principles is the respondents' rejection of violence as a legitimate means in politics (Table 6.6: Beta-coefficient = 0.13). An overwhelming majority (80 percent) of Hungarian citizens believe that the use of violence to pursue political goals is "never justified." The principle of non-violence appears to be deeply rooted in Hungarian political culture: "The Hungarian transformation was a peaceful, elite-controlled change," a "negotiated revolution... . This idea has had a powerful impact on the political culture: it has promoted the acceptance of such principles as Table 63 Political tolerance - attitudes toward homosexuality, 1999 Strong democrats (%) Weak democrats (%) Undecided (%) Autocrats {%) Tolerant (%) intolerant (% ) Total 36 46 9 9 Total 37 63 Gender Gender Women 34 44 11 10 Women 35 65 Men 38 49 7 7 Men 38 62 Age Age 18-34 years 40 44 7 8 18-34 years 47 53 35-54 years 36 47 8 9 35-54 years 39 61 55-90 years 32 48 11 9 55-90 years 21 79 Education Education Primary 22 47 16 J5 Primary 19 81 Secondary 44 43 7 7 Secondary 39 61 Tertiary 45 48 4 4 Tertiary 49 51 Town size Town size 100-9,999 30 49 10 11 100-9,999 24 76 10,000-99,999 36 47 8 9 10,000-99,999 39 61 100,000-2,000,000 41 44 9 7 100,000-2,000,000 44 56 Subjective dais Subjective class Lower class 24 47 13 16 Lower class 24 76 Working class 27 51 13 9 Working class 29 7) Middle class 46 42 5 7 Middle class 42 58 Upper class 34 49 9 9 Upper class 47 53 154 Christian W. Haerpfer non-violence, self-restraint, political pragmatism and readiness for negotiations. All the political leaders ... wanted to avoid violent solutions" (Bozoki 1999: 108). This principle, however, is in the first place limited to domestic Hungarian politics and does not necessarily apply to international politics, the question of war and Hungarian participation in international conflicts. One-fifth (20 percent) of the respondents believe that violence in politics can be justified under certain circumstances, a result that casts a shadow over Hungarian politics after ten years of political transformations. Gender does not affect the respondent's attitude toward the question of the legitimacy of violence in politics. This is also true for age, level of formal education or the urban-rural dimension. Only the self-ascribed belonging to a social class proved to be of importance: one-quarter of respondents who attribute themselves to the working class (24 percent) and 27 percent of the upper class agree with the general statement that violence could be justified as a mean in politics. Among members of the middle class and the lower class, political violence is not accepted by 80 to 90 percent. The third most important effect on the level of support for democracy was the level of solidarity with the poor (Table 6.5: Beta-coeffi- Table 6.4 Legitimacy of violence in politics, 1999 Never justified (%) Justified (%) Total 80 20 Gender Women 81 19 Men 78 22 Age 18-34 years 79 21 35-54 years 78 22 55-90 years 81 19 Education Primary 81 19 Secondary 80 20 Tertiary 78 22 Town size 100-9,999 80 20 10,000-99,999 78 22 100,000-2,000,000 81 19 Subjective class 1 ,ower class 88 12 Working class 76 2-1 Middle class 82 18 Upper class 73 27 Hungary 155 cient = 0.i0). The absolute majority of 75 percent of all citizens indicated a high level of solidarity with the poor (Table 6.5). This is especially true for women, the middle-aged, for respondents with an average level of formal education, in villages and small towns and in the lower class, which is most affected by poverty itself. Solidarity with the poor is somewhat less expressed by the higher classes, especially by the middle class and the upper class, in big cities and by men. The relative effect of important political principles and values on the level of public support for democracy indicates that the concept of democracy in Hungary is not just limited to a set of abstract rules in the political arena, but is based on the acceptance of a pluralist society and its core political values. Impact of political motivation and political participation on support of democracy Political motivation In this section, the concept of political motivation is operationalized by variables measuring j "political discussion with peers," "political interest" Table 6.5 Solidarity wi th the poor, 1999 High(%) Medium (%) Low (%) Total 75 23 2 Gender Women 76 22 3 Men 73 25 2 Age 18-34 years 73 24 3 35—54 years 78 21 1 55-90 years 75 22 3 Education Primary 73 24 3 Secondary 79 21 1 Tertiary 73 24 3 Town size 100-9,999 77 21 2 10,000-99,999 78 22 0 100,000-2,000,000 70 25 5 Subjective class Lower class 84 16 0 Working class 79 19 9 Middle class 72 25 3 Upper class 68 29 3 156 Christian W. Haerpfer Table 6.6 Impact of political principles on support of democracy (model A; OLS multiple regression; dependent variabler democracy-autocracy index) Political principles Beta Significance R2 ■ XI Tolerance 0.24 0.00 — X2 Illegitimacy of violence as a political means 0.13 0.00 - X3 Solidarity with the poor 0.10 0.00 - Explained variance by model A - - 0.08 and the respondent's evaluation of the "importance of politics." The frequency of political discussions with peers serves as the main indicator of political motivation or involvement. In the survey IV percent of the respondents discuss political matters frequently with their friends (Table 6.7). The majority of 56 percent of Hungarian citizens discuss political issues now and then, whereas 27 percent never discuss politics with their friends. We found a strong influence of gender on the frequency of political discussions with friends: men discuss politics more often than women. Every fifth man is involved in political discussions with his friends frequently (21 percent), whereas only 13 percent of all Hungarian women frequently debate politics. One-quarter of Hungarian men do not talk politics (24 percent) with their friends. When women are meeting, one-third of them never talk about politics (30 percent). There is a linear relationship between age and the frequency of political debate with friends. The older the citizen, the more he/she discusses politics with friends. Among the younger generation, only 10 percent are discussing political matters frequendy. Roughly one-fifth of all citizens between 35 and 54 years of age, and of those aged more than 55 years, are involved in political debates with their friends frequently. The higher the respondents' level of formal education, the more frequent are political debates among friends. In the category of respondents with the lowest level of formal education, only 11 percent discuss politics with their friends frequently. This share increases to 16 percent among those who received secondary education. The highest frequency of political debate with friends was found in the group of respondents who received tertiary education. Every fourth Hungarian university graduate discusses politics with friends frequently (26 percent). There is also a clear relationship between the urban-rural dimension and the frequency of political discussion. The larger the size of the local community, the more often political discussions take place with friends. In rural areas and small towns, only 14 percent of the population discuss politics regularly. The proportion of politically interested persons goes up to 17 percent in medium-sized towns. We found the highest frequency of political debate in cities with more than 100,000 inhabitants (19 percent) and among the middle and the upper classes. One-fifth of the middle class and 16 percent of the upper class are discussing politics frequently. Hungary 157 Table 6.7 Political discussion with friends, 1999 Frequently (%) Occasionally (%) Never (% ) Total 17 56 27 Gender Women 15 57 30 Men 21 55 24 Age 18-34 years 10 62 28 35-54 years 20 62 19 55-90 years 22 43 35 Education Primary 11 49 40 Secondaiy 16 59 26 Tertiary 26 60 14 Town size 100-9,999 14 54 32 10,000-99,999 17 55 27 100,000-2,000,000 19 58 23 Subjective class Lower class 7 34 59 Working class 15 51 34 Middle class 20 61 19 Upper class 16 69 15 A second indicator for political involvement is the assessment of the importance of politics (Table 6.8). It had, however, no impact on the public level of support for democracy. Siighdy less than one-third of the respondents (27 percent) state that politics is important in their life, while the remaining 73 percent indicate that politics is not important in their everyday life. The difference between men (25 percent) and women (29 percent) with regard to importance of politics in their life is rather small. Instead, we found a linear relationship with age. The older the respondent, the more he or she is inclined to think that politics is important. Hence, we found that 43 percent of the older generation is convinced that politics is important for their own life. The impact of the Second World War, communist rule until 1989 and the advent of democracy after 1989 might have contributed to the comparatively high esteem politics enjoys among older citizens. This pattern is unusual, since it is usually the middle generation who is found to be most interested in politics - as these are the citizens who participate most actively in all spheres of live. In Hungary, only 25 percent of the middle-aged indicate that they think politics is important in their lives and only 24 percent of those between 18 and 34 years gave a similar answer. The vast majority of 76 percent of all young Hungarians at the end of the twentieth century regarded their family, their work or their peers more important than politics. 158 Christian W. Haerpfer Table 6.S Importance of politics, 1999 Important (% ) Not important (%) Total 27 73 Gender Women 29 71 Men 25 75 Age 18-34 years 24 76 35-54 years 25 75 55-90 years 43 66 Education Primary 23 77 Secondary 27 73 Tertiary 32 68 Town size 100-9,999 24 76 10,000-99,999 25 75 100,000-2,000,000 32 68 Subjective class Lower class 18 82 Working class 25 75 Middle class 29 71 Upper class 32 68 We also found a positive relation between higher levels of formal education on the one hand and the respondents' evaluation of the importance of politics. The higher the respondent's level of formal education, the more important is politics for her or his life. Approximately one-third of university graduates {32 percent) are convinced that politics is important for their lives. In the category of respondents who received secondary education, we found 27 percent who deem politics as important in their life, while only 23 percent of respondents who received primary education report that politics is important for them. There was also a linear relationship between size of local community and the importance of politics. The importance of politics in the respondent's life grows with the size of the local community. In cities with more than 100,000 inhabitants, including Budapest, the capital city with nearly two million residents, one-third of the population (32 percent) indicated that politics is important for their own lives. Another well-pronounced relationship was found with subjective class membership. The higher the self-ascribed position of a Hungarian citizen on the social ladder, the more important he or she thinks politics to be in her or his life (18 percent of lower class respondents as opposed to 32 percent upper class respondents). "Interest in politics" was selected as an additional indicator of political Hungary 159 motivation (Table 6.9). With regard to interest in politics, the Hungarian public is divided in two groups of equal size: those who are interested in politics (50 percent) and those who are not (50 percent). Significantly more men than women are interested in politics (55 percent versus 46 percent). With regard to age, there is no linear relation but nevertheless a pattern. The middle-aged generation and the elderly are more interested in politics than the younger generation. We found the highest levels of interest in politics in the group of respondents aged 35 to 54 years (56 percent). The elderly are representative for the overall distribution of political interest: 50 percent of older Hungarians (55 years and above) indicate an interest in politics, the other half does not. A lineai' relationship is found for the respondents' level of formal educa tion and their reported level of interest in politics. That is, the higher the level of education, the greater the interest in politics. Among those who received primary education only, 38 percent indicate that they are interested or very interested. This proportion increases to 49 percent in the group of those who received secondary education. In the group of respondents with the highest level of formal education, the level of respondents interested in politics increased to 62 percent. With regard to the urban-rural divide, we found a distinction between rural villages and very small towns on the one Table 6.9 Interest in politics, 1999 Interested (%) Not interested (%) Total 50 50 Gender Women 46 54 Men 55 45 Age 18-34 years 44 56 35-54 years 56 44 55-90 years 50 50 Education Primary 38 62 Secondary 49 51 Tertiary 62 38 Town size 100-9,999 45 55 10,000-99,999 52 48 100,000-2,000,000 52 48 Subjective class Lower class 29 71 Working class 44 56 Middle class 54 46 Upper class 63 37 J 60 Christian W. Haerpfer hand (45 percent interested)., and somewhat larger towns and cities on the other (52 percent interested). Subjective social class also had a strong impact upon interest in politics. The level of political interest increases with higher social status. Among the lower class, only 29 percent indicate interest in politics, the other 71 percent are not interested at all. The proportion of politically interested citizens goes up to 44 percent among members of die working class; in the middle class it grows to 54 percent. With 63 percent, political interest reaches its highest level in the upper class. Non-institutionalized political participation In this section, we first discuss the readiness to participate in a lawful demonstration because we use this single variable in our multivariate models to explain support of democracy. Second, we describe results generated by an index of non-institutionalized political participation (which includes "signing a petition," "joining in boycotts" and attending lawful demonstrations"). With respect to participation in lawful demonstrations, Hungarian citizens are divided: 50 percent would never participate in a political demonstration, whereas the other half might participate or had already Hungary 161 participated in such demonstrations. Actual participation in the period of 1989 to 1999 is reported by 9 percent of citizens, and 41 percent say that they might do so if there were a political issue which they believe should be addressed. Respondents who have already demonstrated are found primarily among the middle-aged (35 to 54 years), among respondents who received a high level of formal education, in cities and among the middle class. Potential participants were identified primarily among men, the young (18 to 34 years), respondents with a higher level of formal education, in cities and among the middle class. Participation in demonstrations is not very popular among 55 percent of women, the elderly (55 years and older), among those who only received a low level of formal education, in rural villages and among members of the lower and the working class. We also measured the respondents' readiness to participate or their actual participation in three different types of protest behavior (signing a petition, joining in boycotts and attending a lawful demonstration). The information has been combined in an index of protest behavior (Table 6.11). A total of 15 percent of respondents are represented at the highest level of protest behavior. These very active citizens are found more often among men (18 percent) than among women (13 percent). A medium level of protest behavior is reported by 39 percent of the respondents. Table 6.10 Participation in lawful demonstrations, 1999 Table 6,11 Index of protest behavior, 1999 Have done (%) Might do (%) Would never do (%) High<%) Medium (%) Low(%) Total 9 41 50 Total 15 39 46 Gender Gendei Women 9 37 55 Women 13 39 48 Men 10 46 44 Men 18 39 43 Age Age 18-34 years 8 46 46 18-34 years 17 39 44 35-54 years 13 43 44 35-54 years 22 40 39 55-90 years 8 35 58 55-90 years 8 39 53 Education Education Primary 4 31 66 Primary 4 35 61 Secondary 6 47 47 Secondary 13 39 48 Tertiary 17 48 35 Tertiary 29 42 29 Town size Tmim size 100-9,999 7 35 58 100-9,999 9 39 53 10,000-99,999 11 43 47 10,000-99,999 19 37 44 ] 00,000-2,000,000 10 45 45 100,000-2,000,000 18 41 41 Subjective class 76 Subjective class Lower class 2 22 Lower class 2 33 64 Working class 7 39 54 Working class 8 41 51 Middle class 10 48 42 Middle class 22 36 43 Upper class 18 32 50 Upper (kiss 22 50 28 J 62 Christian W. Haerpfer Combined, these two groups of protesters constitute a politically active majority of 54 percent of the population. The remaining 46 percent are not active and it would be difficult to mobilize them for political action. These passive citizens are over-represented among women (48 percent, while men are 43 percent). There is an expressed relation between age and the inclination to protest. The politically most active age group is the cohort aged 35 to 54 years: 22 percent indicate a high level of readiness for non-institutionalized political action, or have already participated that way. Among the young, 17 percent are ready to participate in all three types of political protest. Finally, 53 percent of the elderly indicate a very low level of readiness for political action. We also found a clear-cut correlation between the formal level of education and the readiness to participate in political action. Participation levels rise with higher levels of formal education. Whereas only 4 percent of Hungarians who received a low level of formal education indicate a high potential for participation in protest behavior, this share increases to 29 percent for respondents who received a high level of formal education. Participation in protest behavior or the willingness to participate in it is primarily found in towns and cities, and only to a lower extent in the countryside. Approximately one-fifth of all urban dwellers indicate participation or the willingness to participate in political action. In the countryside, we found only one-tenth of citizens who can be regarded as potential participants in protest behavior. The majority of Hungarian villagers (53 percent) are not ready to participate in protests. The readiness to express demands by means of political protest appears to be a characteristic of the middle classes. Of those respondents who subjectively classify themselves as middle class, 22 percent indicate that they are ready for this type of political action or to have used it already. Among respondents who describe themselves as members of the working class, we identified only 8 percent who are willing to participate in a similar manner. Selected correlates of the democracy-autocracy index discussed above will now be included in a multivariate regression model as independent variables to test their joint explanatory power. The two indicators are "political discussion with friends" and the "readiness to participate in lawful demonstrations." Results are presented in Table 6.12. Together, the Table 6.12 Impact of political motivation and political participation on support of democracy (model B; OLS multiple regression; dependent variable: democracy-autocracy index) Beta Significance X<1 Political discussions with friends 0.10 0.00 _ X5 Participation in demonstrations 0.15 0.00 - explained variance by model B - - 0.08 Hungary 163 two independent variables explain 8 percent of the variance in the democracy-autocracy scale. They both contribute almost equally to this result, with political motivation (political discussion with friends, beta-coefficient = 0.19) having a little bit the upper hand as compared to political protest (attending a lawful demonstration, beta coefficient = 0.15). Impact of education on support of democracy In post-communist societies, the proportion of "democrats" is highest among respondents who received a high level of formal education. This section of the analysis, again, makes use of results of the New Democracv surveys and relies on Haerpfer's index of democracy. Thus, we are in a position to show the development of support for democracy over time (1994, 1996, 1998) for the strategically important social group of citizens with higher education. This is the group from which future political and social elites are selected, and in this respect it is very important to know whether or not it is in support of democracy as a form of government. In central Europe, in 1998, on average 79 percent of Hungarian citizens who had obtained or were in the process of earning an academic degree supported democracy as a regime type (as measured by the Haerpfer democracy index). For many reasons, the political orientations of this social group can be considered crucial for the persistence of the new democratic regimes. They should be the social backbone of the new democracies. In the central European countries, the strength of citizens supporting democracy increased from 74 percent in 1994 to 79 percent in 1998 (Table 6.13). Hungary is the country with the largest share of democrats among the higher educated (76 percent in 1994 and 88 percent in 1998). The expansion of support for democracy was most dramatic in Poland: in 1994, only 65 percent of all Polish graduates supported the young Polish democracy, while in 1998 their share increased to 86 percent. The Czech pattern was quite different from the pattern in Hungary and Poland. Of all Czech graduates, 95 percent supported democracy in 1994; however, developments in Czech party politics and government apparendy led to a decrease in 1998 (80 percent). The process of democratization in Slovak society was much slower than in Hungary, Poland and the Czech Republic. It also happened at a somewhat lower general level. In 1994, only 58 percent of Slovaks who achieved higher formal education supported democracy as a regime type. Their share increased to 72 in 1998 at the end of the era of Prime Minister Meciar. In Slovenia, we find a similar pattern as in the Czech Republic, a shift from democratic euphoria to greater "realism": in 1998, only 71 percent of the respondents who received tertiary education in Slovenia supported democracy, which puts Slovenia in the same group with Slovakia, well behind Hungary, Poland and the Czech Republic. Thus, Hungarian and Polish societies belong to 164 Christian W. Haerpfer Table 6.13 Education and support of democracy: percentage of democrats among respondents with tertiary education, central and eastern Europe (Haerpfer democracy index; New Democracy Barometer surveys), 1994-8 1994 (%) 1996 (%) 1998 (%) Change 1994-8 Central Europe 74 76 79 5 1. Hungary 76 65 88 12 2. Poland 65 75 86 21 3. Czech Republic 95 88 80 -15 4. Slovakia 58 72 72 14 5. Slovenia - 80 71 -9 South-eastern Europe 67 60 62 -5 1. Romania 79 69 77 -2 2. Bulgaria 61 60 74 13 3. Croatia 62 50 56 -6 4. FRY - - 41 - Northern Europe 42 44 - 2 1. Estonia 60 62 - 2 2. Lithuania 38 40 - 2 3. Latvia 27 29 - 2 Eastern Europe 31 20 37 6 1. Belarus 32 21 47 15 2. Ukraine 36 19 26 -10 3. Russia 25 - - - Sources: NDB 3 (1994), NDB 4 (1996), NDB 5 (1998), New Balüc Barometer 1994, New Baltic Barometer 1996. the class of "mature democracies," supported by almost 90 percent of democrats. The Czech Republic, Slovakia and Slovenia fulfill the criteria of a consolidating democracy (60 percent to 80 percent democrats among the higher educated), while the Czech Republic seems to have crossed the threshold of a mature democracy in that social group (70-80 percent). In south-eastern Europe, the level of support for democracy of respondents with a higher level of formal education is significantly lower than in central Europe. Only 67 percent of the south-east European graduates and students qualify as democrats in 1994 and their share decreased to 62 percent in 1998. In south-eastern Europe, we are able to identify two different patterns. The first one was characteristic for countries neighboring the Black Sea and the second for the republics of the former Yugoslavia. Romania and Bulgaria qualify for the category of consolidating democracy, with a share of democrats well above 70 percent. In Romania, the percentage of democrats among Romanian graduates decreased from 79 percent in 1994 to 77 percent in 1998. The strong relief felt by the Romanian public after the end of the old Ceauscescu regime probably caused the rather high level of support in 1994. In Bulgaria, we observe an Hungary 165 increase of support for democracy by respondents who received a high level of formal education, from 61 percent in 1994 to 74 percent in 1998. Thus, Bulgaria and Romania can be labeled "consolidating democracies" with regard to respondents in the sector of tertiary levels of formal education. The pattern is different in Croatia and Serbia/Montenegro. In Croatia, the share of democrats in the group of graduates and university students decreased from 62 percent in 1994 to 56 percent in 1998. In the Federal Republic of Yugoslavia, only 41 percent of citizens who received tertiary education supported democracy. This low level of support for democracy in the group of respondents with a high level of formal education in Croatia and Serbia/Montenegro is probably caused by the earlier authoritarian regimes of Milosevic and Tudjman and the experience of war. These experiences apparently made it difficult even for those groups who are most likely to embrace democratic principles to develop democratic attitudes. Croatian society fulfils the criteria of an emerging democracy (40 percent up to 60 percent of democrats) with regard to citizens who received terdary levels of formal education. Serbia, however, is just above the threshold of an emerging democracy. In the Baltic countries, the level of democratization among respondents with a high level of formal education is much lower than in central and south-eastern Europe. Only 42 percent of Baltic citizens in this social group qualify as democrats in 1994. Their proportion increased only slightly in 1996 (44 percent). A clear majority of graduates support democracy in Estonia (62 percent). At 40 percent, the share of democrats is much lower in Lithuania, and in Latvia we find only 29 percent of respondents with a high level of formal education who qualify as democrats. In eastern Europe, the share of respondents who received tertiary education and fit the category of democrats increased from 31 percent in 1994 to 37 percent in 1998. Surprisingly, we find an increasing level of public support for democracy in Belarus. In 1994, only 32 percent of respondents with a higher level of formal education qualified as democrats. However, their share increased to 47 percent in 1998 - despite the fact that during these years President Lukashenko's regime turned more and more into a dictatorship. Thus, Belarus seems to be the interesting case. The society is becoming more democratic at the grass-roots, while the government loses more and more of its already small stock of democratic characteristics. An opposite trend could be observed in Ukraine, where we find a share of 36 percent democrats in 1994 among the respondents with a high level of formal education. Their share decreased to 26 percent in 1998. In Russia - on the low end - only a minority of 25 percent of respondents with a tertiary level of formal education could be characterized as democrats in 1998. 166 Christian W. Haerpfer Impact of civil society on support of democracy The degree to which civil society is realized in a given country can be measured in many ways. One highly consensual method of measurement is to consider membership and activity in voluntary associations. We will also follow this measurement strategy. One third of Hungarian respondents are active members of at least one voluntary association (Table 6.14). Volunteering for active service is more prevalent among men (35 percent) than women (29 percent). The most active age group participating in voluntary associations are middle-aged respondents (25 to 54 years of age; 35 percent). Young people (31 percent) are slightly more active than the elderly (29 percent). The group most active in voluntary associations are respondents with a high level of formal educadon (44 percent). One-quarter of those who completed primary or secondary education are active members. The urban-rural divide has apparently no impact on the level of membership and activity in voluntary organizations. We did find an influence of subjective social class. The higher the (self-ascribed) social status, the more active is a respondent. Only 20 percent who reported belonging to the lower class are active members in voluntary associations. This share increases to 25 percent if a respondent says he is a member of the working Table 6.14 Membership in voluntary organizations, 1999 Active members (%) Not active (%) Total 32 68 Gender Women 29 71 Men 35 65 Age 18-34 years 31 69 25-54 years 35 65 55-90 years 29 71 Education Primary 25 75 Secondary 26 74 Tertiary 44 56 -,. i J own sue 100-9,999 30 70 10,000-99,999 32 68 100,000-2,000,000 32 68 Subjective class Lower class 20 80 Working class 25 75 Middle class 36 64 Upper class 40 60 Hungary 167 class. More than one-third of the middle class is active in the third sector and its organizations (36 percent). The most active social class, however, is the upper class, with a reported 40 percent of active members in voluntary associations. Law-abidingness. is one of the most important characteristics of a mature democratic citizen. This is particularly important precondition to ensure the rule of law in a new democracy. In Table 6.15, we present the level of the respondents' subjective law-abidingness as an index. This index combines the acceptance of three patterns of behavior: "not to claim government benefits without entitlement" (Item 1); to "pay a ticket using public transport" (Item 2); to "pay all your taxes" (Item 3). The overall result based on this index indicates that two-thirds of the Hungarian population accept the rule of law, while one-third does not - to a varying degree. Women are slightly more law-abiding (68 percent) than men (66 percent). More than 70 percent accept the rule of law in the middle-aged generation (73 percent) and among the elderly (72 percent). The young are less law-abiding. Only 56 percent of young Hungarians accept the rule of law in full; 36 percent report violating the rule of law now and then, and 8 percent do not accept the rule of law - as defined here - at all. Apparently there is a demand for educational activities that Table 6.15 Law-abidingness, 1999 High(%) Medium (%) Low(%) Total 67 28 5 Gender Women 68 27 5 Men 66 29 5 Age 18-34 years 56 36 8 35-54 years 73 24 3 55-90 years 72 24 4 Education Primary 68 26 6 Secondary 69 26 5 Tertiary 65 30 5 Town she 100-9,999 68 28 4 10,000-99,999 70 25 5 100,000-2,000,000 64 30 7 Subjective class Lower class 62 27 11 Working class 69 25 5 Middle class 69 27 4 Upper class 62 34 4 168 Christian W. Haerpfer target Hungarian youth. Level of formal education has no influence on the acceptance of the rule of law. Results regarding the urban-rural divide are interesting. The rule of law is accepted most in villages (68 percent) and small towns (70 percent). In the big cities and in the capital, Budapest, a lower share of respondents reports abiding to the new laws and regulations (65 percent). More than one-third of the urban population (37 percent) are either not prepared to follow the law at all, or they try to avoid some of it sometimes. The strongest pillars of the rule of law in Hungary are the middle class (69 percent) on the one hand, and the working class (69 percent) on the other. The lower classes (62 percent), as well as the upper classes (62 percent), are less law-abiding than the middle and the working classes. In the latter part of this chapter, civil society has been operationalized as the level of citizens' participation in voluntary organizations on the one hand, and their level of law-abidingness on the other (Table 6.16). Relating these two specific democratic attitudes to the democracy-autocracy index we learn that - in relative terms - the most important predictor of the two indicators is the degree to which citizens participate in civil society (membership in voluntary associations; beta-coefficient = 0.14), followed by law-abidingness (beta-coefficient = 0.12). Democratic consolidation in Hungary - conclusions The last step of this analysis is to test a multivariate regression model of the correlates of support of democracy in Hungary in the year 1999 (Table 6.17). This regression model combines elements of all the dimensions discussed in the previous sections. Hungary can be characterized as a consolidated democracy as far as public support for democracy as an ideal form of government is concerned: 82 percent of Hungarian citizens fit the categories of either strong or weak democrats. We distinguish four groups of predictors for the level of democratic support: political principles (dimension A), political motivation and participation (dimension B), the quality of civil society (dimension C) and the level of formal education (dimension D). The variance explained by the regression amounts to 19 percent, almost one-fifth. Level of formal education is a powerful predictor of Table 6.16 Impact of civil society support of democracy (model C; OLS multiple regression; dependent variable: democracy-autocracy index) Civil society Beta Significance K2 X7 Member in associations 0.14 0.00 X8 Law-abidingness 0.12 0.00 Explained variance by Model D - — 0.03 Hungary 169 Table 6.17 A multivariate regression model to explain support of democracy in Hungary, 1999 (OLS multiple regression; dependent variable (Y): democracy—autocracy index) Dimensions Beta Significance A Political principles XI Tolerance 0.18 0.00 X2 Illegitimacy of political violence 0.15 0.00 X3 Solidarity with poor 0.10 0.00 B Political motivation and participation X4 Political motivation 0.13 0.00 X5 Protest behavior 0.09 0.01 C Civil society X6 Law-abidingness 0.12 0.00 X7 Membership in voluntary organizations 0.08 0.02 D Social structure X8 Level of formal education 0.15 0.00 Variance explained 0.19 support for democracy (beta-coefficient = 0.15). Other socio-economic variables such as age, urban-rural differences or gender are unimportant after controlling for the overpowering effect of education. Law-abidingness (beta-coefficient = 0.12) and civic participation (membership in voluntary organizations; beta-coefficient = 0.08) also contribute in a significant fashion. Two-thirds of the Hungarian respondents believe in the rule of law, not just in the abstract but as a guiding principle in everyday life. One-third of the populace are active in or members of voluntary organizations. This is considered to be an important prerequisite for a pluralist, representative democracy (Putnam 1993a, 1993b). Political motivation (beta coefficient = 0.13) and protest behavior (beta-coefficient = 0.09) contribute to explanations on a similar level. One-tenth of the respondents already participated in political demonstrations and another 40 percent are prepared to participate. Regarding other modes of protest behavior, we found support (54 percent) and rejection (46 percent) almost equally distributed (summary index combining signing a petition, joining in boycotts and attending a lawful demonstration). These findings indicate that democratic citizens in Hungary do not regard democracy as a spectator game in which one has to vote once every four years. Rather, democratic principles, values and activities have entered the political life of Hungarian citizens in a number of other ways. Hence, there is reason to believe the "strong democrats" particularly have embraced the concept of a participatory democracy. The most important effects on support for democracy and rejection of autocracy, however, were related to basic democratic principles such as tolerance (beta-cor-fTicient = 0.18), the illegitimacy ol political violence 170 Christian W. Haerpfer (beta-coefficient = 0.15) and solidarity with the poor (beta-coefficient = 0.10). These democratic principles have also been found to enjoy an even greater support among the Hungarian political elite (Linz and Stepan 1996; Tokes 1997; Agh 1998; Derleth 2000). Apparently there is a congruence in these value orientations between the Hungarian public and the Hungarian elite. This strengthens our argument that democracy is safely rooted in Hungarian society at the end of the twentieth century. Notes 1 My index of democracy consists of nine different items: Item 1: negative rating of communist political regime in the past; Item 2: positive rating of new democracy or current political regime; Item 3: optimism about the future of democratic parliaments; Item 4: support for democratic national parliament; Item 5: rejection of authoritarian leader as alternative to democracy; Item 6: rejection of a military regime as alternative to democracy; Item 7: rejection of monarchy as alternative to democracy; Item 8: rejection of return to communist political regime as alternative to democracy; Item 9: optimism about the future of democracy. 2 The democracy-autocracy index consists of two democracy items and two autocracy items. The two democracy items are: Item 1: positive rating of "Having a democratic system"; Item 2: positive rating of "Democracy may have problems but it is better than any other form of government." The two autocracy items are: Item 3: positive rating of "Having a strong leader who does not have to bother with parliament and elections"; Item 4: positive rating of "Having the army rule." The sum of the scores of the autocracy items (3, 4) is subtracted from the sum of the scores of the democracy items (I, 2), resulting in a scale which runs from +6 to —6. The groups used here are defined as follows: A. strong democrats <+5-+6); B. weak democrats ( + 1—1-4); C. undecided citizens (0); D. autocrats (— 1—— 6). Missing values are excluded list-wise. 3 The process of consolidation of democracy at the meso-level and at the macro^ level has to be measured with other methods. This is also true for the extent of democratization of elites and of institutions. However, I am focusing here at the extent of democratization on the micro-level of transforming societies. Hungary 17] Putnam, R.D. (1993a) Bowling Alone: the Collapse and Revival oj'American Community, New York: Simon & Schuster. Putnam, R.D. (1993b) Making Democracy Work: Civic Traditions in Modern. Italy, Princeton: Princeton University Press. Swain, N. (1993) "Hungary," in White, S., Blatt, J. and Lewis, P.G. (eds) Developments in East European Politics, Basingstoke, Hampshire: Macmillan. Tokes, R.L. (1996) Hungary's Negotiated Revolution: Economic Reforms, Social Change, and Political Succession, Cambridge: Cambridge University Press. Tokes, R.L. (1997) "Party Politics and Political Participation in Postcommunist Hungary," in Dawisha, K. and Parrot, B. (eds) Tlie Consolidation of Democracy in East-Central Europe, Cambridge: Cambridge University Press. References Agh, A. (1998) The Politics of Central Europe, London: Sage Publications. Bozoki, A. (1999) "Democracy in Hungary, 1990-1997," in Kaldor, M. and Vejvoda, I. (eds) Democratization in Central and Eastern Europe, London: Frances Pinter. Bruszt, L. (1990) "1989: the Negotiated Revolution in Hungary," Social Research 5,2. Derleth, W.J. (2000) The Transition in Central and Eastern European Politics, Upper Saddle River, New Jersey: Prentice Hall. Haerpfer, C.W. (2002) Democracy and Enlargement in Post-Communist Europe: the Demoralisation of the General Public in Eifleen Central and Eastern European Countries, 1991-1998, London: Routledge. Linz, j.J. and Stepan, A. (1996) Problems of Democratic Transition and Consolidation: Southern Europe, South America, mid Post-Communist Europe, Baltimore: |ohns Hopkins University Press.