Stefan Stojkovic s2583437 Supervisor: Dr. M.A.M. Nagtzaam Parties, Parliaments and Democracy Second reader: Prof.dr P. Kopecky
Democratic dissatisfaction and Populist Voting in
East-Central Europe
Submitted on June 15, 2020 Word count: 8879
Abstract
While it has been well established that political dissatisfaction is related to voting for populist parties, we know less about what specifically this dissatisfaction has to do with the functioning of democracy. This question of the relationship between how democracy is perceived to function and party preferences has attracted certain attention in established democracies, but almost none in post-communist East-Central European countries. To fill this gap, I ask (1) what aspects of dissatisfaction with democracy are related to voting for populist parties in this region and (2) to what extent dissatisfaction drives support for populist parties. The study employs European Social Survey 2012 and includes 6 East-Central European countries. I run a comparative quantitative study on the demand side and compare populist parties’ voters to non-populist voters and abstainers. Results suggest that support for democratic protest voting model is very country-specific, and that overall patterns with regards to the suggested model are hardly identified in post-communist democracies.
Introduction
Just as the link between established parties and voters has been declining for years (Ignazi, 1996; Mair, 2006; Van Biezen et al., 2012; Whiteley, 2011), so has the representative function that political parties perform. This decline has largely been followed by the rise of populist radical right parties in North-Western European countries on the one hand, and leftist populist movements in Southern Europe, on the other. This trend hase become especially noticeable after the European debt and Refugee crisis (Hernández & Kriesi, 2016) and, interestingly, has been seen not just as a danger to European party systems, but also as the opportunity for their realignment (for instance, Kriesi, 2014: 361).
With increasing number of citizens giving support to parties that challenge political mainstream in the aftermath of the aforementioned crises, the question of why voters vote for populists has understandably become the subject of a rigorous empirical scrutiny (in this study, by populism I mean political parties that mainly thrive on crises and political dissatisfaction). On the one hand, we have learned that the nature of structural conflicts in the European political arena has changed, so that new political issues – such as immigration or globalization – have become important dividing lines influencing elections to a higher degree than before (Hooghe & Marks, 2018; Kriesi, 2010). On the other hand, we have also learned that recent crises have deepened the distrust in established political parties and contributed to dissuading voters from supporting them (Akkerman et al., 2017; Zhirkov, 2014).
Most of the research on populist voting has been done in older or established democracies in Western Europe. Taking into account a long history of party systems in this region, significant scholarly attention has understandably been paid to populist parties who appear to challenge well-established political actors. For these reasons, in the paragraphs that follow I first review the literature on voting for populist parties in established democracies. I
then move on to analyze what is different East-Central Europe and why it is relevant to study populism in this region. After having reviewed the main findings relevant to this study, I proceed to the next section where I formulate hypotheses.
Who votes for populist parties?
In a nutshell, voters who vote for populist parties have in common that hold anti-immigration and Eurosceptic attitudes (Goodwin, 2011; Oesch, 2008; Rooduijn, 2018; Rydgren, 2008). Socio-demographically, the support base of populist parties is predominantly composed of younger males and somewhat less educated citizens (e.g. Mudde, 2009). They also report distrust in institutions (Ziller & Schübel, 2015) and discontent with their political system (Arzheimer, 2009; Kehrberg, 2015). Even though discontent is commonly related to electoral volatility (Dejaeghere & Dassonneville, 2017) citizens voting for populist parties have not shown high electoral volatility in recent years. In other words, political dissatisfaction that accompanies support for populist parties has not faded away after the financial crisis, but persisted and even stabilized the voter base of these parties (Voogd & Dassonneville, 2018). For years after the 2008 crisis, this group of voters has turned out to be the loyal electorate, making the parties they support credible challengers of both leftist and rightist mainstream parties in Europe (Oesch & Rennwald, 2018).
In comparative studies, dissatisfaction has often been treated holistically (Bowler et al., 2017; Rooduijn, 2018; Zhirkov, 2014). Put differently, we know that dissatisfaction drives citizens to support populist political alternatives but we know less about what exactly this dissatisfaction refers to. In a rare recent attempt to deconstruct the concept of dissatisfaction in populist parties’ voting, Hernández (2018) breaks it down into multiple indicators that measure the level of (dis)satisfaction with different aspects of democracy in established European democracies. He shows that discontent is accompanied by citizens’
normative expectations about how democracy is supposed to work. In other words, he argues that a vote for (populist) challengers is not just a protest vote, but that it also signals a change in a democratic system that voters would like to see.
Disaggregating the concept of discontent or dissatisfaction is especially relevant during or after major crises, as satisfaction with democracy declines during these events (Kitschelt, 2002). In Western democracies, the last decade has been marked by the weakening of established parties and strengthening of populist, most often radical right populist, parties and movements. So far, we have learned a great deal about populist voting in established democracies. We have also learned that dissatisfaction with how democracy functions in most cases relates to the support for populist alternatives, rather than established parties (Hernández, 2018). When it comes to East-Central Europe, however, we know less about how dissatisfaction with different aspects of a democratic system is related to populist voting. On the one hand, shedding light on that problem is scientifically relevant because it contributes to further understanding the political competition in young CEE democracies. On the other hand, it is a politically relevant puzzle, as it may reveal systemic shortcomings on which populism in this region thrives.
East-Central Europe
Before defining the research problem, it is important to outline a few particularities of party systems in East-Central Europe (compared to Western European systems). Above all, Central and Eastern European party systems are less institutionalized – meaning that parties do not have stable roots in society – links between parties/politicians and voters are more personalistic, party identification is lower, programmatic voting is less present and volatility levels are higher (Casal Bértoa, 2013; Mainwaring & Torcal, 2006). These factors have contributed to a different nature of party competition in young democracies of CEE countries.
For instance, parties that led the transition were largely challenged on the basis of competence and probity, rather than policy platforms (Stanley, 2017). While radical right populism was mostly successful in the years right after the fall of Berlin wall, the 21st century has seen the success of centrist populist parties largely thriving on transition fatigue and on blaming governing parties (or former elites) for everything that is not going well (Mudde, 2017; Stanley, 2017; Učeň, 2007).
The scholarship agrees on that populist parties in CEE countries did not necessarily need to develop recognizable policy platforms because playing on discontent cards has often been sufficient for electoral success. However, these discontent-based explanations have often been treated monolithically – discontent with the system, with democracy. I have already outlined above why a more in-depth, rather than a monolithic approach to dissatisfaction with democracy, is more informative and relevant, and while there have been such attempts in established democracies, they seem to be lacking in the CEE region. Therefore, to understand how citizens evaluations of satisfaction with democracy affect party competition between populist and other parties in East-Central Europe, I ask to what extent voting for populist parties is related to different indicators of dissatisfaction with a democratic system in East-Central European countries. Again, the answer to this question will tell us more about (1) the political competition in the CEE region and (2) to what extent different aspects of the functioning of a democratic system motivate the populist parties’ electorate.
One may suggest that dividing parties into populists and other is problematic, given the lack of established political parties in the Western sense, which may blur the political competition analysis. Yet, the lack of established parties is not a problem per se. My aim is not to explore what threatens the status quo, but rather to understand whether discontent with particular aspects of the functioning of democracy feeds the support for populist parties in
East-Central Europe. Moreover, as the convergence between Western and East-Central European party systems (Enyedi & Casal Bértoa, 2018) demonstrates, the differences between the two are in degree and they show the trend towards diminishing. This further suggests that there are no strong theoretical reasons to believe that exploring functioning of democracy in the context of CEE populists vs non-populists party competition would not give us valuable insights. In the end, we will better understand a widespread political dissatisfaction in East-Central Europe, the concept on which most of the party competition in the region is arguably based (Stanley, 2017).
To this end, I employ a comparative demand side quantitative study. The defining features of populist parties in the region are anti-elitism and anti-establishment (Stanley, 2017), i.e. these parties are arguably the voice of protest. For this reason. I rely on protest voting theoretical model expecting that this protest will be expressed through dissatisfaction with a democratic system. Protest voting model defined though democratic discontent seems to be especially relevant among CEE countries. In other words, political competition in the region is defined by widespread and largely unspecified political dissatisfaction on which populist parties thrive (Stanley, 2017; Učeň, 2007). My intention is to specify the nature of this dissatisfaction and see if we can observe any patterns across cases.
Formulation of hypotheses
To approach the problem of the relationship between evaluations of how democracy functions on the one hand, and party preferences on the other, I begin with defining indicators through which I measure the satisfaction with different aspects of democracy. Through these indicators – derived from common theoretical accounts about democracy – I operationalize my (democratic) protest voting model. Finally, having defined indicators, I end the section by deriving two general hypotheses for this study.
To test the theory of democratic protest voting, I shall, above all, consider the question of freedom and fairness of elections. The reason for this is not only because elections are embedded in the relevant modern definitions of democracy (Schumpeter, 1976: 269), but also because they are important even to some of the harshest autocracies in the world. The fact that the regimes like Russia or China care about “dressing their windows” (Gandhi & Przeworski, 2007; Przeworski, 2018) with elections – even though people maybe get to vote but not really to choose in these elections – speaks about the symbolic power of the concept which is why I cover it in this study.
In addition to Schumpeterian definitions of democracy, it would also be reasonable to take into account classical definitions that he reacts to (Schumpeter, 1976: 250-268). These classical conceptions developed by enlightenment thinkers commonly hold that there is a Common Good which can be recognized by politically sophisticated rational individuals. What follows from this understanding of democracy is that people should be given a final say (e.g. by referendum) on at least most important social issues. The reason to include a direct
say as an indicator of democracy in this study does not only come from the fact that this
concept belongs to classical definitions. It is also that direct democracy is perhaps more often than not an integral part of the contemporary populist right-wing parties agenda (De Blasio &
Sorice, 2018) – the reasons for which I do not analyze here – even though this way of exercising politics is not necessarily mirrored by their voters (Bowler et al., 2017).
Aside from elections and referendums, I also consider concepts approximating equality and liberty, the main principles of liberal democracy (Fukuyama, 2011), given that Western liberal democracy has largely served as a role model to post-communist European states during their transition period. First, I consider government’s protection against inequality as an indicator of democracy. I base this decision on findings that poverty has a number of negative consequences for the sustainability of a democratic polity (Haushofer & Fehr, 2014) and that raising inequalities are highly undesirable among the public, even though often not punished electorally (Bartels, 2005). Moreover, government’s protection against poverty might be a particularly important consideration for citizens in post-communist European countries where citizens have historically recent experience of the extensive government’s interventions in the economy. Second, I also consider equality before the law as a concept that is commonly identified with the modern notions of democracy in theory and practice. Third, I also look at the freedom of expression. The reason for this lies in the fact that personal freedoms to choose and express oneself have been cross-culturally more strongly tied to people’s understanding of democracy, than any procedural components of the concept (Ottemoeller, Marcus & Mease, 2001).
Finally, I find two more concepts relevant to include in this study: protection of minority rights (by the state) and the extent to which politicians in a given country take into account the views of other European governments. While the decision to include the former is self-telling, I find the latter relevant in the contemporary globalized world where populist parties continue to campaign on real or alleged discontent with transferring power beyond national borders. While the term does not target this transfer of power specifically, it does
contain the element of ‘other’ which is perhaps sufficient to invoke the notions of international interdependence.
Although theoretically grounded, this list of indicators is certainly not exhaustive when it comes to all potential notions associated with democracy that may be important for citizens. However, I find it sufficient to cover basic democratic principles and to inform my conclusions about whether voting for populist parties can be differentiated from other groups on the basis of expectations from a democratic system in East-Central European countries. When I say ‘other groups’, I mean (1) voters of non-populist parties or just ‘other parties’ and (2) abstainers. The reason for including non-voters/abstainers is that I recognize that some politically dissatisfied citizens have likely just given up the electoral participation. In line with Dassonneville et al., (2015), abstainers are expected to be more dissatisfied with the system as a whole and to have lost genuine interest in politics. Given this holistic nature of their dissatisfaction, these citizens are expected to be less able to articulate and target their dissatisfaction specifically. On the contrary, one may expect populist voters to be more emotionally involved and, therefore, more able to express their dissatisfaction and identify the source of it.
On the basis of this short overview, I derive two general hypotheses that will be tested in the study: H1 – the more dissatisfied a voter is with his/her democratic system, the more likely s/he is to vote for a populist party than for a non-populist party or to abstain from voting. H2 – the more important a voter finds a particular aspect of democracy, the effect of dissatisfaction will be stronger.
Instead of seeking support for H1 and H2 per indicator, I will rather comment in terms of the extent to which my results support these hypotheses. As I pool together different populist parties while looking for what unites/divides them, this gives my research problem
an explorative tone, which is why I decide to keep the hypotheses on a somewhat broader level.
Method
I employ multinomial regressions in order to analyze populist parties’ voters against non-populist parties’ voters and abstainers. First, these three groups will be analyzed in a pooled model that brings together populists, non-populists (also referred to as voters of ‘other parties’ in this study) and abstainers from six CEE countries. Second, I break down the study into case-by-case analyzes to see what the similarities/differences across countries are.
Data and Party selection
I employ European Election Survey (EES) 2012, round 6. The reason for using this data is that it focuses specifically on democracy. Therefore, this round provides a convenient pool of democracy-related items for this analysis. Even though the data is somewhat outdated, there are no strong reasons to believe this should be a problem for theory building.
Countries
The selection of countries is conditioned upon data availability. European Election Survey 2012 contains the data from the following CEE countries: Albania, Bulgaria Czechia, Estonia, Hungary, Kosovo, Lithuania, Poland, Slovakia, Slovenia, Ukraine. However, as I have no party classification data for Albania, Kosovo and Ukraine, these countries are excluded from the empirical analysis. In addition, my party selection sources described below identify no significant populist party in Estonia and Slovenia at the time of country’s national elections prior to 2012 data collection. In Slovenia, the sources that I am using do consider the Slovenian National Party as populist, but this party did not make it to the parliament which is why I do not consider it in this study. Moreover, the Slovenian Democratic Party is a borderline case treated as populist by only one source (Inglehart & Norris, 2016: 44), which
does not meet the criteria to be analyzed in this study. By contrast, Estonia does not seem to have witnessed any credible populist parties during the 2011 parliamentary election. For these reasons, my final selection of countries includes: Bulgaria Czechia, Hungary, Lithuania, Poland, and Slovakia.
There is no reason to believe that the countries that I have excluded will hurt the representativeness in my study. On the one hand, Estonia and Slovenia are left out due to the lack of populist parties supply. Given that my focus is on what differentiates populists in CEE, it is reasonable not to consider those CEE countries where there are simply no relevant populist parties registered at a given time. On the other hand, the dropping of Albania, Kosovo and Ukraine due to the lack of data on populist parties classification may appear more problematic. However, the recent history of party systems in these countries does not dramatically differ from those included in the analysis. Therefore, I have no strong reasons to expect that the results there would be substantially different. This is why I conclude that the group of countries selected in this study is fairly representative of the CEE region.
Parties
As indicated previously, I take into account only politically significant parties that participated in the last national elections before the data was collected for European Social Survey 2012. Political significance is defined as reaching the threshold for a seat in national parliaments. In other words, all political parties in a given country that won at least a seat in their national parliament are taken into consideration.
The selection of populist parties is based on three sources: classification of populist parties by Inglehart and Norris (2016: 44), the PopuList database (Rooduijn et al., 2019) and van Kessel’s review of populist parties in Europe (Van Kessel, 2015: 33-73). Whereas the first two are based on expert surveys, the latter presents a rich literature review and discourse
analysis of commonly identified populist parties in Europe. Due to discrepancies in party classification across databases, I have decided to consult multiple sources and classify a party as populist only if it is classified as such in at least two out of three aforementioned sources. All politically significant parties in a given election that were not classified as populist, will be labeled as non-populist parties. Lastly, all respondents who claim they did not vote in a given election will be treated as abstainers.
Table 1. The selection of populist parties in seven CEE countries
Country Party Year of national elections Election results (%)1 Sample/survey-based results (%) Bulgaria GERB Attack
Order, Law and Justice
2009 39.72 9.36 4.13 57.2 2.1 0.4
Czechia Public affairs 2010 10.88 10.1
Hungary Fidesz Jobbik 2010 52.73 16.67 63.2 7.3
Lithuania Way of Courage
Order and Justice 2012
7.99 7.31
0.8 7.0
Poland Law and Justice 2011 31.89 31.6
Slovakia Ordinary People and
Independent Personalities 2012 8.56 5.6
As Table 1 shows, 11 parties meet the criteria to be categorized as populist parties. The number of politically successful populist parties for a given period of time varies per country from 3 in Bulgaria to just 1 in Czechia and Poland. In Table 1 we can also see that this selection covers populist parties of different electoral strength, ranging from 4.23% in the case of Order, Law and Justice Party to 52.72% that Fidesz achieved in Hungary in 2010 parliamentary elections. In order to estimate to what extent the survey data represents vote share that these parties won in reality, I also display survey-based voting results (far right
columns in Tables 1 and 2). This data shows that survey reports on voting fairly approximate the actual election results. Considerable disparities can be noted in the case of the Way of Courage party, which had 7.99% in the national election, but only 0.8% vote share among survey respondents (Table 1). Having said that, I conclude that, nevertheless, the ratio between the populists and non-populists parties vote share in the survey fairly approximates the actual voting ratio.
Table 2. The selection of non-populist parties in seven CEE countries
Non-populist political parties (Table 2) in this study are parties that participated in their national elections prior to the 2012 ESS data collection, reached the electoral threshold,
2 The data from European Election Database: https://nsd.no/european_election_database/
3 No DPS party in the choice list within the dependent variable, even though this party made it to the parliament
Country Party Year of national elections Election results2 (%) Sample/survey-based results (%) Bulgaria3 Coalition for Bulgaria
Blue Coalition 2009 17.70 6.76 34.5 2.3 Czechia
Communist Party of Bohemia and Moravia
Czech Social Democratic Party Civic Democratic Party
TOP 09 2010 11.27 22.09 20.22 16.71 15.0 36.7 21.8 14.4 Hungary Hungarian Socialist Party
Hungary’s Green Party 2010
19.30 7.48
22.9 5.2 Lithuania
Social Democratic Party Homeland Union Liberal Movement AWPL 2012 18.37 15.08 8.57 5.83 17.0 9.9 1.2 3.9 Poland Civic Platform Palikot’s Movement Polish People’s Party Democratic Left Alliance
2011 39.18 10.02 8.36 8.24 48.4 4.5 6.0 7.5 Slovakia Most-Hid
Slovak Democratic and Christian Union – Democratic Party
Freedom and Solidarity
2012 6.90 6.10 5.88 5.3 10.0 6.4
but do not meet the criteria to be classified as populist. However, there are exceptions that I make when it comes to borderline cases. In other words, (1) there are parties that fall into non-populist party group because they are labeled populist according to only one of three party classification sources. Additionally, (2) certain parties were seen as populist by the sources I refer to until some point in the past, but lost their populist characteristics by the national elections in question.
Instances of the first scenario are Direction – Social Democracy (SMER) in Slovakia and the Labour Party in Lithuania. The first one is treated as populist until 2006 after which it reportedly lost its populist features (Rooduijn et al., 2019; Van Kessel, 2015: 65). Similarly, the second one is considered populist during the 2004 election campaign but not thereafter (Rooduijn et al., 2019; Van Kessel, 2015: 58). As regards the second scenario, the Christian Democratic Movement in Slovakia is seen as populist only by Inglehart and Norris (2016: 44) but not by the other two sources. In both of the aforementioned scenarios, I consider such parties as borderline cases and decide to leave them out of the empirical analysis to avoid them blurring the differentiation between populist and non-populist parties. I shall, however, also run the analysis with these parties in the non-populist group and report if significant changes in the output are observed.
Variables
Independent variables are classified in four sets. The first set is composed of variables where respondents were asked to say how important different aspects of democracy were for them: (1) free and fair elections, (2) equal treatment before the law (3) having a final say on political issues by voting in referendums, (4) that the government protects against poverty, (5) their politicians take the view of other EU governments into account, (6) minority rights are protected and that (7) everyone is free to express political views. The second set consists
of the very same variables where respondents were now asked to estimate to what extent each of the aforementioned indicators (1-7) applies to their country. Both sets of variables range from 1 – not at all important for democracy/does not apply at all [to my country], to 10 –
extremely important for democracy/applies completely. As for the third set, I create
interaction terms out of the first set of importance indicators and the second set of assessment indicators. Given that these interaction terms have a range of 1-100, I divide them by 10 to facilitate the interpretation of coefficients. Finally, as discussed in the introduction, the fourth set consists of control variables. Those are gender, age and education. Furthermore, I consider anti-immigration attitudes, Euroscepticism and interest in politics.
Here I briefly elaborate how control variables are operationalized. First: Age, expressed in years. Second: Gender, recoded so as to 0 means female and 1 male. Third: Interest in politics, a 4-point ordinal variable which is treated as metric in the analysis. Originally, this variable ranged from 1 – very interested in politics, to 4 – not at all interested. This variable is, however, reversed before the analysis so as to higher scores mean more interest in politics. Fourth: EU unification, a 10-point scale where respondents were asked to estimate in a scale from 0 – the unification gone too far, to 10 – unification should go further. Fifth: Immigration, again a 10-point scale, asking respondents to estimate the extent to which immigrants generally make a country worse or better place to live. Sixth: Education – here I rely on a standardized education variable based on International Standard Classification of Education (ISCE). This is a 7-point ordinal variable which I treat as metric in the analysis. The ISCE levels are: 1 – less than lower secondary, 2 – lower secondary, 3 – lower tier upper secondary, 4 – upper tier upper secondary, 5 – advanced vocational, 6 – lower tertiary BA level education, 7 – higher tertiary MA level education. Finally, I include Trust in institutions as a variable made out of the following 10-item scales: Trust in the legal system, Trust in the police, Trust in politicians, Trust in political parties, Trust in national’s parliament. After a
one-factor solution was identified in the factor analysis for trust-variables, the scale was made by adding them up and dividing by 10.
All independent variables are mean centered (except gender) to account for multilevel data structure in my analysis and for multicollinearity issues. Furthermore all four sets of variables are included in the analysis simultaneously. Descriptive statistics for independent variables is shown in Table 3 (appendix).
It is important to point out that the variable – freedom to express one’s political views (“political views” in Table 3) has a considerably lower response rate than other variables. As I have also determined that this response rate is relatively equally distributed among selected countries, the inclusion of this variable dramatically decreases my number of cases. This may hurt the accuracy of my analysis especially where the number of cases in a group is already low as it is, for instance, in the populist voting group in Czechia or Slovakia whose populist categories are represented by voters of only one party (Table 1). For this reason, I will check if this variable significantly contributes to the explanation of my dependent variable, and if it does not, I will comment on it but display the version of the empirical analysis with leaving this variable out in order not to lose cases.
As regards dependent the variable, it refers to vote choice in previous national elections, i.e. first national election prior to the European Election Study Round 6 data collection (see national election years in Tables 1 and 2). The question was who did you vote
for in previous elections. For the purpose of the analysis, all of those who voted for one of the
non-populist parties (Table 1) were labelled 0 and set as a reference group. Those who responded they voted for populist parties (Table 2) were marked by 1 and abstainers by 2. The proportion of respondents per country per dependent variable is displayed in Table 4.
Table 4. Descriptive statistic for dependent variable
Country Populist parties voters (%) Other voters (%) Abstainers (%) N Invalid cases Bulgaria 49.3 24.3 26.4 1092 1167 Czechia 6.6 60.1 33.3 1012 996 Hungary 48.5 21.0 30.5 1093 920 Poland 20.3 49.5 30.1 1035 862 Lithuania 8.3 20.0 71.7 701 1407 Slovakia 8.0 38.3 53.6 636 1210
Note: the percentage of voters based on variables that will be included in the empirical analysis when excluding “freedom to express one’s political views”
Results
Bellow I first show and interpret the results from the pooled model, after which I move to a country-by-country analysis. Results are interpreted by the set of variables: first the set of importance indicators, then assessment and, finally, interactions. Table 4 summarizes the pooled analysis. The voters of non-populist parties are the reference category.
Table 5. Multinomial logistic regression analysis of the probability of voting for populist parties
populists abstainers
(constant) -2.223*** (.148) -.570*** (.090)
Importance indicators
Free and fair elections .025 (.034) -.094** (.027)
Courts .010 (.034) .006 (.029)
Referendums .013 (.028) -.021 (.025)
Protection against poverty .077** (.024) .068* (.021)
Other governments’ views -.008 (.017) -.001 (.016)
Minority rights -.055* (.022) -.053** (.010)
Assessment indicators
Free and fair elections -.001 (.018) -.033* (.016)
Courts -.061** (.019) -.037 (.018)
Referendums .017 (.017) .026 (.016)
Protection against poverty .030 (.019) .019 (.018)
Other governments’ views .070*** (.017) .039 (.016)
Minority rights .021 (.018) .001 (.016)
Importance x Assessment
Free and fair elections .025 (.097) -.080 (.076)
Courts .086 (.102) -.070 (.084)
Referendums -.020 (.079) .036 (.071)
Protection against poverty -.031 (.082) .034 (.069)
Other governments’ views .069 (.054) -.124** (.053)
Minority rights -.018 (.069) -.026 (.062) Control variables Age Gender -.018*** (.003) .031 (.080) -.029*** (.002) .070 (.073) Interest in politics -.147** (.052) -.914*** (.051) EU unification -.049** (.016) -.010 (.015) Immigration -.033 (.019) -.023 (.018) Education Trust in institutions Country-fixed effects -.094*** (.025) .048* (.022) ✓ -.211*** (.024) -.058** (.020) ✓ -2LL 9482.117 .361 .408 5505 Cox and Snell’s R2
Nagelkerke’s R2
N
Note: multinomial logistic regression coefficients with standard errors in brackets. ***p < 0.001, **p < 0.01, * p < 0.05
Even though I report regression coefficients in table 5, I use the divide by 4 rule for interpretation in order to obtain voting probability estimations and facilitate the explanation of coefficients. Besides, while I reflect on all significant results, (1) I do not elaborate the effects of control variables in detail and (2) my interpretation is more focused on the comparison between two voters groups than on the comparisons with abstainers. Finally, what I treat as statistical significance is the coefficients that exert statistical effect on at least 95% confidence level.
First of all, on the basis of pseudo R-Squareds (Nagelkerke and Cox and Snell in Table 4), I conclude that explanatory power of the model is good. With regards to normative/importance coefficients, we see that for every point increase in how important respondents see for a democracy that government protects against poverty, the probability of voting for populist parties increases by 1.93% compared to voting for other parties. Additionally, for every point increase in the importance of the minority rights protection, the probability of voting for populist parties decreases by 1.38%.
When it comes to assessment indicators, respondents are 1.53% less likely to vote for populist parties for every point increase on the scale assessing to what extent it applies in a country that courts treat everybody the same. The respondents are also 1.75% more likely to vote for these parties for every point increase in their perception of how much their governments take into account the views of other governments. Out of the control variables, we see that populist parties voters are younger, less interested in politics, less educated, more opposed to the EU unification and somewhat have somewhat higher trust in institutions.
As regards the abstaining from voting, respondents are more likely to abstain the less they find important free and fair elections and the protection of minority rights (when compared to non-populist parties’ voter group). Their assessment of the free and fair elections is also the lowest among the analyzed groups. Abstainers are the youngest group,
least educated, least interested in politics and they have the lowest trust in institutions. In general, those who abstain from voting appear to be closer to voters of populist parties than to voters of other parties.
Important to note is, however, that the substantive effect of these coefficients is rather low. Based on the pooled model (Table 5), therefore, there seems to be a very limited support for H1, while H2 is not supported. However, before discussing these findings in a greater detail, I move to case-by-case analyzes to see if there are relevant differences across countries with regards to support for my hypotheses.
Table 5. Multinomial logistic regression per country analyses of the probability of voting for populist parties Populists
(BG) Abstainers (BG) Populists (HU) Abstainers (HU) Populists (CZ) Abstainers (CZ)
(constant) 1.158*** (.174) .711*** (.194) .936*** (.158) .098 (.183) -2.311** (.253) -.767*** (.142) Importance indicators
Free and fair elections -.018 (.096) -.138 (.111) .285*** (.076) .089 (.076) .136 (.123) -.095 (.056)
Courts -.072 (.087) -.150 (.097) -.066 (.093) -.028 (.102) -.118 (.135) .022 (.064)
Referendums .089 (.058) .019 (.066) -.010 (.077) .065 (.086) .134 (.109) .001 (.054)
Protection against poverty .029 (.056) .114 (.074) .041 (.056) .070 (.064) .088 (.080) .045 (.043) Other governments’ views .029 (.039) -.038 (.046) -.085 (.043) -.036 (.060) -.091 (.070) .038 (.041) Minority rights -.141** (.044) -.123* (.049) -.022 (.056) -.130* (.059) .001 (.080) .016 (.044) Assessment indicators
Free and fair elections .072 (.041) .029 (.115) .110* (.045) .086 (.048) .022 (.069) -.035 (.037) Courts -.124** (.040) .147** (.049) -.099* (.044) -.023 (.049) .047 (.068) .008 (.039)
Referendums .039 (.034) .088* (.110) .049 (.048) .021 (.053) -.014 (.061) .059 (.035)
Protection against poverty -.054 (.047) .044 (.134) .038 (.042) .029 (.047) -.002 (.074) -.025 (.044) Other governments’ views .055 (.034) -.165 (.120) .109** (.041) .010 (.047) -.009 (.072) .042 (.042) Minority rights .055 (.033) .174 (.123) .001 (.044) -.035 (.047) .016 (.066) -.050 (.037) Importance x Assessment
Free and fair elections -.150 (.281) -.457 (.320) .434 (.227) .551* (.221) .339 (.408) -.071 (.187)
Courts .316 (.255) .189 (.293) .148 (.247) -.258 (.259) -.029 (.334) .210 (.182)
Referendums .431** (.167) 1.161* (.571) -.135 (.218) .253 (.253) -.829** (.291) -.123 (.160) Protection against poverty -.072 (.196) .312 (.296) -.092 (.182) -.185 (.203) -.145 (.254) -.129 (.139) Other governments’ views .001 (.120) -.055 (.140) -.271* (.136) -.411** (.155) -.325 (.247) -.049 (.140) Minority rights .044 (.135) .103 (.147) -.129 (.182) .093 (.188) .241 (.283) -.053 (.146) Control variables ✓ ✓ ✓ ✓ ✓ ✓ -2LL 2000.575 .226 .258 1092 1853.035 .322 .347 1093 1483.918 .212 .260 1012 Cox and Snell’s R2
Nagelkerke’s R2 N
Table 5. Continued
Populists
(PL) Abstainers (PL) Populists (SK) Abstainers (SK) Populists (LT) Populists (LT)
(constant) -.430* (.170) .091 (.151) -1.307*** (.307) .110 (.194) -.942* (.378) 1.336*** (.249) Importance indicators
Free and fair elections -.053 (.073) -.154* (.064) .029 (.158) -.276** (.085) -.112 (.175) -.053 (.126)
Courts -.177* (.089) -.028 (.087) -.083 (.156) -.058 (.088) .042 (.163) .015 (.105)
Referendums .148* (.074) .049 (.062) .070 (.149) -.074 (.081) .091 (.157) -.002 (.096)
Protection against poverty .109 (.064) .083 (.056) .309* (.129) .150* (.068) .315* (.146) -.097 (.079) Other governments’ views .011 (.037) .060 (.035) -.065 (.082) -.145** (.049) .016 (.078) -.039 (.056) Minority rights -.113 (.073) -.133* (.066) -.121 (.117) -.029 (.069) -.056 (.108) -.036 (.082) Assessment indicators
Free and fair elections -.151*** (.039) -.101** (.036) -.169 (094) -.198 (.062) -.278** (.095) -.141 (.062)
Courts -.001 (.051) -.071 (.046) -.129 (.114) .003 (.047) .199* (.099) .147 (.073)
Referendums -.023 (.044) -.067 (.040) -.190 (.084) .016 (.057) -.026 (.094) .022 (.065)
Protection against poverty -.100 (.052) -.001 (.045) .026 (.100) .095 (.060) .309** (.102) .229 (.073) Other governments’ views .114** (.041) .053 (.038) -.093(.088) .092 (.054) .018 (.096) .043 (.066) Minority rights .036 (.060) .170** (.057) .126 (.088) .203*** (.055) -.006 (.086) -.084 (.059) Importance x Assessment
Free and fair elections -.234 (.210) -.296 (.186) -.206 (.493) -.575* (.255) -.256 (.433) -.105 (.292)
Courts -.007 (.362) .574 (.340) -.129 (.615) -.523 (.333) -.622 (.503) -.755* (.317)
Referendums -.298 (.234) -.418* (.202) .620 (.441) .682** (.225) .048 (.467) .268 (.285) Protection against poverty .356 (.273) .275 (.231) -.228 (.431) -.046 (.313) .384 (.516) 694 (.305) Other governments’ views .074 (.124) -.103 (.118) -.407 (.298) -.080 (.177) .521 (.272) .004 (.179) Minority rights -.034 (.298) -.474 (.286) -.102 (.361) -.179 (.212) .243 (.322) -.208 (.227) Control variables ✓ ✓ ✓ ✓ ✓ ✓ -2LL 1857.001 .239 .274 1035 910.092 .315 .377 636 705.716 .410 .522 701 Cox and Snell’s R2
Nagelkerke’s R2 N
Note: multinomial logistic regression coefficients with standard errors in brackets. ***p < 0.001, **p < 0.01, * p < 0.05
Table 5 summarizes the results of multinomial logistic regressions by country. The non-populist voter group is again set as the reference category. First of all, pseudo R’s show a good fit of the models, with Lithuania indicating the highest model-fit values.
With regards to importance/normative indicators of democracy, for every point increase in the importance of minority rights, the likelihood of voting for populist parties decreases by 3.5% in Bulgaria. What plays a role in Hungary in this context is free and fair elections. In other words, the more important free and fair elections are considered, the probability of voting for populists increases by 7.13%. Slovakian and Lithuanian voters of populist parties seem to be concerned with the protection against poverty: the likelihood of supporting populist parties raises by 7.73% and 7.85% in these counties respectively, for every point increase in the importance of government’s protection against poverty. Lastly, polish respondents are 4.43% more likely to support populist parties in Poland for every point decrease in the importance that courts treat everybody the same in a democracy and 3.7% more for every point increase in the importance of referendums. As for the abstainers, people are more likely not to vote the less important they see minority rights (Bulgaria, Hungary and Poland), free and fair elections (Poland and Slovakia), taking into account other government’s views (Slovakia) and the more important they find protection against poverty in a democracy (Slovakia).
When respondents are asked to say to what extent different indicators of democracy apply to their country, we get the following difference between populist voters and voters of other parties: citizens are more likely to vote for populist parties in Bulgaria and Hungary the worse they assess equality before the law. In other words, for every point decrease in this indicator, the likelihood of voting for these parties raises by 3.1% (Bulgaria) and 2.48% (Hungary). Additionally, Hungarian voters are more likely to support populist parties in their countries the more they assess their governments take into account the views of other
government – 2.73% more likely for every point increase on the scale – which is something they have in common with polish supporters of populist parties (2.85% increase in probability). Polish and Lithuanian voters have in common that they are more likely to support populist parties, the worse they assess freedom and fairness of elections (3.78% and 6.95% respectively). In the end, Lithuanians are more probable to support their populist party the more they assess their government protects against poverty (7.26% for every point increase on the scale) and the more they believe courts treat everybody the same (4.96%). On the other hand, citizens are more likely to abstain from voting the better they assess equality before the law in Bulgaria, the worse than other parties (but similar to populist) they assess the freedom and fairness of elections in Poland, and the more they believe minority rights are protected in their countries (Poland and Slovakia)
With regards to interaction effects, we see that there is a joint effect of referendum indicators in two cases (Bulgaria and Czechia) and an interaction effect of taking into account other government’s views (Hungary). While the interaction effect found in Hungarian case could be considered substantively irrelevant, the referendum interaction effects seem to carry substantive weight. For this reason, I show and comment on these bellow, after which I move on to discuss the implications of these results for populism in East-Central Europe.
Figure 1. Interaction effect of referendum indicators in Bulgaria
Figure 1 shows the relationship between the assessment variable for the applicability of referendums in Bulgaria (horizontal axis) and probability of voting for populist parties in this country (vertical axis) when (a) normative referendum variable (how important…) has values one standard deviation above population mean (blue line), (b) normative referendum variable has values one standard deviation bellow population mean (red line) and (c) normative referendum variable has values equal to population mean. We see that for people who find it more important to have a direct say in politics by voting in referendums, as their perception of the presence of direct democracy in their country increases, so does the probability of voting for populist parties. By contrast, for people who find it less important to engage in politics through referendums, the probability of voting for populist parties decreases with the increase in the applicability estimation. In other words, the more these people see the presence of referendum practices the less likely they are to vote for populists.
Figure 2. Interaction effect of referendum indicators in Czechia
The setup for Figure 2 is the same as the one in Figure 1: assessment on horizontal axis, voting probability on vertical and colors depict different levels on the normative variable. Having said that, the interaction between referendum indicators in Czechia (Figure 2) exerts an effect that I generally expected from most of interaction terms as postulated in the H2. Put differently, those who find referendums very important are considerably more likely to vote for populist parties, the more they perceive the absence of referendums in their country. With regards to those with medium and low scores on the normative dimension, their voting probability stays roughly the same on different levels of applicability dimension. The implication of these results, together with results commented above, are discussed in the next section.
Discussion and conclusion
Previous research has shown that people who vote for populist parties are largely motivated by discontent or political dissatisfaction. Voting for those labeled as populists is, therefore, often considered to be a protest vote. Even though there have been comparative studies that explored the concept of democratic discontent in established democracies, such studies are lacking in East-Central European democracies where discontent and populism are even more widespread. For these reasons, my aim has been to fill this gap by analyzing the political competition between relevant populist and non-populist political parties in East-Central Europe and see whether voters’ evaluations of how their democracies function tell us something about populism in this region. In addition to relevant populist and non-populist parties, I have also added abstainers to understand where their position is in the context of democratic discontent, relative to the other two groups.
First of all, my findings suggest that – with regards to dissatisfaction with democratic system – we can hardly speak of populism monolithically in East-Central Europe. In other words, there is not enough empirical evidence to suggest there is a unified group of populist voters cross-nationally in the region. Country-by-country studies have further shown that, for instance, Lithuanian and Slovakian populist voters care more about protection against poverty, while Hungarian populist parties care more about free and fair elections (Table 5). There are also some opposing effects, such as the negative relationship between assessment of equality before the law and the probability of populist voting in Bulgaria and Hungary on the one hand, and the positive relationship between the same indicator and populist voting in Lithuania on the other. Moreover, interaction effects (Figure 1 and 2) just add to this picture: when having a more direct say on political issues is very important for citizens, the probability of voting for populist parties increases in Bulgaria if they think this possibility of directly affecting politics is present, but decreases in Czechia in the same scenario.
There are perhaps two reasons why the empirical support for H1 and H2 has been very limited. The first one is methodological, while the second one might be labeled more conceptual and requires further empirical scrutiny. First, the groups of populist and non-populist parties might be too heterogenous. On the one hand, when compared to West-European populism that is filled by radical right parties and, therefore, relatively homogenous, populist parties in CEE are not exclusively radical right, but also mainstream catch-all parties that even often hold political power (Stanley, 2017). On the other, the non-populist party group is ideologically very heterogenous. In this regard, I have run the analysis with and without the Communist Party of Bohemia and Moravia – the most radical non-populist party in my sample according to the party selection sources – to see if results differ either on the level of a single-country study or in the pooled model. There have been, however, no significant changes.
The second reason, as indicated above, might be more conceptual and it concerns the way how democracy is conceived of among the CEE populist parties electorate. In other words, I started with the protest voting theoretical background and focused on a protest expressed through dissatisfaction with democracy. When deriving hypotheses, I then largely relied on theories developed in established Western democracies while seeking support for my expectations in a democratic system with very different historical experience. This could also explain non-findings and serve as a suggestion for future studies. In other words, democracy might not be primarily associated with equality before the law or minority rights among populist voters in the CEE region. Reasons to believe that this might be the case come from voters of the main polish populist party which has weekend the Constitutional Tribunal (Pappas, 2016) or from Hungarian Fidesz party voters who continue to support the party whose leader openly rejects liberal democratic principles (Körösényi & Patkós, 2017).
Therefore, future studies on the topic may want to begin with better describing the concept before proceeding with making causal inferences. To this end, interviews or focus groups with populist parties’ voters would help defining indicators for measuring democracy in the CEE region. Further case studies could also focus on exploring why some indicators of the functioning of democracy are relevant in, say Lithuania, while others are in Hungary.
References
Akkerman, A., Zaslove, A., & Spruyt, B. (2017). ‘We the people’or ‘we the peoples’? A comparison of support for the populist radical right and populist radical left in the Netherlands. Swiss Political Science Review, 23(4), 377–403.
Arzheimer, K. (2009). Contextual factors and the extreme right vote in Western Europe, 1980–2002. American Journal of Political Science, 53(2), 259–275.
Bartels, L. M. (2005). Homer gets a tax cut: Inequality and public policy in the American mind. Perspectives on Politics, 3(1), 15–31.
Bértoa, F. C. (2013). Post-communist politics: On the divergence (and/or convergence) of East and West. Government and Opposition, 48(3), 398–433.
Bowler, S., Denemark, D., Donovan, T., & McDonnell, D. (2017). Right-wing populist party supporters: Dissatisfied but not direct democrats. European Journal of Political
Research, 56(1), 70–91.
Dassonneville, R., Blais, A., & Dejaeghere, Y. (2015). Staying with the party, switching or exiting? A comparative analysis of determinants of party switching and abstaining.
Journal of Elections, Public Opinion and Parties, 25(3), 387–405.
De Blasio, E., & Sorice, M. (2018). Populism between direct democracy and the technological myth. Palgrave Communications, 4(1), 1–11.
Dejaeghere, Y., & Dassonneville, R. (2017). A comparative investigation into the effects of party-system variables on party switching using individual-level data. Party Politics,
23(2), 110–123.
Dogan, M. (2005). Erosion of confidence in thirty European democracies. Comparative
Sociology, 4(1–2), 11–54.
Enyedi, Z., & Casal Bértoa, F. (2018). Institutionalization and De-institutionalization in Post-communist Party Systems. East European Politics and Societies, 32(3), 422–450.
Fukuyama, F. (2011). Poverty, Inequality, and Democracy: Dealing with Inequality. Journal
of Democracy, 22(3), 79–89.
Gandhi, J., & Przeworski, A. (2007). Authoritarian institutions and the survival of autocrats.
Comparative Political Studies, 40(11), 1279–1301.
Goodwin, M. (2011). Europe’s Radical Right: Support and Potential. Political Insight, 2(3), 4–7.
Haushofer, J., & Fehr, E. (2014). On the psychology of poverty. Science, 344(6186), 862– 867.
Hernández, E. (2018). Democratic discontent and support for mainstream and challenger parties: Democratic protest voting. European Union Politics, 19(3), 458–480. Hernández, E., & Kriesi, H. (2016). The electoral consequences of the financial and
economic crisis in Europe. European Journal of Political Research, 55(2), 203–224. Hooghe, L., & Marks, G. (2018). Cleavage theory meets Europe’s crises: Lipset, Rokkan, and
the transnational cleavage. Journal of European Public Policy, 25(1), 109–135. Ignazi, P. (1996). The crisis of parties and the rise of new political parties. Party Politics,
2(4), 549–566.
Inglehart, R. F., & Norris, P. (2016). Trump, Brexit, and the rise of populism: Economic
have-nots and cultural backlash.
Kehrberg, J. E. (2015). The demand side of support for radical right parties. Comparative
European Politics, 13(5), 553–576.
Kitschelt, H. (2002). Popular dissatisfaction with democracy: Populism and party systems. In
Democracies and the populist challenge (pp. 179–196). Springer.
Körösényi, A., & Patkós, V. (2017). Liberal and illiberal populism: The leadership of Berlusconi and Orbán. Corvinus Journal of Sociology and Social Policy, 8(3), 315– 337.
Kriesi, H. (2010). Restructuration of partisan politics and the emergence of a new cleavage based on values. West European Politics, 33(3), 673–685.
Kriesi, H. (2014). The populist challenge. West European Politics, 37(2), 361–378. Mainwaring, S., & Torcal, M. (2006). Party system institutionalization and party system
theory after the third wave of democratization. Handbook of Party Politics, 11(6), 204–227.
Mair, P. (2006). Ruling the void: The hollowing of western democracy. New Left Review, 42, 25–51.
Mair, P. (2009). Representative versus responsible government. MPIfG Working Paper 09/8. Cologne: Max Planck Institute for the Study of Societies
Mudde, C. (2009). Populist radical right parties in Europe. Cambridge University Press Cambridge.
Mudde, C. (2017). Politics at the Fringes?: Eastern Europe’s populists, racists, and extremists. In The Routledge handbook of East European politics (pp. 254–263). Routledge.
Oesch, D. (2008). Explaining workers’ support for right-wing populist parties in Western Europe: Evidence from Austria, Belgium, France, Norway, and Switzerland.
International Political Science Review, 29(3), 349–373.
Oesch, D., & Rennwald, L. (2018). Electoral competition in Europe’s new tripolar political space: Class voting for the left, centre-right and radical right. European Journal of
Political Research, 57(4), 783–807.
Ottemoeller, D., Marcus, R., & Mease, K. (2001). Popular definitions of democracy from Uganda, Madagascar, and Florida, USA. Journal of Asian and African Studies, 36(1), 113–132.
Pappas, T. S. (2016). Distinguishing liberal democracy’s challengers. Journal of Democracy,
27(4), 22–36.
Przeworski, A. (2018). Why bother with elections? John Wiley & Sons.
Rooduijn, M. (2018). What unites the voter bases of populist parties? Comparing the
electorates of 15 populist parties. European Political Science Review, 10(3), 351–368. Rooduijn, M., Van Kessel, S., Froio, C., Pirro, A., De Lange, S., Halikiopoulou, D., Lewis,
P., Mudde, C., & Taggart, P. (2019). The populist: An overview of populist, far right,
far left and eurosceptic parties in Europe.
Rydgren, J. (2008). Immigration sceptics, xenophobes or racists? Radical right-wing voting in six West European countries. European Journal of Political Research, 47(6), 737– 765.
Schumpeter, J. A. (1976). Capitalism, socialism and democracy (1942). J. Econ. Literature,
20, 1463.
Stanley, B. (2017). Populism in central and Eastern Europe. The Oxford Handbook of
Populism, 1, 140–158.
Stefanovic, D., & Evans, G. (2019). Multiple Winning Formulae? Far Right Voters and Parties in Eastern Europe. Europe-Asia Studies, 71(9), 1443–1473.
Učeň, P. (2007). Parties, populism, and anti-establishment politics in East Central Europe.
SAIS Review of International Affairs, 27(1), 49–62.
Van Biezen, I., Mair, P., & Poguntke, T. (2012). Going, going,... gone? The decline of party membership in contemporary Europe. European Journal of Political Research, 51(1), 24–56.
Voogd, R., & Dassonneville, R. (2018). Are the supporters of populist parties loyal voters? Dissatisfaction and stable voting for populist parties. Government and Opposition, 1– 22.
Whiteley, P. F. (2011). Is the party over? The decline of party activism and membership across the democratic world. Party Politics, 17(1), 21–44.
Zhirkov, K. (2014). Nativist but not alienated: A comparative perspective on the radical right vote in Western Europe. Party Politics, 20(2), 286–296.
Ziller, C., & Schübel, T. (2015). “The pure people” versus “the corrupt elite”? Political corruption, political trust and the success of radical right parties in Europe. Journal of
Appendix
Table 3. Descriptive statistics for independent variables
Variable N Min Max Mean SD
Fair elections importance 11802 -8.88 1.12 .01 1.75
Fair elections assessment 11467 -6.14 3.86 -.06 3.01
Courts importance 11871 -9.10 .90 -.01 1.69 Courts assessment 11543 -3.51 6.49 -.00 2.74 Referendums importance 11622 -8.40 1.60 .06 1.94 Referendums assessment 11329 -4.72 5.28 -.04 2.98 Poverty importance 11850 -8.57 1.43 -.04 2.10 Poverty assessment 11773 -2.77 7.23 -.02 2.56
Others government’s views
importance 10859 -6.51 3.49 -.04 2.55
Other government’s views
assessment 10248 -5.35 4.65 .01 2.54
Minority rights importance 11613 -7.98 2.02 -.04 2.27
Minority rights assessment 11186 -6.51 3.49 -.06 2.66
Political views importance 7374 -6.34 3.66 .03 2.49
Political views assessment 8946 -8.36 1.64 .01 1.84
Elections interaction 11324 -3.43 5.45 .08 .54
Courts interaction 11436 -4.61 3.19 -.03 .44
Referendums interaction 11130 -4.43 3.96 .02 .57
Poverty interaction 11634 -6.20 2.38 -.07 .52
Other government’s views
interaction 9870 -3.03 3.48 -.00 .74
Minority rights interaction 10982 -2.79 5.19 .04 .67
Political views interaction 7138 -3.06 5.30 .09 .49
Age 12108 -31.27 50.73 .01 17.73 Gender 12055 .00 1.00 .45 .49 Interest in politics 12044 -1.14 1.86 -.02 .84 EU unification 10539 -5.38 4.62 .04 2.68 Immigration 10882 -4.84 5.16 -.00 2.32 Education 12072 -2.87 3.13 .04 1.67 Trust in institutions 11348 -3.27 6.73 .05 2.07 Valid N (listwise) 4970
