Political Disaffection in 3 rd Wave Democracies in Europe Comparative Study Comparative Study on Political Disaffection among 3 rd Wave Democracies of Europe based on the data legally provided by European Social Survey 2002/2003. Aleksander Kucel Department of Political and Social Sciences Pompeu Fabra University Barcelona, August 2004
Abstract The following paper discusses the concept of political disaffection in context of 3 rd wave democracies in Europe. As 3 rd wave democracies are accounted here Spain, Portugal, Greece, Italy, Poland, Czech Republic, Hungary and Slovenia. The group of 3 rd wave democracies is being divided into two subgroups, namely 3 rd Wave Matured Democracies and 3 rd Wave Young Democracies. Significant differences are being discussed and described. Mechanisms causing the differences among the discussed groups are proposed for further consideration. 2
List of Contents List of Contents 3 1. Introduction 4 2. Political support in Europe. Descriptive analysis. 6 3. Factor analysis of political variables16 4. Regressions on political disaffection 28 5. Conclusions 34 3
1. Introduction Political support is a multidimensional phenomenon. There are many different approaches to question of measuring and characterizing political support. Colloquial understanding of this concept is that people support the politicians that are at the given moment in power. This sort of view is by all means unclear. To order a bit the scenery let us say that concepts such as support for democracy and political disapproval are two distinct problems. They may be somehow interconnected but in principle they describe two different attitudinal surfaces of our understanding of politics. A structured, rigorous studying of political attitudes dates back to famous book of Almond and Verba, The civic culture. Political attitudes and democracy in five nations, published in 1963, in which political culture was treated as a black box in which we put everything that is nonobservable directly in political and economic life. Twelve years later Easton published a paper in which he proposed a new construction of studying political life based on concepts of specific and diffuse support. However a real blossoming of influential works on the question of political attitudes and political support we encounter in the 1990 s, with such classical already works as Barnes Politics and Culture and methodological article of Reisinger The renaissance of a rubric: political culture as concept and as theory. Mentioning those works certainly is just meant to provide some taste of the literature in the topic of political support. It is obvious that the state of literature in this discipline is much broader than the one jotted above. As Torcal proposes we can divide the study of attitudinal change and formation in politics into two main groups: the Culturalist model established by Lerner and the Rational Culturalist model. The culturalist model claims that attitudes change slowly and are heavily based in the history, since socialization is a slowmoving historical process. Rational culturalists claim, on the contrary, that culture and attitudes can change quickly as a result of political and/or economic events, institutional changes and through process of rational adaptation and adult learning. After rational culturalists we claim here that political attitudes may change quickly and that people adapt their attitudes to the situation in which they have to live. However we do not answer the question what is the time scale for such changes. We do not attempt to answer what it means that some attitudes change quickly and some others rather more slowly. Therefore we operate here within the mainstream of rational political culturalism. Rational 4
political culturalists in our view claim that political science ought to benefit from psychology, sociology and anthropology as well as from economics. For us, two societies sharing all the same institutional and structural conditions can nevertheless differ in their political dynamics if their political cultures differed significantly, while at the same time we allow for moderately quick changes in those cultures through operation of various nonobservable factors. One of such factors in our opinion (and often decisive one) is comprised of expectations. What people expect the incumbents to do in the politics, what kind of functioning of political system they desire in our opinion shapes in the long as well as in the short run the political attitudes of societies. Therefore in our analysis we attempt to demonstrate the significant differences between modern societies in a given moment of time. If we manage successfully to demonstrate that European 3 rd wave democracies are not homogenous within their group and that the differences between them are statistically significant this may suggest that our path of research is correct because, in our view, only different expectations may diversify strongly enough political attitudes between societies. This does not exclude the space for operation of another mechanism mechanism of rational adaptation. Actually without this mechanism rationale of our claim ceases to exist. Therefore to demonstrate our point we have to assume that societies rationally adapt their political expectations in the medium and long run. Our proposition cannot by any means be considered a theory since we narrow our analysis only to provision of some possible explanations. Plausibility of them depends heavily on the state of art in the discipline of political support, which on the other hand evolves constantly. Our claims should rather be treated as suggestions of mechanisms ruling political support than any coherent independent theory. In the following sections we describe first the present situation about political support in discussed group of countries using standard descriptive statistics and tables of frequencies, later on we present factor analysis in order to create indexes from our variables that we use in regressions in the last section. Those regressions are meant to provide main support for our argument. In the last section of this paper we provide conclusions. In appendices we provide full results of statistical analyses. 5
2. Political support in Europe. Descriptive analysis. In this section we provide a broad view on political support in Europe. We discuss a group of 22 countries. The complete list countries contains: Austria, Belgium, Switzerland, Czech Republic, Germany, Denmark, Spain, Finland, France, United Kingdom, Greece, Hungary, Ireland, Israel 1, Italy, Luxembourg, Netherlands, Norway, Poland, Portugal, Sweden, Slovenia. One of the most important contemporary works discussing political support and specifically even democratic support is the book edited and coauthored by Pippa Norris Critical citizens. Global support for democratic governance. Norris in the chapter 11 proposes 6 hypotheses concerning political support and discusses them briefly in light of empirical research based on World Value Survey from two different periods of time. However she tends to mix democratic support with more general concept of political support. We assume in here that political support is a general concept concerning support for institutions in the country and support for political incumbents and their actions. We exclude however support for democracy from the concept of political support following Torcal in his work Political Disaffection and Democratization. History in New Democracies 2. We find it however very instructive to quote the hypotheses of Norris concerning political support for their intuitive character and clarity. These are: H1: Winners have higher confidence in governing institutions than losers. H2: The existence of widespread political rights and civil liberties should be associated with higher levels of public confidence in the regime. H3: Parliamentary systems, where all parties continue to have a stake in the policymaking process, should generate greater system support than winnertakeall presidential systems. H4: Countries with twoparty and moderate multiparty systems should have the highest levels of institutional support. H5: Federal systems should produce higher levels of institutional support than unitary states. H6: Majoritarian systems should produce less institutional support than proportional systems. 1 Israel is not geographically placed in Europe but frequently in articles and discussions it is treated as a European country, thus such little geographical inconsistence shall be omitted. 2 It is worth noticing that there are more articles clarifying the concept of political support. One of the excellent examples is the work of Ruth Lane Political culture: residual category or general theory. 6
We do not discuss those hypotheses at length because we assume that they refer rather to the countries that we call later Old Democracies which for structural reasons remain outside our scope in this work. However the hypotheses of Norris can provide a very fruitful source for reflection on the further evolution of what we call 3 rd Wave Young Democracies especially while confronted with contemporary development of political life in our control group called 3 rd Wave Matured Democracies. The book edited by Norris provides us also with another very useful concept, namely democratic crisis. It is nowadays, a very fashionable to speak in parliaments of various countries about the crisis of democracy. In fact, for decades political oppositions in democratic countries used the argument of democratic crisis against their opponentincumbents. Is democracy really permanently in crisis? The answer for that is provided in the title to the unpublished yet book of Torcal called Disaffected but democrats. In fact what we observe presently is rather an example of speed up in democratic evolution rather than a crisis of democracy. As we will demonstrate in this section many societies are unhappy with how democracy works in their countries but on the other hand they want to protect democracy from parties that would threaten this political regime. This animosity in the public opinion is only spurious because in fact we have to distinguish between support for democracy as for the best political regime and satisfaction of the democratic performance in the country. As Mishler and Rose demonstrate in chapter 4 of Norris, book the dynamics of democratic support in PostCommunist Europe shows that democracy becomes year by year more appreciated system. This positive trajectory of democratic support does not however have to go in pair with satisfaction of democratic functioning in those countries. Having in mind all the assumptions made before we can now proceed to the analysis of frequencies of some basic political indicators among our sample of countries. All the data used in this work come from European Social Survey provided free legally online at www.europeansocialsurvey.org. Our variable choice is a limited version of the original database system of ESS. All the calculations were performed with use of ESSdesigned dweight variable for weighting cases in the database. The first variable we present on the sample of 22 mentioned earlier countries is Satisfaction with the government. We computed it as a percentage of population of each country with additional use of other weighting ESSdesigned variable called pweight. 7
Satisfaction with the government (%) Country Unsatisfied Satisfied Very satisfied Total Austria 46 45 9 100 Belgium 19 71 10 100 Switzerland 15 69 16 100 Germany 57 41 2 100 Denmark 16 56 28 100 Finland 13 71 17 100 France 36 57 7 100 United Kingdom 35 57 7 100 Ireland 39 54 7 100 Israel 48 42 10 100 Luxembourg 9 58 33 100 Netherlands 35 61 4 100 Norway 34 61 4 100 Sweden 20 69 12 100 Table 1. Satisfaction with the government in Old Democracies. Source: own computations based on ESS data. Satisfaction with the government (%) Country Unsatisfied Satisfied Very satisfied Total Spain 37 55 7 100 Greece 41 49 10 100 Italy 40 55 5 100 Portugal 50 43 6 100 Table 2. Satisfaction with the government in 3 rd Wave Matured Democracies. Source: own computations based on ESS data. Satisfaction with the government (%) Country Unsatisfied Satisfied Very satisfied Total Czech Republic 46 50 4 100 Hungary 29 56 15 100 Poland 60 37 3 100 Slovenia 38 53 9 100 Table 3. Satisfaction with the government in 3 rd Wave Young Democracies. Source: own computations based on ESS data. The tables 13 presented above show percentages of populations of respective countries satisfied with the government of their country. The average dissatisfaction with the government in Old Democracies group is 30% while already for the group of 3 rd Wave Matured Democracies the respective average percentage of dissatisfaction amounts 42%. Not much different is the dissatisfaction level in the third group of countries called 3 rd Wave Young Democracies where the figure is 43%. Much greater differences we observe in the 8
levels of medium satisfaction (in tables indicated as Satisfied ) where the percentages are respectively: 58%, 51% and 49%. Among the old democracies the greatest dissatisfaction exhibit Germany 57% and Israel 48%. In the group of 3 rd wave matured democracies the greatest dissatisfaction with the government we observe in Portugal (50%) and the lowest in Spain (37%). In the last group absolutely outstandingly great dissatisfaction level is in Poland with the percentage reaching 60% of respondents. In the group of Very satisfied respondents we observe that the greatest percentages characterize Luxembourg (33%) and Denmark (28%) among the countries called old democracies, with Greece (10%) leading in the 3 rd wave matured democracies group and Hungary (15%) in the 3 rd wave young democracies group. The next set of tables (tables 46) shows average index of trust in politics. The index was created from two variables: trust in politicians (trstplt) and trust in country s parliament (trstprl). The index is a simple arithmetic average of these two variables. Index of average trust in politics (%) Country Distrust Trust Strong trust Total Austria 50 45 5 100 Belgium 36 61 3 100 Switzerland 22 72 6 100 Germany 54 45 1 100 Denmark 17 72 10 100 Finland 28 66 5 100 France 50 48 2 100 United Kingdom 47 50 2 100 Ireland 50 47 4 100 Israel 54 42 4 100 Luxembourg 27 66 8 100 Netherlands 27 71 2 100 Norway 28 67 5 100 Sweden 28 65 7 100 Table 4. Index of average trust in politics in Old Democracies. Source: own computations based on ESS data. Index of average trust in politics (%) Country Distrust Trust Strong trust Total Spain 51 46 3 100 Greece 52 42 5 100 Italy 48 50 2 100 Portugal 61 39 1 100 Table 5. Index of average trust in politics in 3 rd Wave Matured Democracies. Source: own computations based on ESS data. 9
Index of average trust in politics (%) Country Distrust Trust Strong trust Total Czech Republic 64 35 1 100 Hungary 46 49 5 100 Poland 71 28 1 100 Slovenia 63 35 2 100 Table 6. Index of average trust in politics in 3 rd Wave Young Democracies. Source: own computations based on ESS data. Similar pattern as with the political satisfaction is observed in terms of political trust in our 3 groups of countries. The greatest trust exhibit old democracies with average of strong trust in politics reaching 5% and the lowest 3 rd wave young democracies with the level of strong trust equal to only 2% of respondents. Respective numbers for those who trust moderately in politics are: 58% for Old democracies group, 44% for 3 rd Wave Matured Democracies and 37% for 3 rd Wave Young Democracies. The strongest trust, again similarly like in case of political satisfaction, is noted in Denmark 10% and in Luxembourg 8%. Respective leaders in the two remaining groups in strong trust in politics are Greece with 5% and Hungary 5%. Polish respondents again manifested the highest distrust in politics reaching the level of 71%. Germany and Israel distrust the politics in 54% and Portuguese respondents in 61%. Interestingly however Greece being the leader in the group of 3 rd wave matured democracies in terms of strong trust exhibits at the same time moderately high level of distrust amounting to 52% of respondents. Another curious observation comes from the column 3 (Trust) of table 4 where Switzerland and Netherlands declare more than 70% trust in politics reaching the level of Denmark and Luxembourg leaders in this category. Since most of the influential works on the topic of political support and democratization take into account satisfaction of the citizens with the state of country s economy, we decided to follow this mainstream and present respective tables frequencies for our groups of countries. The following tables (79) present mentioned data. 10
Satisfaction with the present state of economy (%) Country Unsatisfied Satisfied Very Satisfied Total Austria 22 61 17 100 Belgium 19 68 13 100 Switzerland 24 65 11 100 Germany 62 36 2 100 Denmark 6 49 45 100 Finland 9 66 25 100 France 49 48 3 100 United Kingdom 21 66 13 100 Ireland 32 59 9 100 Israel 67 30 3 100 Luxembourg 6 56 38 100 Netherlands 16 75 10 100 Norway 15 57 29 100 Sweden 22 67 11 100 Table 7. Satisfaction with the present state of country s economy in Old Democracies. Source: own computations based on ESS data. Satisfaction with the present state of economy (%) Country Unsatisfied Satisfied Very Satisfied Total Spain 25 65 9 100 Greece 51 42 7 100 Italy 35 62 3 100 Portugal 61 38 1 100 Table 8. Satisfaction with the present state of country s economy in 3 rd Source: own computations based on ESS data. Wave Matured Democracies. Satisfaction with the present state of economy (%) Country Unsatisfied Satisfied Very Satisfied Total Czech Republic 49 48 3 100 Hungary 39 57 4 100 Poland 68 31 1 100 Slovenia 41 52 7 100 Table 9. Satisfaction with the present state of country s economy in 3 rd Wave Young Democracies. Source: own computations based on ESS data. This variable follows the pattern wellestablished by the two previous variables: political satisfaction and political trust. In the group of old democracies the average level of dissatisfaction with the present state of economy in the country (among the group countries) is 27%, while the same frequency for 3 rd wave matured democracies amounts to 43% and grows to 49% for 3 rd wave young democracies. Again the most satisfied with the national economy are the citizens of Denmark (45%) and Luxembourg (38%). In the group of 3 rd wave matured democracies the most dissatisfied with the economic performance of the country are 11
Portuguese people. Some surprise brings Greece where despite a great trust in politics and major dose of satisfaction with the political functioning people exhibit 51% of dissatisfaction with economic performance of their country, while at the same time the number of Greek citizens very satisfied with present state of country s economy equals 7% almost as much as in Spain which is the leader in the group in terms of satisfaction with the state of economy in the country. In the third group (3 rd wave young democracies) again the highest amount of citizens dissatisfied with economic functioning of the country is noted in Poland with 68% of respondents being dissatisfied and only 1% (the lowest in all the countries) very satisfied. Interesting results provides decomposition of the index of political trust into two variables from which it was initially created: trust in parliament (trstprl) and trust in politicians (trstplt). Tables 1012 provide the results of the decomposition of the index of trust. The general pattern observed in the tables below suggests that people in old and 3 rd wave matured democracies trust more their parliaments than politicians. It may be the result of poor performance of political incumbents combined with deeper trust in parliament as a stable and reliable institution. The only significant difference exhibits the medium level of trust group of respondents in some countries where trust in politicians is not lower than trust in parliaments. However distrust and strong trust follow previous pattern, with Denmark keeping its leading position in strong trust (30% in parliament and 14% in politicians) together with, this time, Sweden (29% of strong trust in parliament and 8% in politicians). It can be noticed that in old democracies the medium trust and satisfaction play much more important role than in 3 rd wave democracies in general. Let us see the following tables for more detailed look. 12
Country Variable Distrust Trust Strongly Trust Total Austria Trust in the parliament 24 57 18 100 Trust in the politicians 52 42 6 100 Belgium Trust in the parliament 23 66 11 100 Trust in the politicians 34 62 4 100 Switzerland Trust in the parliament 12 70 18 100 Trust in the politicians 20 72 8 100 Germany Trust in the parliament 34 57 10 100 Trust in the politicians 50 47 3 100 Denmark Trust in the parliament 11 59 30 100 Trust in the politicians 16 70 14 100 Finland Trust in the parliament 15 62 23 100 Trust in the politicians 28 66 7 100 France Trust in the parliament 31 60 8 100 Trust in the politicians 47 50 3 100 United Kingdom Trust in the parliament 31 58 11 100 Trust in the politicians 44 52 3 100 Ireland Trust in the parliament 35 52 12 100 Trust in the politicians 46 48 6 100 Israel Trust in the parliament 34 48 17 100 Trust in the politicians 55 40 5 100 Luxembourg Trust in the parliament 15 64 22 100 Trust in the politicians 27 63 10 100 Netherlands Trust in the parliament 18 73 9 100 Trust in the politicians 22 74 4 100 Norway Trust in the parliament 16 62 23 100 Trust in the politicians 27 67 6 100 Sweden Trust in the parliament 15 57 29 100 Trust in the politicians 28 64 8 100 Table 10. Trust in the country s parliament and politicians in Old Democracies. Source: own computations based on ESS data. Interestingly in the table above the wellknown pattern from previous tables vanishes. In fact there appears to be a group of Scandinavian countries (Sweden (64%), Norway (67%), Finland (66%), Denmark (70%)), where medium trust in politicians is greater than in the parliament. Netherlands and Luxembourg classify also close to this group with almost equal medium trust in parliaments and politicians. Similar decomposition for the group of 3 rd wave matured democracies does not however bring any revelation and keeps following previously established pattern from tables 19. In the columns 3 and 5 however the previously mentioned and earlier widely discussed pattern returns. The general conclusion from the above table can be that trust in parliament is much greater than in politicians (usually double) in the groups of citizens manifesting strong trust and at the same time distrust in parliament is much smaller than distrust in politicians in the group of people exhibiting general distrust in politics. 13
Country Variable Distrust Trust Strongly Trust Total Spain Trust in the parliament 27 61 12 100 Trust in the politicians 52 44 4 100 Greece Trust in the parliament 32 50 18 100 Trust in the politicians 53 40 8 100 Italy Trust in the parliament 24 66 10 100 Trust in the politicians 48 50 2 100 Portugal Trust in the parliament 31 61 7 100 Trust in the politicians 63 36 1 100 Table 11. Trust in the country s parliament and politicians in 3 rd Wave Matured Democracies. Source: own computations based on ESS data. Table 11 follows earlier conclusions and moreover it exhibits the same trends as table 5. Spain and Greece show the biggest trust in politicians and parliament in the column of strong trust while Italy leads in category of medium trust. The greatest distrust both in parliament and politicians is observed in Portugal as it was before in table 5. Country Variable Distrust Trust Strongly Trust Total Czech Republic Trust in the parliament 50 46 4 100 Trust in the politicians 57 41 2 100 Hungary Trust in the parliament 28 52 19 100 Trust in the politicians 44 49 7 100 Poland Trust in the parliament 52 44 4 100 Trust in the politicians 67 32 1 100 Slovenia Trust in the parliament 42 50 8 100 Trust in the politicians 59 38 3 100 Table 12. Trust in the country s parliament and politicians in 3 rd Wave Young Democracies. Source: own computations based on ESS data. Once again Poland classifies as a country with greatest distrust in parliament and politicians and lowest frequencies for strong trust. Hungary on the other hand proves to be outstanding in the group of 3 rd wave democracies and exhibits the greatest strong trust in the parliament and politicians in the whole group, comparably big as many of old democracies. 14
Satisfaction with functioning of democracy (%) Country Dissatisfied Satisfied Very satisfied Total Old Democracies Austria 21 52 27 100 Belgium 17 67 16 100 Switzerland 7 53 39 100 Germany 24 56 20 100 Denmark 4 41 55 100 Finland 8 61 30 100 France 28 56 16 100 United Kingdom 24 61 15 100 Ireland 27 57 17 100 Israel 29 52 19 100 Luxembourg 8 47 45 100 Netherlands 10 73 16 100 Norway 10 64 27 100 Sweden 12 58 30 100 3rd Wave Matured Democracies Spain 14 66 19 100 Greece 20 48 32 100 Italy 23 65 12 100 Portugal 28 64 7 100 3rd Wave Young Democracies Czech Republic 30 56 14 100 Hungary 28 58 14 100 Poland 40 52 7 100 Slovenia 36 55 9 100 Table 13. Satisfaction with functioning of democracy (frequencies countrywise). Source: own computations based on ESS data. With those tables we get a clear pattern that with some exceptions (like Hungary in table 12) there are two strongly different groups of democratic countries: old democracies and 3 rd wave democracies, however we can observe also significant differences within the group of 3 rd wave democracies which preliminarily validates our division of this group into two subgroups 3 rd wave matured and 3 rd wave young democracies. 15
3. Factor analysis of political variables In this section we perform factor analysis to figure out which variables change in the same way so that we can find which nonobservable factors cause that changes. Factor analysis helps especially when some variables can be grouped on the basis of their common changes that means that we can group some variables that simply change in the same fashion. In our case factor analysis is especially interesting since its results may shed new light on our perception of the variables we analyze in the political support context. Having those factors is crucial for further analysis because we will use them in section 4 to create indexes and combined variables that further we will use in linear regressions to provide evidence to our main claim. We performed our factor analysis for all three groups of countries that we specified earlier, that is for Old Democracies, 3 rd Wave Matured Democracies and 3 rd Wave Young Democracies. Before conducting actual factor analysis it is always recommended to perform KaiserMeyer Olkin (KMO) and Bartlett s tests which measure how sensible it is to perform factor analysis in a given data set. The KaiserMeyerOlkin Measure of Sampling Adequacy measure was high (the KMO statistic equal 0.838) and the significance level of the Bartlett s test also indicated sensibility of performing factor analysis in the group of Old Democracies. The table below contains the results of the tests. KMO and Bartlett's Test(a) KaiserMeyerOlkin Measure of Sampling Adequacy 0.838 Bartlett's Test of Sphericity Approx. Chi Square 83,768.877 a. Groups of countries: old, matured, young = Old democracy df 351 Sig. 0.000 Table 14. KMO and Bartlett s tests for Old Democracies. Source: own computations based on ESS data. The results of the factor analysis should be conclusive since the tests indicate that our ESS data qualifies for the analysis. The results of the analysis are presented below. We present the results for each group separately. Only the final results are being presented 3. Eventually after performing the analysis we decided to drop the group of old democracies and in the last section of this paper we compare in regressions only two groups of countries, namely 3 rd Wave Matured Democracies and 3 rd Wave Young Democracies. Therefore our results do not embrace the whole initial sample of countries but it does not diminish any theoretical 3 More information on the performed factor analysis can be found in Appendix A. 16
findings. We were initially interested in the countries called commonly 3 rd wave democracies and specifically the differences between defined earlier two subgroups within this group. Therefore we can without loss of generality drop the group of old democracies and keep our results conclusive. Below we present a principal component matrix which is used to filter out hidden factors that drive variance of our variables. In the table 15 (below) we bolded those numbers that show which nonobservable factor is comprised of which variables from our data set. Thus if a column 1 in table 15 contains 8 bolded numbers then those 8 variables, to which the numbers belong (we read that in a row), comprise the primary nonobservable factor. Interestingly such algorithm of constructing factors leads directly to production of 8 factors each of which is built from the variables in our data set. The meaning of each factor is arbitrary and is always subject to criticism for its biases or nonobjectivity. However in some cases it happens that factors are comprised of variables of some clearly chosen type, which can be interpreted without any difficulty and plausibility of such interpretation is selfenforcing. Obviously not always such situation will take place but most often the most important factors (like factor 1, factors 2 or at most 3) are easy to interpret in our study. Difficulty of interpretation depends also heavily on indepth understanding of the initial variables thus 2 nd part of our study, despite the fact that a bit lengthy proves to be very fruitful here, where we need to see through our variables to interpret nonobservable factors properly. In any case interpretation of the factors in the factor analysis is not always necessary. If factor analysis is conducted only as a additional tool for grouping variables before running regressions then interpretation of the factors does not come into scope. 17
TV watching, news/ politics/current affairs on average weekday 0.036 0.062 Radio listening, news/ politics/current affairs on average weekday Component 1 2 3 4 5 6 7 8 0.038 0.022 Newspaper reading, politics/current affairs on average weekday 0.053 0.237 0.021 0.278 0.632 0.037 0.029 0.059 0.641 0.055 0.056 0.003 0.642 0.026 0.053 0.081 0.008 0.055 0.022 Most people can be trusted or you can't be too careful 0.245 0.070 0.737 0.106 0.027 0.023 0.021 Most people try to take advantage of you, or try to be fair 0.165 0.054 0.794 Most of the time people helpful or mostly looking out for themselves 0.169 0.026 0.731 0.701 0.744 0.045 0.091 0.029 0.049 0.024 0.016 0.033 0.005 0.044 0.055 0.043 0.028 0.037 0.020 How interested in politics 0.039 0.096 0.012 0.359 0.042 0.058 Politics too complicated to understand 0.029 0.221 0.079 0.058 0.074 0.051 Making mind up about political issues 0.674 0.059 0.061 0.154 0.053 0.004 0.041 0.239 Politicians in general care what people like respondent think 0.654 0.201 0.120 0.063 0.007 0.029 0.218 0.219 Politicians interested in votes rather than peoples opinions 0.631 0.199 0.109 0.023 Trust in country's parliament Trust in the legal system Trust in politicians How satisfied with present state of economy in country 0.619 How satisfied with the national government 0.745 How satisfied with the way democracy works in country 0.708 The less government intervenes in economy, the better for country 0.169 0.074 0.780 0.139 0.125 0.049 0.016 0.002 0.585 0.103 0.174 0.091 0.824 0.128 0.149 0.249 0.195 0.010 0.036 Government should reduce differences in income levels 0.073 0.096 Employees need strong trade unions to protect work conditions/wages Ban political parties that wish overthrow democracy Important in life: politics 0.029 0.017 0.044 0.045 0.039 0.076 0.042 0.013 0.204 0.279 0.052 0.042 0.074 0.279 0.178 0.060 0.141 0.101 0.120 0.053 0.016 0.077 0.058 0.105 0.029 0.022 0.088 0.040 0.288 0.162 0.018 0.018 0.061 0.056 0.019 0.036 0.021 0.035 0.013 0.384 0.748 0.015 0.021 0.016 0.101 0.119 0.201 0.317 0.237 0.186 0.328 0.032 0.026 0.054 0.732 0.114 0.055 0.033 0.857 0.194 0.661 0.006 0.269 0.026 0.079 Employment status 0.019 0.008 0.088 0.029 0.760 0.007 Year of birth 0.038 0.682 0.044 0.123 0.317 0.081 Gender 0.045 0.015 0.045 0.281 0.003 0.024 Years of fulltime education completed 0.043 0.294 0.071 0.620 0.126 0.139 Feeling about household's income nowadays 0.186 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. 0.008 0.239 0.244 0.218 0.471 0.071 0.035 0.168 0.002 0.102 0.095 0.031 0.028 0.007 0.764 0.095 0.112 0.076 Table 15. Results of factor analysis for Old Democracies. Source: own computations based on ESS data. 18
For the group of Old Democracies we defined the following factors: Factor F1 (Trust in politics and satisfaction with it) Politicians in general care what people like respondent think Politicians interested in votes rather than peoples opinions Trust in country's parliament Trust in the legal system Trust in politicians How satisfied with present state of economy in country How satisfied with the national government How satisfied with the way democracy works in country Factor F2 (Interest in politics) How interested in politics Politics too complicated to understand Making mind up about political issues Important in life: politics Factor F3 (Social trust) Most people can be trusted or you can't be too careful Most people try to take advantage of you, or try to be fair Most of the time people helpful or mostly looking out for themselves Factor F4 (Personal characteristics) Employment status Year of birth Years of fulltime education completed Factor F5 (Media) TV watching, news/ politics/current affairs on average weekday Radio listening, news/ politics/current affairs on average weekday Newspaper reading, politics/current affairs on average weekday Factor F6 (Social coverage desires) Government should reduce differences in income levels Employees need strong trade unions to protect work conditions/wages 19
Factor F7 (Support for democracy as a system) Ban political parties that wish overthrow democracy Factor F8 (Gender) Gender The importance of each factor declines with the number of the factor. The most important for the group of Old Democracies are Trust in politics and satisfaction with it, Interest in politics and Social trust. The least important is gender. It may suggest that gender in old democracies is not a politically influential factor in the society. In Old Democracies group of countries what revealed to be mostly important was trust in politicians, politics and satisfaction coming from the functioning of democracy, country s economy, and national government. This factor as the most important can be understood that political support in Old Democracies depends very much on people s trust in politics and their satisfaction with how things go in the country. Second most important factor for the analyzed group of countries comprised of variables describing interest in politics. It seems very intuitive to say that people in Old Democracies if interested in politics may donate political support, otherwise they do not consider political issues as important things in their lives. Interestingly, social trust proves to be important for political support in Old Democracies too. One might risk a hypothesis that if people trust each other they will be more prone to rely their trust in politics too. Last but not least is to notice that media do not play very important role in formation of political support in Old Democracies. We shall observe the evolution of this factor in the analyses for 3 rd Wave Democracies especially for the Young subgroup. Below we present corresponding results of the factor analysis on the group of 3 rd Wave Matured Democracies. Again KMO and Bartlett s tests indicate correctness of our decision to run factor analysis on this sample. KMO and Bartlett's Test(a) KaiserMeyerOlkin Measure of Sampling Adequacy. 0.792 Bartlett's Test of Sphericity Approx. Chi Square 20,721.581 df 351 Sig. 0.000 a. Groups of countries: old, matured, young = 3rd wave matured democracy Table 16. KMO and Bartlett s tests for 3 rd Wave Matured Democracies. Source: own computations based on ESS data. 20
The results of factor analysis for the group of 3 rd Wave Matured Democracies indicate different set of factors and this is certainly understandable and desirable result. To remind we want to demonstrate that within the group of 3 rd wave democracies there are two subgroups that are significantly different between each other. However quietly we also desire that the both subgroups were significantly different from the Old Democracies group. Otherwise our results would be completely inconclusive. After demonstrating that factors for all 3 groups differ we will drop the old democracies from our analysis, since our quiet assumption of old democracies being significantly different from 3 rd wave democracies will be already supported by the factor analysis. In the analysis below, for the group of 3 rd Wave Matured Democracies we notice, that first three factors are to some extent similar to those observed in the group of Old Democracies but however they are already interestingly different. Political support in this group is mostly shaped by people s interest in politics and their education level. The more they are able to comprehend political issues the more interested in politics they are. Trust in politics comes into scope as second most important factor in the group of 3 rd Wave Matured Democracies which means that it lost on importance as we moved from Old Democracies. That is a first important difference between the analyzed groups. Another important observation is that political satisfaction took place of social trust of Old Democracies. Political satisfaction as the third most important factor for 3 rd Wave Matured Democracies group establishes another difference between this group and the Old Democracies group. Media factor did not change its position and remained 5 th factor on the scale of importance. As we move across groups of democracies in Europe we notice that different factors shape people s political attitudes. This preliminary observation is very plausible because it shows structural differences between societies of different democracies. 21
TV watching, news/ politics/current affairs on average weekday 0.112 0.091 Radio listening, news/ politics/current affairs on average weekday 0.134 Component 1 2 3 4 5 6 7 8 0.053 Newspaper reading, politics/current affairs on average weekday 0.351 0.076 0.033 0.016 0.092 0.046 0.599 0.003 0.038 0.125 0.085 0.521 0.224 0.085 0.562 Most people can be trusted or you can't be too careful 0.137 0.086 0.054 0.775 0.052 Most people try to take advantage of you, or try to be fair 0.038 0.143 0.044 0.806 0.005 0.064 0.065 0.037 0.271 0.061 0.115 0.245 0.042 0.014 0.008 Most of the time people helpful or mostly looking out for themselves 0.061 0.091 0.739 0.037 0.131 0.001 0.008 How interested in politics 0.092 0.027 0.116 0.710 0.294 0.231 0.035 Politics too complicated to understand 0.090 0.590 0.067 0.001 0.074 0.078 0.064 Making mind up about political issues 0.691 0.073 0.106 0.165 0.064 0.042 0.142 Politicians in general care what people like respondent think 0.288 0.677 0.049 0.074 0.005 0.143 0.128 Politicians interested in votes rather than peoples opinions 0.168 0.689 0.058 0.130 0.165 0.009 0.093 Trust in country's parliament 0.007 0.612 0.458 0.130 0.042 0.105 0.049 Trust in the legal system 0.577 0.171 0.064 0.198 0.028 0.033 0.278 Trust in politicians 0.127 0.689 0.435 0.109 0.070 0.010 0.071 How satisfied with present state of economy in country 0.010 0.177 0.749 0.119 0.072 How satisfied with the national government 0.000 0.167 0.801 0.012 How satisfied with the way democracy works in country 0.028 0.160 0.774 0.077 0.045 The less government intervenes in economy, the better for country 0.262 0.057 0.115 0.123 Government should reduce differences in income levels 0.055 0.030 0.141 0.021 Employees need strong trade unions to protect work conditions/wages 0.058 Ban political parties that wish overthrow democracy 0.229 0.107 0.160 0.291 0.231 Important in life: politics 0.592 0.360 0.096 Employment status 0.135 0.002 0.034 Year of birth 0.043 0.006 0.104 Gender 0.023 0.043 0.064 Years of fulltime education completed 0.563 0.120 0.006 Feeling about household's income nowadays 0.326 0.065 0.225 0.011 0.034 0.085 0.092 0.029 0.102 0.135 0.207 0.157 0.038 0.008 0.000 0.050 0.034 0.094 0.148 0.048 0.056 0.200 0.024 0.039 0.018 0.213 0.044 0.051 0.054 0.021 0.120 0.046 0.025 0.010 0.055 0.712 0.650 0.026 0.281 0.045 0.681 0.146 0.066 0.038 0.114 0.578 0.104 0.061 0.155 0.105 0.112 0.332 0.068 0.604 0.076 0.713 0.138 0.385 0.138 0.082 0.147 0.042 0.744 0.132 0.132 0.042 0.122 0.027 0.056 0.286 0.440 Table 17. Results of factor analysis for 3 rd Wave Matured Democracies. Source: own computations based on ESS data. 22
For the group of 3 rd Wave Matured Democracies we defined the following factors: Factor F1 (Importance and understanding of politics with education) How interested in politics Politics too complicated to understand Making mind up about political issues Important in life: politics Years of fulltime education completed Factor F2 (General political trust) Politicians in general care what people like respondent think Politicians interested in votes rather than peoples opinions Trust in country's parliament Trust in the legal system Trust in politicians Factor F3 (Political satisfaction) How satisfied with present state of economy in country How satisfied with the national government How satisfied with the way democracy works in country Factor F4 (Social trust) Most people can be trusted or you can't be too careful Most people try to take advantage of you, or try to be fair Most of the time people helpful or mostly looking out for themselves Factor F5 (Media influence and age of respondent) TV watching, news/ politics/current affairs on average weekday Radio listening, news/ politics/current affairs on average weekday Newspaper reading, politics/current affairs on average weekday Year of birth Factor F6 (Social coverage desires and democratic support) Government should reduce differences in income levels Employees need strong trade unions to protect work conditions/wages Ban political parties that wish overthrow democracy Factor F7 (Employment and gender) Employment status Gender Factor F8 (Private economic perception and liberalism) The less government intervenes in economy, the better for country Feeling about household's income nowadays 23
These factors were specified for the group of 3 rd Wave Matured Democracies. In fact they differ significantly from the factors specified for old democracies, therefore our assumption already confirms. However to drop the group of Old Democracies we need to prove that factor analysis for 3 rd Wave Young Democracies will be also different. Unlikely to the old democracies 3 rd wave matured democracies are characterized in much greater respect by understanding of politics and education of respondents than by political trust or satisfaction. While in the old democracies political trust and satisfaction were crucial in 3 rd wave matured democracies these factors dropped to second and third positions. Similarly social trust which was important in old democracies lost on importance while we moved to analyze 3 rd wave matured democracies. Media influence and social coverage desires advanced in 3 rd wave matured democracies group which can suggest that people who live in moderately young democracies desire more support from the government and are more prone to get influenced by media than those who live in democracy for generations. It is actually (in our opinion) Lipset s getting used to the system that decides about the importance of these two last factors. People who lived in democracy for decades raised their offspring in understanding of democratic rule, but those who experienced political transformation in their lives have to get used to the new situation. In fact this explanation does not stand in contrast to rationalculturalist explanation if we consider that people raised in different conditions will have different expectations. It is as if to say that those who got as an initial endowment democratic understanding have their expectations much more realistic than those who experienced system switch. The following table presents results of the last factor analysis for the group of 3 rd Wave Young Democracies. Here the KMO and Bartlett s tests prove sensibility of our analysis. KMO and Bartlett's Test(a) KaiserMeyerOlkin Measure of Sampling Adequacy. 0.831 Bartlett's Test of Sphericity Approx. Chi Square 12,977.035 df 351 Sig. 0.000 Table 18. KMO and Bartlett s tests for 3 rd Wave Young Democracies. Source: own computations based on ESS data. Table 19 provides results of the factor analysis for 3 rd wave young democracies. 24
TV watching, news/ politics/current affairs on average weekday 0.004 Component 1 2 3 4 5 6 7 8 0.106 0.656 0.040 Radio listening, news/ politics/current affairs on average weekday 0.004 0.201 0.642 0.022 Newspaper reading, politics/current affairs on average weekday 0.049 Most people can be trusted or you can't be too careful 0.193 Most people try to take advantage of you, or try to be fair 0.126 0.220 0.035 0.051 0.579 0.009 0.008 0.014 0.021 0.059 0.051 0.233 0.043 0.006 0.209 0.015 0.102 0.037 0.120 0.755 0.076 0.090 0.048 0.797 0.093 0.010 Most of the time people helpful or mostly looking out for themselves 0.196 0.033 0.030 0.732 0.034 How interested in politics 0.049 Politics too complicated to understand 0.052 Making mind up about political issues 0.036 Politicians in general care what people like respondent think 0.266 Politicians interested in votes rather than peoples opinions 0.270 Trust in country's parliament 0.715 Trust in the legal system Trust in politicians 0.697 0.691 0.523 0.716 0.704 0.106 0.019 0.105 0.029 0.059 0.536 0.053 0.066 0.097 0.213 0.081 0.158 0.068 0.066 0.032 0.009 0.101 0.029 0.034 0.083 0.761 0.080 0.006 0.024 0.081 0.079 0.201 0.006 0.005 0.013 0.102 0.011 0.146 0.095 0.007 0.021 0.129 0.778 0.034 0.003 0.059 0.145 0.223 0.072 0.105 0.099 0.064 0.174 0.362 0.094 0.090 0.063 0.073 0.038 0.024 0.109 0.056 0.195 0.075 How satisfied with present state of economy in country 0.734 0.034 0.000 0.151 0.054 0.175 0.010 How satisfied with the national government 0.797 0.021 How satisfied with the way democracy works in country 0.733 0.046 The less government intervenes in economy, the better for country 0.034 0.118 0.094 0.091 0.064 0.088 0.003 0.059 0.017 0.092 0.009 0.118 0.084 0.012 0.293 0.400 0.084 0.125 0.076 0.131 0.139 0.175 Government should reduce differences in income levels 0.017 0.000 0.000 0.059 0.077 0.760 0.114 0.049 Employees need strong trade unions to protect work conditions/wages 0.017 0.040 0.089 0.036 0.098 0.754 0.060 0.099 Ban political parties that wish overthrow democracy 0.014 0.065 0.035 0.086 0.029 0.086 0.041 0.831 Important in life: politics 0.175 0.477 0.049 0.187 0.340 0.053 0.004 0.339 Employment status 0.047 0.028 0.026 0.074 0.046 0.007 0.781 0.002 Year of birth 0.024 0.076 0.479 0.024 0.052 0.049 0.375 0.250 Gender 0.036 Years of fulltime education completed 0.024 Feeling about household's income nowadays 0.271 0.622 0.020 0.139 0.124 0.239 0.135 0.065 0.080 0.092 0.170 0.239 0.640 0.003 0.132 0.137 0.321 0.025 0.162 0.330 0.058 0.138 Table 19. Results of factor analysis for 3 rd Wave Young Democracies. Source: own computations based on ESS data. 25
From the factor analysis for the group of 3 rd Wave Young Democracies we got yet another set of factors. Factor F1 (Political trust and satisfaction) Trust in country's parliament Trust in the legal system Trust in politicians How satisfied with present state of economy in country How satisfied with the national government How satisfied with the way democracy works in country Factor F2 (Understanding politics and gender) How interested in politics Politics too complicated to understand Gender Making mind up about political issues Factor F3 (Media and political importance with age) TV watching, news/ politics/current affairs on average weekday Radio listening, news/ politics/current affairs on average weekday Newspaper reading, politics/current affairs on average weekday Important in life: politics Year of birth Factor F4 (Social trust) Most people can be trusted or you can't be too careful Most people try to take advantage of you, or try to be fair Most of the time people helpful or mostly looking out for themselves Factor F5 (Opinion about politicians) Politicians in general care what people like respondent think Politicians interested in votes rather than peoples opinions Factor F6 (Social coverage desires and liberalism) The less government intervenes in economy, the better for country Government should reduce differences in income levels Employees need strong trade unions to protect work conditions/wages Factor F7 (Individual economic characteristics) Employment status Years of fulltime education completed Feeling about household's income nowadays Factor F8 (Democratic support) Ban political parties that wish overthrow democracy 26
As we already mentioned for the group of 3 rd Wave Young Democracies we obtained in factor analysis yet another set of factors which accomplishes part of our task proving that 3 rd wave democracies are different from old democracies. The most surprising is the position of the gender so high in the ranking of the variables explained by nonobservable factors. Political trust and satisfaction in young democracies has similar position like in the old democracies but position of media and understanding of politics are also different than in the previous two groups. Media became 3 rd most important factor in the group of 3 rd Wave Young Democracies. Gender which leveled up to the second factor in the previous 2 groups of countries played a very minor role. Interestingly political trust in the group of 3 rd Wave Young Democracies came to its position noted first for Old Democracies group. However it is not the same political trust in both groups. Factor 1 for Old Democracies contained 2 more variables than factor 1 for 3 rd Wave Young Democracies : politicians care what people think, and politicians care only about the votes. Inexistence of those two variables seems very important here. Those variables exactly show at least partly that we should think of political disaffection among 3 rd Wave Young Democracies as of resulting from disappointment of political expectations of people. People believe in the electoral propaganda in 3 rd Wave Young Democracies to a much greater degree than in old democratic societies. It is disappointed expectations that afterwards turn people out from politics in 3 rd wave young democracies and lack of those disappointed expectations that keep people trusting in politics in old democracies. As we saw above sometimes comparing factors in factor analyses may prove very useful and educative. Thus as we said earlier if it is possible factors in the factor analysis should be interpreted otherwise it becomes dry statistical tool. At this moment we can proceed to the next section and analyze the regressions on the joint and divided sample of 3 rd wave democracies. 27