alys l an ary is Deta Hu i ng How democratically elected politicians get away with autocratizing Hungary

IDENTITY, PARTISANSHIP, POLARIZATION How democratically elected politicians get away with autocratizing Hungary

IDENTITY, PARTISANSHIP, POLARIZATION Authors Elena Avramovska is a senior researcher for democracy, society and youth at the Friedrich-Ebert-Stiftung’s(FES) Office “Democracy of the Future”. Her research has been funded by the European Commission, the Swiss National Research Foundation and the Open Society Foundation. Avramovska has contributed to publications with Oxford University Press and Routledge. András Bíró-Nagy is the director of Policy Solutions and senior research fellow at the HUN-REN Centre for Social Sciences. His main areas of expertise include European integration, and the values of Hungarian society. His book,“The Social Democratic Parties in the Visegrád Countries”(co-edited with Ania Skrzypek), was published by Palgrave Macmillan in 2023. Ludwig List is a senior researcher for empirical social and trend research at the Friedrich-Ebert-Stiftung’s(FES) Office “Democracy of the Future”. His research focuses on income distribution, productivity and employment. Johanna Lutz is the director of the Friedrich-Ebert-Stiftung’s(FES) Office“Democracy of the Future”. She has worked on democratization and democracy support also in her previous roles as Deputy Head of the FES EU Office in Brussels and as President of the European Network of Political Foundations in Brussels. Milan Svolik is Professor of Political Science at Yale University. He studies the politics of authoritarian regimes, democratization and democratic backsliding. Svolik’s research has won awards and funding from the American Political Science Association, the Guggenheim Foundation and the National Science Foundation. Áron Szászi is an external expert at Policy Solutions and a public affairs consultant at Grayling Hungary. His expertise are political analysis, public opinion research and survey experiments. Currently, his professional interests lie in political polarization, Euroscepticism and public health-related attitudes. Acknowledgments The authors would like to thank their team members Filip Milačić, Krisztina Stefán and Margarete Lengger for their support. We would also like to thank Florian Müller for the fantastic graphic design. 2

How democratically elected politicians get away with autocratizing Hungary Table of Contents Executive Summary 4 Introduction 8 Research Design 10 Democratic knowledge and preferences 14 Voter Punishment for Democratic Violations 26 Which Undemocratic Positions Are Voters Most Willing and Least Willing to Punish? 33 Democratic attitudes and Punishment Rates 38 Socio-economic, demographic and psychological profile of democratically minded voters 40 Tolerating Democratic Violations 44 Polarization 56 Conclusion 58 References 61 Imprint 62 3

IDENTITY, PARTISANSHIP, POLARIZATION EXECUTIVE SUMMARY 1 Before testing whether Hungarian voters would trade off democracy for partisan or policy interests, we first assessed their understanding of democracy. We developed a measure to evaluate their levels of democratic competence, and our results indicate that the Hungarian respondents have lower democratic competence compared to the respondents from the other European countries we analyzed earlier. Democratic competence How high is the share of people who are democratically competent? Sweden Estonia Germany Spain Poland Serbia Ukraine Hungary 20% 40% 60% 85% 81% 80% 80% 78% 76% 75% 72% 80% 2 Would Hungarian voters electorally punish undemocratic political candidates? Yes, but less so than the voters in the other countries we analyzed. An undemocratic candidate would lose-4.8% of the vote share. Average punishment rates all countries All figures in percent Sweden Poland Serbia Estonia Germany Ukraine Spain Hungary punishment -2% -4% -4.8 -5.9 -6% -7.1 -7.9 -7.8 -7.5 -8% -9.2 -10.0 4

How democratically elected politicians get away with autocratizing Hungary 3 The Hungarian voters do not punish all democratic violations equally. They punish violations in the sphere of electoral fairness more than violations in the sphere of checks and balances or civil liberties. Average punishment by key elements of liberal democracy All figures in percent Electoral Fairness Checks and balances Civil liberties punishment -2% -4% -4.3 -5.8 -6% -8% -8.2 4 Not all Hungarian voters equally punish democratic violations. Older citizens, women, individuals with higher education and net income, and those with a more pessimistic outlook on economic prospects tend to be more punitive towards democratic violations. Characteristics of respondents who showed more willingness to punish undemocratic politicians Women (especially those who are democratically competent) Higher educated ODEMACY CR Higher income Older Pessimistic economic prospects 5

IDENTITY, PARTISANSHIP, POLARIZATION 5 When it comes to party support and punishment of democratic violations, our results show that the supporters of FideszKDNP punish democratic violations the least: Which party supporters are least likely to punish democratic violations? Punishment by select European parties, according to voting behavior, all figures in percent Ukraine Opposition Platform for Life Germany AfD Spain VOX and PP Sweden Sweden Democrats Estonia EKRE Serbia SNS Poland PiS Hungary Fidesz-KDNP punishment -1% -2% -2.7 -3% -3.4 -4% -4.0 -4.3 -4.9 -4.9 -4.6 -5% -5.4 6 While the respondents in all the countries we studied are“partisan first, democrats second”, party loyalty in Hungary is stronger than in the other European countries we analyzed. An undemocratic candidate that would otherwise on average lose-4.8% of the vote share in Hungary, would be rewarded with a gain of 35.9% of the vote share if that candidate belongs to the respondents’ preferred party. The general effect of co-partisanship Select European countries. All figures in percent 31.2 Final reward +30% compensation 36.0 +25% 21.7 19.9 18.0 +20% 15.9 15.6 15.4 +15% 11.5 +10% 27.6 22.9 23.5 23.7 25.1 29.1 +5% 21.5 0% punishment -4.8 Hungary -5.9 Spain -7.1 -9.2 Poland Ukraine Average punishment -7.8 -7.9 -7.5 Estonia Serbia Germany Compensation by voters’ party loyalty -10.0 Sweden -5% -10% 6

How democratically elected politicians get away with autocratizing Hungary 7 While in the other countries we studied earlier, the respondents were primarily willing to trade off democracy for identityrelated policies, the Hungarian voters would trade off democracy for all policies we investigated: with the strongest effect observed concerning the rights of same-sex couples. Trade-offs: what are democratic standards most likely to be exchanged for? All figures in percent punishment compensation Average 1 Punishment -4.8% Final reward Partisanship Same-sex couples' rights +14.9% +36.0% 2 10.1% 31.2% 3 Immigration +10.3% 5.5% Defense policy +10.0% 5.2% Education and income tax +9.6% 4.8% Energy taxation +9,0% 4.2% -5% How to read this graph: 1 The regular punishment of undemocratic behavior among left-leaning voters equals-4.8%. 0%+5%+10% 2 In this example,+36.0 percentage points is the absolute compensation rate that candidates receive if they belong to the voter's preferred party. +15%+20%+25%+30% 3 Relative compensation(or sometimes punishment) is the evolution of vote shares or the electoral reward/punishment that candidates ultimately receive when both the original punishment and compensation are taken into account. 8 We also examined whether polarization plays a role Which issues let politicians get away with when it comes to the voters’ propensity to trade off undemocratic behavior? democracy for specific policy issues. On the one hand, we asked the respondents how important these policies would be for them, and on the other, we measured the Issues displayed according to their importance to voters and degree of polarization Size of the bubble represents the average trade-off reward for each policy levels of polarization surrounding them. Our findings 8 indicate that, in comparison to the previously examined countries, Hungary and Ukraine exhibit the highest levels of polarization. Regarding specific policy domains, 7 our findings indicate that Hungarian respondents demonstrate the highest polarization concerning the rights of same-sex couples, followed by the policy area 6 Energy taxation Education/ income tax on taxes and education, immigration, energy taxes and Defense policy lastly, the area of defense policy. 5 more important (1 10) Immigration 4 3 0,2 0,3 Same-sex couples’ rights 0,4 0,5 0,6 more polarizing(0,0- 1,0) 7

IDENTITY, PARTISANSHIP, POLARIZATION INTRODUCTION Why do voters support candidates who violate democratic norms? Autocratization A process opposite of democratization that involves the weakening of institutional constraints, civil liberties and political rights. 1 Avramovska(2023) In recent years, Hungary has experienced continuous democratic decline. Once lauded as a democratic success story in post-communist Europe, Hungary has observed a significant erosion of democratic values and the rule of law. Since 2010, Viktor Orbán and his Fidesz party have exploited constitutional amendments to consolidate power, influencing the judiciary and imposing constraints on media freedom. Moreover, starting from 2015, Orbán has embraced far-right anti-migrant rhetoric, restricted LGBTQIA+ rights, and maintained close ties with Russia, despite Russia’s full-scale invasion of Ukraine(Scheppele 2022, Bernhard 2021, Ágh 2016, Cianetti et al. 2018, Vachudova 2020). While Hungary is among the countries that have experienced some of the sharpest processes of autocratization, it is not the only one. The slow dismantling of democratic institutions and norms, coupled with an embrace of illiberalism, has occurred in many countries around the world in recent years(Nord et al. 2024). However, despite the increasing cases of democratic erosion, scholars and experts have yet to reach a consensus on the factors that fuel the rise and persistence of illiberal, populist, and right-wing parties that gradually erode the democratic fabric of their countries. In the search for explanations for recent trends of democratic erosion, scholars have considered a variety of explanations, such as the rise of populist rhetoric, gradual institutional dismantling, or increasing polarization(Bartels 2023, Bermeo 2016, Visnovitz and Jenne 2021; Nalepa, Vanberg and Ciopris 2018; McCoy& Somer, 2019; Grzymala-Busse 2019, Grzymala-Busse et al. 2020). However, with some exceptions(e.g. Svolik 2019, Lewandowsky and Jankowsk 2023; Svolik et al. 2023; Saikkonen and Christensen 2023; Wunsch& Gessler 2023), scholars have paid less attention to the role of voters, especially outside the United States. If voters had to choose between supporting democracy and adhering to party loyalty or policy interests, what would they prefer? Who would put democracy above all else? While the diagnosis of democratic deficits may begin with an examination of the role of politicians, it must also include the crucial role of voters(Svolik et. al 2023). This report builds on our previous conjoint experiment 1 of seven European countries(Estonia, Germany, Poland, Serbia, Sweden, Spain and Ukraine) and looks in depth at the case of Hungary. It seeks to answer the question: Why do voters continue to support politicians who violate democratic principles? The study is based on a survey experiment conducted in Hungary with 1000 respondents. First, it aims to understand whether and under what circumstances voters are willing to vote for politicians who disregard democratic norms, as long as they belong to the voters’ preferred parties or pursue policies in line with their views. Second, it aims to explore whether voters’ willingness to prioritize policy alignment over democracy is increased by polarization. 8

How democratically elected politicians get away with autocratizing Hungary STRUCTURE OF THE STUDY Our study is structured as follows: First, our study focuses on assessing the respondents’ knowledge of and attitudes toward democracy before addressing the crucial question of whether voters in Hungary would be willing to compromise democracy for alternative priorities. To this end, we present a series of specific questions that assess their understanding and command of democratic principles and processes. In particular, this section examines three aspects: 1) support for democracy, 2) satisfaction with democracy, and 3) knowledge of democracy. It also provides insight into the profile of respondents who value democracy most by analyzing variables such as gender, age, net income, education, and media consumption. Second, we explore the degree to which survey participants hold political candidates responsible for suggesting actions that defy democratic standards. We also analyze the characteristics of voters who are more inclined to hold such breaches accountable. Additionally, we investigate how partisan allegiances influence the accountability of individuals for undemocratic conduct, while examining various other factors such as age, gender, education, political awareness, religious affiliation, and more. Through this examination, we aim to gain deeper insights into the specific demographics and characteristics that contribute to a greater propensity to punish democratic transgressions. Building on the previous sections, which explored different dimensions of punishing democratic transgressions, then we examine the conditions under which voters may be willing to compromise democracy in favor of other interests. This investigation will focus on two key aspects: a) First, we will examine the impact of party loyalty on voters’ decisions to overlook democratic violations and continue to support their preferred political candidates. By analyzing the role of partisanship in shaping voters’ responses, we can gain insights into the extent to which party affiliation may affect their willingness to prioritize other considerations over democratic principles. b) Secondly, we will explore whether voters are willing to compromise democracy in exchange for specific policy positions. By analyzing voters’ attitudes to different policy areas, we can gauge the potential trade-offs they are willing to make, and shed light on the relative importance they attach to policy issues versus democratic values. Through these findings, we aim to gain a better understanding of the circumstances and factors that may influence voters’ willingness to trade democracy for alternative interests, providing valuable insights into the dynamics between democratic principles and voter decision-making. Finally, the study explores the relationship between polarization and the extent to which voters are willing to overlook undemocratic behavior by politicians in exchange for policies that align with their preferred social, economic, or foreign policy agendas. By examining this dynamic, we aim to assess whether polarization acts as an amplifier, increasing voters’ willingness to forgive and support politicians who engage in undemocratic behavior, as long as their policy positions are consistent with their own beliefs and interests. These findings will shed light on the complex interplay between political polarization, voter behavior, and the prioritization of policy objectives over democratic principles. It will also explore whether voters are more willing to compromise democracy in exchange for policy benefits in areas that generate polarization or in areas that voters consider“important” but are not inherently socially divisive. This exploration will provide valuable insights into how polarization may shape voter attitudes toward democratic norms and the extent to which policy considerations may override democratic values in politically charged contexts. 9

IDENTITY, PARTISANSHIP, POLARIZATION RESEARCH DESIGN DATA COLLECTION Our data set is based on a commissioned representative survey composed of two parts: a questionnaire query and a candidate choice experiment. The survey was conducted by the Hungarian polling company,“Závecz Research,” between June 27, 2023 and July 11, 2023 and had a total of 1014 respondents. Interviews were conducted online and had an average length of 25 minutes. Quota selection was done according to the following criteria of representativeness: gender, age, region, and education. CANDIDATE CHOICE EXPERIMENTS In addition to the standard survey questions – support for democracy, policy and party preferences, interest in politics, economic performance, etc. – our survey also included two sets of candidate choice experiments(ten choices per set). In each round, respondents had to choose between two hypothetical candidates, one from Fidesz-KDNP and one from a different party. Each candidate was described with four randomly assigned attributes: party affiliation, two relevant policy positions, and either a neutral or an undemocratic position. In the latter case, a candidate was portrayed as supporting a measure that violates one of three key principles of a liberal democracy: electoral fairness, checks and balances, and civil liberties. Respondents were then asked two questions about the candidates: which candidate they preferred and how likely they would be to vote for the chosen candidate in the next election. In total, each respondent participated in 20 different candidate choice scenarios, resulting in a grand total of 20,000 experiment combinations. These experiments allowed us to isolate the effect of a candidate’s attempt to undermine democracy on their electoral prospects. Example of the candidate profile from the Hungary survey 10

How democratically elected politicians get away with autocratizing Hungary POLICY POSITIONS In designing the policy positions of the political candidates, we followed several criteria: Positions should be uniquely classifiable along a single dimension(e.g., economic left-right and pro-/ anti-immigration); the expected distribution of respondent’s most preferred policies should span all policy positions and reflect the underlying distribution of preferences in the electorate; and the selection and content of issue areas should reflect the most prominent contentious public policy issues. Taxes and education Raise personal income taxes and use the revenue to increase education budget Keep personal income taxes and education budget as they are Cut personal income taxes and reduce education budget accordingly Energy taxation Raise taxes on coal and gas to lower the price of solar and wind energy Lower taxes on coal and gas even if it results in lesser use of solar and wind energy Immigration Allow immigration regardless of the country of origin Allow immigration from the EU, but only family-based immigration from outside of the EU Ban immigration from outside of the EU Ban all immigration regardless of the country of origin The use of minority languages* Only hungarian should be used when communicating with state offices Hungarian speakers should have the right to use their language when communicating with state offices Same-sex couples’ rights National versus EU-level policy on defense Same-sex couples should have the right to marry and adopt children Defense policy for Hungary should be decided at the EU level Same-sex couples should have the right to marry, but not adopt children Same-sex couples should not have the right to marry or adopt children Defense policy for Hungary should be decided jointly by hungarian and EU institutions Defense policy for Hungary should be decided by hungarian institutions alone * Only used in our previous conjoint experiment in Estonia, Serbia, Spain and Ukraine. ** Only used in our previous conjoint experiment in Serbia and Ukraine. EU RUS Relations with the EU and Russia** We should have closer relations with the EU and distance ourselves from Russia We should have close relations with both the EU and Russia We should have closer relations with the Russia and distance ourselves from the EU 11

IDENTITY, PARTISANSHIP, POLARIZATION UNDEMOCRATIC POSITIONS In formulating the candidates’ undemocratic positions, we used the following criteria: First, they had to include violations of basic democratic principles. Second, they should approximate practices that have been employed or attempted by governments and politicians across Europe to undermine the democratic process. Third, the positions should be adopted with the intention of gaining or retaining power, rather than being pursued for their own sake. Fourth, the range should cover positions that are legal but undemocratic, as well as those that are both illegal and undemocratic, reflecting the incremental nature of autocratization. Due to the special political situation present in Hungary, and the dominating presence of Fidesz, we decided to adapt our method used in our previous two studies. More precisely, we did include one Fidesz candidate in every candidate choice experiment. This allowed us to go further in our analysis and distinguish between scenarios in which the respondents’ choice was primarily about voting for their favorite party and those scenarios in which the aim was primarily to prevent the Fidesz candidate from winning the vote. In addition, we aimed to present these positions in a way that avoided any obvious or normative language, such as“undemocratic”,“unconstitutional”, or“violation”. Our aim was to allow respondents to decide for themselves whether a position was a violation of democratic principles, thereby avoiding the potential objection that their punitive response might be influenced by the style of presentation rather than by the content. In selecting undemocratic positions, we followed classic concepts of democratisation by Dahl(1971, 1989) and others (Linz and Stepan 1996), and thus selected undemocratic positions in relation to three key tenets of liberal democracy: electoral fairness, checks and balances, and civil liberties. Electoral fairness Checks and balances Civil liberties “We should direct infrastructure spending to districts that voted for our party.” Opposition version:“If we win, we should direct infrastructure spending to districts that voted for our party.” Key democratic concept: voter autonomy, the government should not abuse public resources for electoral purposes “Laws should be passed without parliamentary debate if criticised by the opposition.” Only plausible for candidates affiliated with the government Key democratic concept: legislative checks on the executive “Encouraged his/her supporters to violently disrupt campaign rallies of his/her political opponents.” Key democratic concept: freedom of assembly “Said the government should discipline judges who publicly criticise it.” Key democratic concept: judicial independence “Said the government should prosecute journalists who accuse the prime minister/president of corruption.” Only plausible for candidates affiliated with the government Key democratic concept: freedom of speech “Supported a proposal for the government to monitor politically critical posts on social media.” Key democratic concept: freedom of speech, right to privacy “Supported a ban on foreign funding for domestic non-governmental organizations critical of the government.” “Supported a ban on foreign funding for labor unions critical of the government” Key democratic concept: freedom of association, speech 12

How democratically elected politicians get away with autocratizing Hungary By randomly assigning undemocratic positions, we were able to compare the vote shares of undemocratic candidates with those of democratic but otherwise identical candidates. A decline in electoral support for undemocratic candidates serves as a measure of the punishment that voters are willing to inflict in defense of democracy. As a result, this research design allowed us to address three fundamental questions regarding the role of ordinary citizens in processes of autocratization: 1. Do voters hold democracy in high regard and value it enough to punish politicians who violate democratic principles? 2. Are voters willing to sanction politicians who violate democratic rules even if doing so means voting against their own political interests or party loyalties? 3. Do voters exhibit a higher tolerance for undemocratic behavior in polarized societies, or is such behavior not correlated with the degree of polarization? CANDIDATE CHOICE EXPERIMENTS VERSUS ATTITUDINAL QUESTIONS ABOUT SUPPORT FOR DEMOCRACY When political scientists assess the global public’s commitment to democracy, they often use direct questions that ask about attitudes towards democracy as a system of government. Recent developments in various countries, however, have raised concerns about the validity of such traditional measures of support for democracy. Despite showing strong commitment to democracy in response to traditional attitudinal questions about public support, many countries have experienced autocratization. This discrepancy may be due to the likelihood that direct questions about support for democracy elicit socially desirable and politically correct responses. The Experiment By randomly assigning undemocratic positions, we were able to compare the vote shares of undemocratic candidates with those of democratic but otherwise identical candidates. A decline in electoral support for undemocratic candidates serves as a measure of the punishment that voters are willing to inflict in defense of democracy. 13

IDENTITY, PARTISANSHIP, POLARIZATION DEMOCRATIC KNOWLEDGE AND PREFERENCES This chapter focuses on assessing respondents’ democratic attitudes and competence. It does so by analyzing a series of specific questions that assess their preferences, understanding and knowledge of democratic principles and processes. In addition to placing the results in an international context, this chapter also presents partisan and socio-demographic patterns. SUPPORT FOR DEMOCRACY THE IMPORTANCE OF LIVING IN A DEMOCRACY As in previous FES surveys, Hungarian respondents were asked to evaluate how important it was for them to live in a democratically governed country. The respondents could mark their opinion on a scale from 1 to 10, where 1 meant that it is not at all important and 10 reflected that it is absolutely important for them. The average response in Hungary was 8.3(8 out of 10 respondents said it was somewhat or very important for them to live in a democracy). The average was the same in Poland, slightly lower in Serbia(8.2) and Estonia(8.1), and higher in Spain(8.4), Ukraine(8.5), Germany(8.5) and Sweden(8.8). “How important is it for you to live in a country that is governed democratically?” Not at all important Mean values Absolutely important Sweden 8.8 Germany 8.5 Ukraine 8.5 Spain 8.4 Hungary 8.3 Poland 8.3 Serbia 8.2 Estonia 8.1 1 2 3 4 5 6 7 8 9 10 8.3 All respondents(average) There are significant differences according to political preferences. While only 60% of voters of the ruling party consider it very important to live in a democratically governed country, this figure rises to 89% among voters of the six-party opposition (supporters of United for Hungary in the 2022 elections). It is notable that even among voters of the far-right Our Homeland, more people say it is important to live in a democratically governed country than among Fidesz voters. 14

How democratically elected politicians get away with autocratizing Hungary “How important is it for you to live in a country that is governed democratically?” Proportion of respondents, according to voting behavior, all figures in percent 80% All respondents 6 14 13 60% 50% 40% 67 20% FideszKDNP United for Hungary Our Homeland Hungarian TwoTailed Dog Party(MKKP) 7 5 510 14 15 5 17 18 89 79 60 68 Other party/No answer 6 15 11 Non-voters of the ´22 elections 9 25 13 68 53 1 corresponds to“not at all important”, 10“very important” 1-3 4-5 6-7 8-10 Due to rounding, totals may not correspond with the sum of the separate figures. The breakdown by media consumption also highlights some major differences. 2 The proportion of people who think it is very important to live in a democratically governed country is 77% among those who mainly consume media critical of the government, and only 61% among those who mainly consume pro-government media. “How important is it for you to live in a country that is governed democratically?” Proportion of respondents, according to media consumption, all figures in percent All respondents 6 14 13 67 Mainly pro-government media consumer 5 18 17 61 Mixed-media consumer 6 14 15 66 Mainly government-critical media consumer 3 9 10 77 No media-consumer/ No answer 11 18 12 58 20% (1-3) 40% 50% 60% (4-5)(6-7)(8-10) 1 corresponds to“not at all important”, 10“very important” Due to rounding, totals may not correspond with the sum of the separate figures. 80% Living in a democracy is important to more people in the older generations. The proportion of those who said they strongly preferred living in a democracy was 82% and 87% respectively in the two oldest age cohorts(respondents aged 56-65 and older), while it was 52% and 57% in the two youngest age groups(respondents aged 18-34 and 35-45 respectively). There are only small gender differences on this question, with women slightly more likely than men to consider living in a democratic country very important(69% vs. 64%). 2 We measured partisan media consumption by asking respondents how often they consume 14 different media outlets. Among the outlets listed, the ones critical of the government were Partizán, ATV, RTL Klub, Telex, 24.hu, 444, Mérce and Hvg.hu. The pro-government outlets were M1, Magyar Nemzet, Kossuth Rádió, TV2, ORIGO and Index.hu. We created categories of media. We defined“pro-government media consumers” as those respondents who regularly(daily or several times a week) consumed at least twice as many of these media as those who were critical of the government. Similarly, we considered“anti-government media consumers” to be those who consumed at least two times as many media outlets that were critical of the government as those that were pro-government. We considered mixed-media consumers to be those in between, while those who said they did not regularly follow the listed mainstream outlets were labelled“no media consumers.” 15

IDENTITY, PARTISANSHIP, POLARIZATION THE SUPERIORITY OF DEMOCRACY “Democracy is better than any other form of government” All figures in percent Sweden 50 43 93 Germany 47 45 92 Spain 44 45 89 Poland 33 56 89 Serbia 24 61 85 Estonia 23 61 84 Hungary 30 53 83 Ukraine 15 38 53 20% 40% strongly agree agree Due to rounding, totals may not correspond with the sum of the separate figures. 60% 80% Media consumption 24% of Hungarians who reportedly abdicated media consumption disagree that“Democracy is better than any other form of government”. We also asked respondents whether they thought democracy was better than any other form of government. A large majority of Hungarian respondents(83%) agreed that democracy is a superior form of government, while less than a fifth(17%) disagreed. The proportion of the pro-democracy camp in Hungary was similar to that in Estonia(84%) and Serbia(85%). Slightly more people considered democracy to be the best political system in Poland(89%), Spain(89%), Germany(92%), and Sweden(92%), while significantly fewer people shared this view in Ukraine(53%). Those with a pro-democracy attitude were in the majority in all political groups, even among non-voters(72%), but the distribution of the responses shows clear partisan differences on this question. Among those who voted for the opposition alliance in the 2022 Hungarian parliamentary elections, 9 out of 10 respondents had a positive view on democracy “Democracy may have problems, but it is better than any other form of government.” Proportion of respondents, according to media consumption, all figures in percent All respondents 13 80% Mainly progovernment media consumer 12 Mixedmedia consumer 9 Mainly governmentcritical media consumer 13 No media consumer/ No answer 6 18 60% 53 54 65 43 50% 53 40% 20% 40 30 29 24 22 Due to rounding, totals may not correspond with the sum of the separate figures. 1 corresponds to“strongly disagree”, 10“strongly agree” 1-3 4-5 6-7 8-10 16

How democratically elected politicians get away with autocratizing Hungary “Democracy may have problems, but it is better than any other form of government.” Proportion of respondents, according to voting behavior, all figures in percent All respondents 13 80% 60% 53 50% 40% FideszKDNP United for Hungary Hungarian TwoOur Tailed Dog Homeland Party(MKKP) 8 6 7 10 20 37 62 61 47 Other Non-voters party/No of the ´22 answer elections 6 24 22 47 57 20% 30 53 27 25 27 26 15 Due to rounding, totals may not correspond with the sum of the separate figures. 1 corresponds to“strongly disagree”, 10“strongly agree” 1-3 4-5 Support for Democracy 6-7 The majority of Hungarian citizens think democracy is 8-10 better than any other form of government. (90%). Half of opposition supporters(53%) strongly agreed with the superiority of democracy and another 37% said they tended to agree. The proportion of“strongly pro-democracy” opinions was much lower among all other political groups(15-27%). However, the overall pro-democracy camp was similarly high among government party voters(89%) and Two-Tailed Dog Party voters(88%), and somewhat lower among far-right Our Homeland voters(73%) and non-voters(72%). Although the pro-democracy view had an absolute majority across all education groups (ranging from 77% to 91%), opinions on democracy show a clear educational pattern. The proportion of respondents who disagreed that democracy was the best form of government was highest among respondents among the less educated(23%) and lowest among respondents with a higher education degree(9%). Slightly fewer female respondents thought that democracy is better than other regime types, despite its drawbacks(79% vs 87%). “Democracy may have problems, but it is better than any other form of government.” Proportion of respondents, according to gender, all figures in percent All respondents 13 80% Male 11 Female 15 1 corresponds to “strongly disagree”, 10“strongly agree” 1-3 4-5 60% 53 50% 40% 54 52 6-7 20% 30 33 27 8-10 Due to rounding, totals may not correspond with the sum of the separate figures. 17

IDENTITY, PARTISANSHIP, POLARIZATION THE EVALUATION OF DIFFERENT FORMS OF GOVERNMENT Next to democracy, respondents were also asked how they viewed different forms of government. Hungarians had the worst opinion of military dictatorships(85% thought they were bad and only 15% thought they were good). A large majority (76%) also disapproved of an autocratic system without an elected legislative body headed by a strong leader, although the proportion of supporters was slightly higher(24%). Two-thirds of respondents(68%), however, had a positive view of a potential government of experts, while only one-third of the respondents considered this alternative to democracy to be a bad form of government. Democracy received the highest support from Hungarians(75%), but a quarter of respondents said they were against democracy(25%). Different forms of government Response options from 1(very bad) to 10(very good). All figures in percent What do you think about having the army rule? 50 35 What do you think about having a strong leader who does not 40 36 have to bother with Parliament or elections? What do you think about having experts, not government, make decisions for the country? 9 24 49 What do you think about having a democratic political system? 8 17 43 20% 40% 50% 60% (1-3) very bad(4-5) fairly bad(6-7) fairly good(8-10) very good Due to rounding, totals may not correspond with the sum of the separate figures. 12 3 19 5 19 33 80% The level of disapproval of authoritarian leadership in Hungary (76%) is similar to that in Poland(76%) and Serbia(77%). In Ukraine(79%), Estonia(81%), Germany(86%), Spain(89%) and Sweden(93%), more people disapprove of a strong leader’s autocracy. “How good or bad do you feel about having a strong leader who does not have to bother with Parliament or elections?” All figures in percent Sweden Germany Spain Poland 65 57 55 43 20 12 3 25 15 3 21 17 6 32 20 5 Hungary 40 Serbia 40 36 19 5 33 21 7 Estonia 26 34 32 7 Ukraine 20 30 35 15 20% 40% 60% 80% very bad fairly bad fairly good very good Due to rounding, totals may not correspond with the sum of the separate figures. “Having a democratic political system is either‘very good’ or‘fairly good’” All figures in percent Sweden 64 28 93 Spain 53 36 89 Germany 47 39 86 Estonia 33 48 81 Ukraine 33 46 79 Serbia 29 49 77 Poland 36 40 76 Hungary 33 43 76 20% 40% 60% 80% very good fairly good Due to rounding, totals may not correspond with the sum of the separate figures. 18

How democratically elected politicians get away with autocratizing Hungary While there is a clear pro-democracy majority in Hungary, it is important to note that Hungary has one of the lowest levels of positive views of democracy(76%) among the European countries surveyed. Notably, the size of the pro-democracy camp is fairly similar in Poland(77%) and Serbia(78%), and slightly higher in Ukraine(80%) and Estonia (82%). There is a small gap on this question between post-socialist countries and Western European countries, where about 9 out of 10 respondents have a favorable view of democracy(87% in Germany, 90% in Spain, and 93% in Sweden). The differences are even greater when strong preferences are compared: for example, the proportion of those who consider having a democratic political system to be very good is 32% higher in Sweden than in Hungary. SATISFACTION WITH DEMOCRACY Our data reflect widespread dissatisfaction with the way democracy works in Hungary. About two-thirds of respondents were dissatisfied with the state of democracy in Hungary(67%) and thought that Hungary was not governed democratically(64%). Perceptions of the current state of Hungarian democracy were clearly determined by political affiliation and media consumption. Within the group of respondents who voted for the six-party opposition alliance in the 2022 elections, the overwhelming majority(94% and 93% respectively) were dissatisfied with the state of Hungarian democracy and believed that Hungary was being governed undemocratically. Far fewer people in the pro-government camp held similar views, but even four in ten Fidesz voters were dissatisfied with the quality of democracy(42%) and felt that the country was governed undemocratically(37%). This finding is in line with a 2022 post-election study 3 by Policy Solutions, which found that a third of Fidesz voters believed that the Orbán government could not be replaced democratically and a fifth of them did not think the election was free and fair. Our study confirms that the majority of Hungarians view the current state of democracy through a partisan lense, but a significant proportion of pro-government voters are aware of the undemocratic ways in which Fidesz operates. The extent of the differences between respondents based on their media consumption on this question is similar to the partisan divide in Hungarian society. The majority of those who mainly consume media critical of the government were dissatisfied(84%) and critical(82%) of the quality of Hungarian democracy, while only the minority of pro-government media consumers held similar views(39% and 35%). The majority of politically disengaged respondents – both those who did not vote in the 2022 elections (74% and 73%) and those who do not consume mainstream media platforms(78% and 74%) – were more critical about the extent to which democratic principles are followed in Hungary. Authoritarian personality traits were associated with perceptions of Hungarian democracy. While respondents who were not authoritarian according to our categori sation were overwhelmingly critical(79% were dissatisfied with democracy and 78% did not think Hungary was governed democratically), those who were highly authoritarian were more positive on this question(54% and 59% respectively). Dissatisfaction with the quality of democracy Hungarians consider democracy important, but are dissastisfied with its quality. 3 Policy Solutions(2022) 19

IDENTITY, PARTISANSHIP, POLARIZATION There are slightly(3-4%) more female respondents than male who are dissatisfied with the state of democracy and think that Hungary is not governed democratically. Furthermore, older respondents and those with a university degree are the most likely to be critical of democracy in Hungary. “How satisfied are you with the way democracy works in your country?” All figures in percent Male 45 21 18 17 Female 48 21 15 16 20% 40% 60% 80% not satisfied rather not satisfied rather satisfied satisfied Due to rounding, totals may not correspond with the sum of the separate figures. “How satisfied are you with the way democracy works in your country?” All figures in percent 18-29 years old 30-39 40-49 50-65 66 or older 39 42 39 55 62 27 23 27 18 22 20 13 8 99 12 13 19 24 20 20% 40% 60% 80% not satisfied rather not satisfied rather satisfied satisfied Due to rounding, totals may not correspond with the sum of the separate figures. “How democratically is your country being governed today?” All figures in percent Male 42 20 18 20 Female 45 21 17 18 20% 40% 60% 80% not democratic rather nondemocratic rather democratic democratic Due to rounding, totals may not correspond with the sum of the separate figures. “How democratically is your country being governed today?” All figures in percent 18-29 years old 30-39 40-49 50-65 66 or older 36 39 38 53 58 26 21 16 29 21 12 24 15 23 11 14 21 7 11 24 20% 40% 60% 80% not democratic rather nondemocratic rather democratic democratic Due to rounding, totals may not correspond with the sum of the separate figures. DEMOCRATIC COMPETENCE Before delving into the experimental segment of our survey, which assesses whether voters are willing to prioritize other interests over democracy, we initially explored their level of democratic competency. To that end, participants in our research were asked to rate 14 different hypothetical political statements. Some of these statements were consistent, and others violated key democratic principles. Respondents were asked to rate these statements on a scale from 1(not democratic at all) to 10(fully democratic). These statements were designed to deliberately mirror the undemocratic positions taken by candidates in our candidate choice experiments, allowing us to test our respondents’ understanding of democracy. We define a“democratically competent” respondent as one who, on average, rates the four democratic items as more democratic than the ten undemocratic items present in the survey. Compared with other European countries where this indicator was measured, democratic competence was lowest in Hungary, where 72% of respondents were democratically competent. The proportion of democratically competent respondents in Hungary was 3% lower than in the second lowest country(Ukraine, 77%) and 13% lower than in the most democratically competent country(Sweden, 85%). The level of democratic competence seems to correlate with the actual state of democracy in a given country. Countries with fewer democratically competent respondents are also those with democratic deficits. At the same time, democratic competence 20

How democratically elected politicians get away with autocratizing Hungary is on average higher in more democratic countries(as measured by the V-Dem Index of Liberal Democracy). It is worth noting, however, that democratic competence shows much less variation than institutional quality. Deviations in“democratic competence” might also indicate that some Hungarians might have a different conception of democracy(Anghel& Jones 2024). Democratic competence How high is the share of people who are democratically competent? Sweden Estonia Germany Spain Poland Serbia Ukraine Hungary 20% 40% 60% 85% 81% 80% 80% 78% 76% 75% 72% 80% Democratic competency versus V-Dem Liberal Democracy Index(2022) 4 Countries displayed according to respondent’s democratic competency and country’s V-Dem Liberal Democracy Index(2022) rating 1.0 Germany Estonia Sweden 0.8 Spain 0.6 0.4 Hungary Poland Serbia 0.2 Ukraine V-Dem Liberal Democracy Index (2022) (0 1) 0.0 70% 75% 80% 85% 90% Share of democratically competent respondents(0- 100%) 4 Varieties-of-Democracy(2022) 21

IDENTITY, PARTISANSHIP, POLARIZATION The proportion of democratically competent respondents was well below average among the highly religious(55%), the politically disengaged (non-voters: 63%, non-mainstream media consumers: 63%), and the less educated(65%). At the same time, the level of democratic competence was above average among non-religious respondents(80%), opposition voters(90%), Two-Tailed Dog Party voters(93%), those who consume mainly media critical of the government(81%), those with a higher education degree(84%), and respondents without the characteristics of an authoritarian personality (87%). Democratic competence Hungarians have the lowest democratic competence among the countries we studied. Democratic competence Hungary Proportion of democratic competent respondents, according to voting behavior, all figures in percent FideszKDNP United for Hungary Our Homeland Hungarian TwoTailed Dog Party(MKKP) Other Non-voters All party/No of the ´22 respondents answer elections 10 7 28 31 29 80% 36 37 Democratically incompetent 60% 50% 40% 72 20% 90 93 69 71 64 63 Democratically competent Due to rounding, totals may not correspond with the sum of the separate figures. Democratic competence Hungary Proportion of democratic competent respondents, according to media consumption, all figures in percent All respondents Mainly progovernment media consumer Mixed mediaconsumer Mainly governmentcritical media consumer Not mediaconsumer/ No answer 28 80% 19 29 30 Democratically 37 incompetent 60% 50% 40% 72 20% 81 71 70 63 Democratically competent Due to rounding, totals may not correspond with the sum of the separate figures. 22

How democratically elected politicians get away with autocratizing Hungary Democratic competence was lowest among the youngest age group(18-34 years old, 65%), and highest among the pre-retired(56-65 years old, 85%), with the other age groups falling somewhere in between(69%-76%). Seven percent more male respondents were considered demo cratically competent than women(76% vs. 69%). Respondents in the lowest net income group were the least democratically competent (64%), while respondents in the highest net income group were the most competent(96% competence rate). Democratic competence Hungary Proportion of democratic competent respondents, according to education, all figures in percent All respondents Lower education Secondary education with completed leaving exam University/ college degree 28 80% 23 16 35 Democratically incompetent Democratic Competence and Age The levels of democratic competence are the lowest among the youngest age group (18-34 years old). 60% 50% 40% 72 20% 77 84 65 Democratically competent Due to rounding, totals may not correspond with the sum of the separate figures. “How democratic do you think the following practice is?” Response options from 1(not democratic at all) to 10(completely democratic). All figures in percent The governments adopts laws without a debate in the parliament. Journalists are prosecuted for criticising the government. The country’s judiciary is staffed with individuals loyal to the governing party. The president encourages his supporters to disrupt opposition campaign rallies. The government cut spending in regions that supported the opposition in the last election. The country´s security agencies collect collect data on their citizens´ internet activity. Unelected experts govern the country. The government banned civil society organizations that receive funds from abroad from operating inside the country. The military removed a corrupt president. 61 60 60 58 58 58 49 48 24 25 18 11 10 21 11 7 20 12 9 21 12 9 22 12 8 21 11 11 21 14 15 24 15 13 18 33 20% 40% 50% 60% not democratic rather nondemocratic rather democratic democratic Due to rounding, totals may not correspond with the sum of the separate figures. 80% 23

IDENTITY, PARTISANSHIP, POLARIZATION “How democratic do you think the following practice is?” Response options from 1(not democratic at all) to 10(completely democratic). All figures in percent Opposition parties organize protests against 20 the government. The countries high court ruled, that a government 18 policy was not constitutional. 24 15 22 18 The prime minister conceded a narrow election defeat. 15 18 14 Journalists frequently disagree with the 14 21 16 presidents policies. 20% 40% 50% 60% not democratic rather nondemocratic rather democratic democratic Due to rounding, totals may not correspond with the sum of the separate figures. 41 42 53 49 80% The majority of respondents recognised which events were indeed undemocratic and which were not. There was one exception, however: a military coup against a corrupt president, which was considered to be somewhat or fully democratic by half of the respondents. A third of respondents considered it to be democratic(33%). Among the democratic statements, the highest rate of misjudgement was found for demonstrations organised by the opposition, with 44% of respondents considering this form of political action to be somewhat or completely undemocratic. A similar proportion of respondents(40%) felt that the Constitutional Court’s review of government policy was undemocratic. About a third of respondents thought it was undemocratic for a prime minister to admit defeat in an election(33%) or for the media to criticise the government(35%). On the latter, the difference between people with different media consumption habits was particularly large. 75% of consumers of media that are predominantly critical of the government considered it fully democratic for journalists to criticise the government, compared to 54% of respondents in the pro-government media bubble, 46% of those who follow heterogeneous sources and 59% of those who do not consume mainstream media. “How democratic do you find the practice of the military removing a corrupt president?” Share of respondents rating this statement with a 9 or 10 on the scale from 1 to 10. Hungary Poland Serbia Spain Germany Estonia Ukraine Sweden 13 10% Due to rounding, totals may not correspond with the sum of the separate figures. 17 16 15 15 22 21 20% 33 30% 24

How democratically elected politicians get away with autocratizing Hungary How democratic do you think the following practice is? “Journalists are prosecuted for criticizing the government.” Proportion of respondents, according to media consumption, all figures in percent Mainly pro-government media consumer Mixed-media consumer Mainly government-critical media consumer No media consumer/ No answer 80% 54 46 59 60% 75 50% 40% 25 25 25 20% 10 19 15 12 11 10 6 3 Due to rounding, totals may not correspond with the sum of the separate figures. 1 corresponds to “not democratic”, 10“democratic” 1-3 not democratic 4-5 rather undemocratic 6-7 rather democratic 8-10 democratic Main Takeaways: In Hungary, while there is widespread acknowledgment of the importance of democracy and its superiority over other forms of government, there is also prevalent dissatisfaction with its current functioning. Pro-democracy sentiments prevail across party lines, yet significant disparities exist, with a majority of opposition voters expressing discontent compared to the opposite view held by many Fidesz supporters. Additionally, the research highlights that compared to the other countries we previously analyzed, Hungarians have a lower democratic competence. 25

IDENTITY, PARTISANSHIP, POLARIZATION VOTER PUNISHMENT FOR DEMOCRATIC VIOLATIONS ARE VOTERS WILLING TO PUNISH UNDEMOCRATIC BEHAVIOR? Punishment “Punishment” is defined as a loss of vote share that a politician experiences for violating democratic Hungary: -4.8% Average punishment Hungarians punish democratic violations less than the voters in the other countries we analyzed. This is the average punishment corresponding to the share of voters willing to defect from an otherwise favored candidate once they adopt an undemocratic position. In the following sections, we make use of regression analyses to calculate the level of punishments while taking into consideration a range of other aspects(e.g., whether the candidate was the respondent’s co-partisan or whether the candidate proposed the respondent’s favorite economic policy). The previous chapter showed that a significant proportion of Hungarian respondents in this study express strong declarative support for democracy. In this chapter, we examine whether this support leads voters to hold politicians accountable for violating democratic norms by not voting for them. This question is of paramount importance for understanding the contemporary erosion of democracies. In Hungary, as elsewhere in Europe and beyond, the erosion of democratic principles is often gradual and incremental. At the ballot box, voters are usually not faced with a stark choice between democracy and autocracy, but with the dilemma of whether to continue supporting politicians who under mine democratic values. This predicament arises even when these politicians present appealing policy proposals or come from the voters’ preferred political parties. Therefore, this chapter aims to reveal the identity of Hungarian voters who prioritize the preservation of democracy over party loyalty, ideological alignment, and pragmatic policy considerations(see also Svolik et al., 2022). Our results show that the average Hungarian voter is willing to hold politicians accountable for violating democratic principles, although to a lesser extent than the voters in our comparative study. The average punishment rate in Hungary is-4.8%, the lowest of all the countries in our survey. To put this in context, Spain has the second lowest average punishment rate at-5.9%, while Sweden has the highest at-10.0%. It’s worth noting that Hungary’s punishment rate is almost half that of Poland(-9.2%), despite the fact that both countries have experienced significant levels of autocratization in recent years. Average punishment rates all countries 5 All figures in percent Sweden Poland Serbia Estonia Germany Ukraine Spain Hungary punishment -2% -4% -4.8 -5.9 -6% -7.1 -7.9 -7.8 -7.5 -8% -9.2 -10.0 5 Contrary to our previous studies, in the Hungarian candidate choice experiment, respondents always had to choose between a Fidesz and a non-Fidesz candidate. While this specificity helps taking into account the political dominance of Fidesz in Hungary, it creates a structural difference in our study designs. Comparisons of punishment rates between Hungary and countries 26 of our prior studies should therefore only be done with due caution.

How democratically elected politicians get away with autocratizing Hungary This suggests that while Hungarian voters are inclined to punish politicians for violating democratic norms, they tend to do so more moderately than in other European countries we have examined. These nuanced differences in voter behavior are crucial for a comprehensive understanding of democratic dynamics in different nations. Our findings also reveal correlations between punishment rates for undemocratic politicians and other factors, such as perceptions of the current state of Hungarian democracy or party affiliation. For example, respondents who are dissatisfied with the state of Hungarian democracy would punish undemocratic politicians at a rate of-7.2%, while those who are satisfied with the situation would impose a punishment rate of only-2.4%. HOW SUPPORTERS OF DIFFERENT PARTIES REACT TO UNDEMOCRATIC BEHAVIOR Looking at punishment rates in Hungary, there is also a clear contrast between supporters of different political parties. Supporters of the Fidesz-KDNP coalition have the lowest punishment rate at-2.7%, while supporters of the Momentum Movement have the highest punitive response at-9.7%. 6 How supporters of different parties react to undemocratic behavior 6 This is how many% of votes candidates would lose if they proposed an undemocratic measure. According to voting behavior, all figures in percent Momentum Movement (MM) Everybody’s Hungary People’s Party (MMNP) Democratic Coalition (DK) Hungarian Two-Tailed Dog Party (MKKP) DialogueGreens JobbikConservatives Other party/No answer Non-voters of the ´22 elections Our Homeland FideszKDNP punishment -2% -2.7 -3.4 -4% -4.3 -5.0 -4.8 -4.8% average punishment -6% -6.4 -7.4 -8% -8.2 -8.0 -9.7 This pattern of varying willingness to punish politicians for democratic violations is also evident in our comparative study. As in Hungary, across the range of countries we have examined, the variation in punishment rates is often greater than the variation observed between countries as a whole. It is important to note, however, that of all the partisan subgroups examined in our study, supporters of the Fidesz-KDNP clearly emerge as the most tolerant when it comes to democratic violations. They show the highest level of leniency among the political factions analyzed, regardless of the specific nation under consideration. To achieve a broader perspective, we will examine supporters of political parties in various countries who demonstrate a similar degree of leniency as supporters of Fidesz-KDNP. In Ser6 For the supporters of other political parties, i.e. Politics Can Be Different(LMP) and the Hungarian Socialist Party(MSZP), our results were not statistically significant, at-5.4% and-4.1% respectively. 27

IDENTITY, PARTISANSHIP, POLARIZATION bia, the party subgroup with the lowest punitive response to such violations is the SNS, with an electoral punishment rate of -4%. In Ukraine, voters of the(now banned) opposition Platform for Life shows a punishment rate of-5.4%. In Germany, non-voters show the lowest rate of electoral punishment at -4.2%, while voters of the AfD punish at-4.9%. Supporters of EKRE in Estonia are remarkably lenient, with a penalty rate of -4.3. In Spain, supporters of Vox and PP show a similar trend, with a penalty rate of-4.9%. These results show that both in Hungary and in six out of seven European countries surveyed, supporters of illiberal right-wing parties and non-voters are the least likely to punish undemocratic behavior. Which party supporters are least likely to punish democratic violations? Punishment by select European parties, according to voting behavior, all figures in percent Ukraine Opposition Platform for Life Germany AfD Spain VOX and PP Sweden Sweden Democrats Estonia EKRE Serbia SNS Poland PiS Hungary Fidesz-KDNP punishment -4.0 -4.3 -4.6 -4.9 -4.9 -5.4 Punishment rates among the non-voters Rate of punishment for democratic violations among non-voters, all figures in percent Serbia Ukraine Poland Spain Estonia Hungary -3.4 Germany -1% -2% -2.7 -3% -4% -5% Sweden punishment -2% -3.3 -4% -4.3 -4.2 -4.9 -4.8 -5.5 -6% -7.5 -8% -8.3 Comparing the punishment rates between Fidesz-KDNP supporters in Hungary and PiS supporters in Poland, especially considering the recent experience of democratic erosion in both countries, reveals comparable levels of punishment. FideszKDNP supporters have a punishment rate of-2.7%, while PiS supporters have a rate of-3.4%. Although these countries share a similar history of democratic violations, there are differences in the electoral consequences of such violations, especially 28

How democratically elected politicians get away with autocratizing Hungary for supporters of opposition parties. In particular, while punishment rates among Hungarian opposition supporters do not exceed-10%, supporters of parties that were previously in opposition in Poland during the data collection period showed a stronger tendency to punish democratic violations. For example, among supporters of Civic Platform(CP), the punishment rate is-13.5%, and among supporters of Lewica, it is -13.3%. AMONG THE PARTY LOYALISTS, WHO ARE THE CITIZENS THAT PUNISH DEMOCRATIC VIOLATIONS THE MOST? When analyzing the impact of educational attainment and party affiliation on the role of democratic principles in vote choice, our results reveal significant differences. In the 2022 elections, voters affiliated with the“Six Party Opposition” with a secondary education would im pose the strongest punishment(-11.4%) on candidates with undemocratic posi tions. Similarly, within the same political camp, voters with a university or college degree impose a penalty of-10.8%. Notably, supporters of the far-right“Our Homeland,” who have a university or college degree, impose a significant punishment of-9.5%. Respondents with a lower level of education generally show less punitive attitudes across most political groups. It is worth noting that among Fidesz voters, only those with a university degree significantly penalize undemocratic behavior(-8.2%). Party support and education Rate of punishment for democratic violations according to past vote 7 and education level, all figures in percent FideszKDNP United for Hungary Hungarian TwoOur Tailed Dog Homeland Party(MKKP) Other party/No answer Non-voters of the ´22 elections Education Across the supporters of all political groups, the respondents with lower levels of education generally punish democratic violations less than those with higher education. college degree University/ -8.2-10.8-9.5 -8.9-5.8 n/a* Secondary education with completed n/a* -11.4 n/a* n/a* n/a* -6.9 leaving exam Lower education n/a*-5.9 n/a* n/a* -4.4 -5.6 *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. The United for Hungary voters are the most punitive towards anti-democratic positions, especially the female opposition supporters(women:-10.3% men: -8.4%). Among Fidesz voters, we see a statistically significant negative reaction to undemocratic behavior only among women(-3.6%), but this is also much smaller than what we see among opposi tion voters or among those who did not vote in the 2022 elections. 7 When measuring party support, we usually use the“future vote” preference measure, which asks respondents which party they would vote for at the next parliamentary election in Hungary. For some observations, we analyzed their party support by looking at the last parliamentary election(past vote) as in the 2022 election, a grand coalition of all opposition parties ran against the incumbent Fidesz-KDNP. 29

IDENTITY, PARTISANSHIP, POLARIZATION Party support in the past elections, gender and punishment rates Rate of punishment for democratic violations according to past vote and gender, all figures in percent FideszKDNP United for Hungary Our Homeland Other party/No answer Non-voters of the ´22 elections Male n/a* -8.4-4.0 n/a* -6.5 Female -3.6-10.3 n/a* -4.9-5.2 *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. Looking at party preferences across age groups, we also found interesting patterns. In particular, respondents in the oldest age group who did not indicate their party preference showed the most significant punishment(-25.6%). Supporters of the“Six Party Opposition” in the youngest age group(18-29 years old) also showed substantial punishment(-12.8%). Similarly, both opposition voters and non-participants in the 2022 elections, who also belong in the 40-49 age group, displayed notable punishment(-10.3 and-9.9% respectively). The majority of age groups within the camp of the“Six Party Opposition” voters show significant and substantive punishment of undemocratic candidates. In contrast, the majority of age groups among Fidesz voters show insignificance in their punitive attitudes. Exceptions are Fidesz voters in the oldest age cohorts and in the 30-39 age group, who show statistically significant punishment(-7.6 and-3.4%). Party support and rural vs. urban Rate of punishment for democratic violations according to past vote and settlement type, all figures in percent FideszKDNP United for Hungary Our Homeland Hungarian TwoTailed Dog Party(MKKP) Other party/No answer Non-voters of the ´22 elections Budapest -3.8*-12.1+4.6*-4.9*-3.5*-7.9 County town -3.2*-10.7+3.7*-3.6* City -1.6*-9-1.6*+1.1* Village -2.6*-6.6-6.7-2.7* Budapest: County town: City: Village: 1,77 M inhabitants central city of each Hungarian county; 40,000- 200,000 inhabitants 10,000- 40,000 inhabitants below 10,000 inhabitants *The punishment or compensation variable was not statistically significant. -0.8* -3.1* -5.2 -5.8 -8.4 -2.3* 30

How democratically elected politicians get away with autocratizing Hungary Respondents aligned with the“Six Party Opposition” show the strongest intolerance of undemocratic politicians across various settlement types, with individuals in Budapest(-12.0% punishment), county towns(-10.7%), and cities(-9.0%) all expressing substantial reductions in vote share. In contrast, respondents who did not vote in the 2022 elections punish to a lesser, but still notable extent, especially in cities(-8.4%) and Budapest (-7.9%). Respondents who favor“Our Homeland” similar punitive attitudes in villages(-6.7%), while the other settlement types cannot be interpreted due to no statistical significance. Our analysis did not reveal any significant punishment reactions for the subgroups of Fidesz-KDNP supporters based on settlement type. We have also broken down the partisan camps according to the level of authoritarianism of the respondents. Voters of the six-party opposition, who showed no signs of authoritarianism, punished undemocratic candidates the most(-13.9 and -11.3%). Among the opposition voters, only the“completely authoritarian” subgroup did not punish undemocratic positions significantly, while among the Fidesz voters only the“rather not authoritarian” group’s punishment exceeded the threshold of statistical significance(-5%). These examples reflect well that authoritarian personality seems to reduce the frequency of democratically conscious voting choices within most partisan groups. Party support and authoritarian traits Rate of punishment for democratic violations according to past vote and authoritarian traits, all figures in percent FideszKDNP United for Hungary Other party/No answer Non-voters of the ´22 elections Not authoritarian -5.4*-14.0-5.4-3.7* Rather not authoritarian -5.8 -11.3 -6.2 -2.9* Somewhat authoritarian -3.1* -6.0-5.7-7.4 Rather authoritarian -1.6*-8.0-1.5*-7.6 Completely authoritarian +1.8*-3.0*-3.6 *The punishment or compensation variable was not statistically significant. -3.4* In the 2022 parliamentary elections, the subgroups of respondents who voted for the six-party opposition alliance were found to be the most democratic, based on the extent to which the undemocratic characteristics of the candidates influenced their vote. Among them, women, those with secondary or higher education, those living in urban areas(especially in Budapest) and those without an authoritarian personality stood out, as these groups punished undemocratic behavior the most. 31

IDENTITY, PARTISANSHIP, POLARIZATION EXPLORING THE ROLE OF MEDIA CONSUMPTION IN ADDRESSING UNDEMOCRATIC BEHAVIOR With regard to the condemnation of specific undemocratic positions, our goal was also to explore potential correlations between the tendency to punish political candidates who sup port undemocratic policies and individuals’ media consumption habits. For this purpose, we categorised media consumers in Hungary into four profiles: predominantly government-critical media consumers, non-media consumers, mixed media consumers, and primarily pro-government media consumers. Our results demonstrate that individuals who predominantly consume government-critical media show almost double the punishment rates(-5.4%) compared to those who predominantly consume pro-government media(-2.8%). However, the highest punishment rate is observed among non-media consumers, at-6.9%. Media consumption and electoral punishment rates According to media consumption, all figures in percent Mainly pro-government media consumer Mixed-media consumer Mainly government-critical media consumer No media consumer/ No answer punishment -2% -2.8 -3.5 -4% -5.4 -6% -6.9 EXPLORING THE ROLE OF RURAL VS. URBAN POPULATION IN ADDRESSING UNDEMOCRATIC CONDUCT Our results also reveal notable differences in punishment rates across settlement types. Specifically, residents of Budapest show a punitive attitude towards undemocratic behavior with a rate of-6.9%, while residents of villages show a comparatively lower punishment rate of-3.6%. Residents of other cities and county towns fall somewhere in between, with rates of-4.5% and-5.2% respectively. Main Takeaways: The section above highlights Hungarian voters’ substantial declarative support for democracy but explores whether this translates into holding politicians accountable for democratic violations. Despite pro-democracy sentiments, voters tend to penalize such transgressions less severely compared to the other European countries we studied. The analysis further reveals correlations between punishment rates and factors such as perceptions of democracy, party affiliation, educational attainment, and settlement type. Party affiliation emerges as a key determinant of punishment rates for undemocratic behavior. Supporters of the ruling Fidesz-KDNP coalition exhibit the lowest punishment rates, indicating a higher tolerance for democratic violations among this group. Furthermore, educational attainment plays a significant role in shaping individuals’ attitudes towards democratic accountability. The analysis reveals that voters with higher levels of education are more likely to impose a penalty on politicians for undemocratic behavior compared to those with lower levels of education. Media consumption habits also influence punishment rates, with government-critical media consumers exhibiting higher rates of electoral punishment. Lastly, settlement type plays a role, with residents of Budapest showing more punitive attitudes towards undemocratic behavior compared to those in villages. 32

How democratically elected politicians get away with autocratizing Hungary WHICH UNDEMOCRATIC POSITIONS ARE VOTERS MOST WILLING AND LEAST WILLING TO PUNISH? Next we sought to determine whether respondent exhibit varying levels of concern towards specific violations of democracy. In our experiments, selected political candidates supported policies that contradicted three fundamental tenets of liberal democracy: electoral fairness, civil liberties, and checks and balances. Our analysis suggests that although Hungary has experienced democratic setbacks across all three pillars, respondents prioritize the protection of certain democratic facets over others. In particular, the are the most concerned about violations of electoral fairness, with an average punishment rate of-8.1%. Violations related to checks and balances have a punishment rate of-5.8%. Conversely, violations related to civil liberties are the least worrying for Hungarian respondents, with a punishment rate of-4.4%. 8 Average punishment by key elements of liberal democracy All figures in percent Electoral Fairness Checks and balances Civil liberties -2% Electoral Fairness Voters value the electoral component of a liberal democracy more than the civil liberties and checks and balances. punishment -4% -4.3 -6% -5.8 -8% -8.2 Looking at specific instances of democratic violations, Hungarian respondents were most likely to penalize undemocratic behavior in scenarios where a candidate advocated favoring districts that voted for his party in the allocation of infrastructure spending. In a 50-50% contest between two political candidates, the candidate who advocated this violation would see his vote share reduced by- 8.5%. The second most punished democratic violation is actions against journalists. In the experiment, a politician who supports measures to discipline judges who criticise the government would see his vote share fall by- 8.2%. The results are to be expected, as both budgetary patronage and media freedom have become contentious issues in Hungary over the past decade. During this period, Hungary has witnessed a significant deterioration in media freedom, accompanied by cases of state capture and violations of the rule of law. In addition, public funds have been allocated to municipalities governed by the ruling parties 8 These numbers include punishment rates for all undemocratic positions for which the results were statistically relevant. 33

IDENTITY, PARTISANSHIP, POLARIZATION (Ghergina and Volintiru, 2023). However, in contrast to Hungarian respondents’ awareness of these two democratic violations, they show the least concern about the democratic violation related to the funding of trade unions. If a politician were to support a ban on foreign funding of trade unions critical of his or her party’s policies, his or her vote share would decrease by only-2.2%. Average punishment for candidates by undemocratic positions All figures in percent All countries average Estonia Germany Hungary Poland Violently disrupt rallies -12.2-11.5-12.4-7.8-11.2 Serbia -11.8 Spain -11.9 Prosecute journalists -10.0-13.0-8.4-8.2-18.2-12.6-3.0 Ban prayers for Muslims -8.6-7.4-11.1 n/a-7.4-6.3-5.7 Discipline judges -8.2-8.4-7.1-6.1-14.1-8.1-5.7 Partisan infrastructure spending -7.3-6.8-6.3-8.5-7.6-5.1-7.5 Monitor social media -6.0-5.5-6.5-2.6-7.6-5.1-2.9 Ban foreign NGO funding Pass laws without parliamentary debate Ban foreign union funding -5.6-8.6-6.1(-1.9*)-4.1-6.3-5.5 -5.4-5.4-3.3-5.5-14.5-9.7-2.5 -5.0-6.4-4.4-2.2-4.6-4.8-6.1 *For the undemocratic position“Supported a ban on foreign funding for domestic non-governmental organizations critical of his party’s policies” our results were statistically insignificant with a punishment rate of-1.9%. Sweden -17.3 Ukraine -10.6 -14.5-5.8 -10.0-4.2 -13.7-2.8 -10.4-8.3 -8.1-6.6 -5.1-4.2 -4.9-4.0 -6.3-3.8 Our findings also reveal discrepancies between supporters of different political parties in their condemnation of undemocratic attitudes. 9 A detailed examination of the most punished democratic violation,“If we win, we should direct infrastructure spending to districts that voted for our party”, highlights these differences quite clearly. The supporters of“Our Homeland” would impose the heaviest penalty for this democratic violation, with-11.1%. In contrast, supporters of Fidesz-KDNP and non-voters would impose the lowest punishment on politi cians for this democratic violation, with punishment rates of-8.7% and-6.8% respectively. 10 Looking at the rates of punishment for this violation of democracy, there are some gender differences. Specifically, male respondents show a punishment rate of-10.0%, while female respondents show a slightly lower rate of-7.2%. There are also differences between different age groups. The 76+ age group shows the highest punishment rate at-12.6%, while the 66-75 age group shows the lowest rate at-6.9%. 9 For the undemocratic position“Supported a ban on foreign funding for domestic non-governmental organizations critical of his party’s policies,” our results were statistically insignificant with a punishment rate of-1.9%. 10 For the following parties, the results were statistically insignificant: Politics Can Be Different(-11.0%), Everybody’s Hungary People’s Party(MMNP)(-6.3), Hungarian Socialist Party(MSZP)(-6.0), Jobbik-Conservatives(-3.7), and Dialogue-Greens(+4-0). 34

How democratically elected politicians get away with autocratizing Hungary Party support in Hungary and punishment of democratic violations Rate of punishment for the undemocratic position“ If we win, we should direct infrastructure spending to districts that voted for our party” and party support. All figures in percent Our Homeland Momentum Movement (MM) Hungarian Two-Tailed Dog Party (MKKP) Democratic Coalition (DK) FideszKDNP Other party/No answer Non-voters of the ´22 elections punishment -9.6 -11.1 -10.8 -10.7 *The punishment variable was not statistically significant. -6.8 -9.3 -8.7* -2% -4% -6% -8% -10% A closer look at the punishment rates for the second most punished undemocratic position, “Said the government should prosecute journalists who accuse the government of corruption,” shows even stronger differences between the supporters of different parties. The supporters of the Jobbik-Conservatives and the supporters of the Hungarian Socialist Party(MSZP) punish this democratic violation with-21.5%, whereas supporters of the Democratic Coalition(DK) punish undemocratic politicians for this violation with-12.7%. For some other party supporters, including Fidesz-KDNP supporters, this violation is not very important, so we cannot observe any statistically significant results. Furthermore, when it comes to the rates of punishment for this violation, we do not observe any differences in the rates of punishment between male and female respondents. While the male respondents punish this infringement with-8.4%, the female respondents punish it with-8.0%. Party support in hungary and punishment of democratic violations Rate of punishment for the undemocratic position“ Said the government should prosecute journalists who accuse the government of corruption” and party support. All figures in percent Jobbik- Hungarian Socialist Our Conservatives Party(MSZP) Homeland Momentum Movement (MM) Democratic Coalition (DK) FideszKDNP Non-voters of the ´22 elections Punishment of specific democratic violations Hungarian respondents were most likely to penalize undemocratic behavior in scenarios where a candidate advocated favoring districts that voted for his party in the allocation of infrastructure spending. -1.6* punishment -19.0 -21.5 -21.5 *The punishment variable was not statistically significant. -13.7 -13.5 -12.7 -5% -10% -15% -20% 35

IDENTITY, PARTISANSHIP, POLARIZATION PUNISHMENT OF INDIVIDUAL UNDEMOCRATIC POLICY POSITIONS AND MEDIA CONSUMPTION Media consumption and punishment of democratic violations Rate of punishment for the undemocratic position“ Supported a ban on foreign funding for domestic non-governmental organizations critical of his party’s policies” and media consumption. All figures in percent +8.4 +8% +6% reward Mainly progovernment media consumer Mainly government-critical media consumer -4.8 +4% No media consumer/ No answer +2% -2% -4% -6% punishment -7.4 -8% Our findings reveal significant differences between media consumer groups in the punishment of political candidates for violating democratic principles. A detailed examination of punishment rates for the democratic violation of“Supporting a ban on foreign funding for domestic non-governmental organizations critical of their party’s policies” effectively highlights these differences. Hungarian citizens who primarily follow media critical of the ruling coalition show a punishment rate of-4.8% for this violation of democracy. Conversely, those who identify themselves as“non-media consumers” impose a punishment rate of -7.4%. However, individuals who primarily consume pro-government media show no inclination to punish politicians for this transgression. Instead, a politician proposing a ban on foreign NGO funding would receive a reward of+8.4%. In addition, our analysis reveals variations in punishment rates among respondents who predominantly consume media critical of the government. The observed differences between media consumption bubbles are not surprising in the Hungarian context. Since coming to power in 2010, Orbán and his government have systematically undermined media plurality and independence, contributing significantly to these different patterns of media consumption and the resulting impact on public perception. In addition, issues such as foreign funding of NGOs have been the subject of contentious debate. In 2017, the Hungarian government implemented a law aimed at imposing restrictions on NGOs that receive foreign funding. The law required these organizations to register as“foreign-funded” if they received more than€20,000 in foreign funding per year. Although the Court of Justice of the European Union ruled 11 that the law violated EU regulations and subsequently forced the Hungarian government to repeal it, the autonomous work of civil society organizations remains under threat. Orbán’s government recently passed a new law, known as the Sovereignty Protection Act, ostensibly aimed at removing foreign financial influence from the country’s political landscape. However, both civil society organizations 12 and opposition parties agree that this law is primarily intended to stifle critical voices in the run-up to the European elections in 2024 and beyond. 11 Court of Justice of the European Union. Judgement in Case C-78/18 Commission vs. Hungary 12 Hungarian Helsinki Committee.“The Proposed Regime Defence Law is Bound to Fail.” November 22, 2023. 36

How democratically elected politicians get away with autocratizing Hungary For some of the other democratic violations, there are no significant differences in punishment rates between consumers of pro-government media and those who are primarily engaged with media critical of the government. For instance, consumers of pro-government media would punish the democratic violation“Stated that the government should prosecute journalists who accuse the government of corruption” with a punishment rate of-11.5%, while respondents who primarily consume media critical of the government would punish this violation with a rate of -11.3%. 13 In the subsequent analysis, we delve deeper into the demographics of voters most inclined to penalize undemocratic behaviors. We scrutinize a range of variables such as democratic preferences, competence, age, gender, education, net income, political knowledge, and religion to discern the traits associated with a heightened propensity for holding violators accountable. By investigating these factors, our aim is to gain a nuanced understanding of the specific demographics and characteristics that foster a stronger inclination to sanction democratic violations. Media and Punishment Consumers of of government-critical media demonstrate higher punishment rates towards certain democratic violations. Main Takeaways: The section above examines Hungarian respondents’ varying levels of concern towards specific violations of democracy, focusing on electoral fairness, civil liberties, and checks and balances. While respondents prioritize protecting electoral fairness the most, violations related to civil liberties are of least concern. Specific instances of undemocratic behavior, such as favoring party-voting districts in infrastructure spending allocation or actions against journalists, receive significant punishment from voters. However, violations related to trade union funding evoke the least concern. There are notable differences in punishment rates based on party affiliation, gender, and age group. For instance, older age groups, particularly those aged 76 and above, tend to exhibit higher punishment rates for certain violations compared to younger age groups. Media consumption patterns further shape punishment rates, with consumers of government-critical media demonstrating higher punitive attitudes towards certain violations. 13 For the“mixed-media” consumer, the results(-0.75) were statistically insignificant. 37

IDENTITY, PARTISANSHIP, POLARIZATION DEMOCRATIC ATTITUDES AND PUNISHMENT RATES We began our investigation by testing whether respondents’ self-reported views of democracy predicted their voting behavior when it came to punishing candidates for being undemocratic. Our results show that with regard to the general evaluation of democracies, only those who think it is good to live in a democratic system punish a candidate for being undemocratic. Among those who think it is“very good” to have a democratic system, being undemocratic leads to a loss of vote share of-8.2%. Among those who think it is“quite good” to have a democratic system, the penalty for being undemocratic is much lower, only- 4.5%. Attitudes towards experts in government correlate with the degree to which voters punish candidates with undemocratic positions. Those who have a more positive view of experts in the government tend to punish undemocratic candidates more severely. Voters who have a“fairly good” or“very good” opinion of experts in government show a significant reduction in the vote share for candidates with undemocratic positions, with a loss of- 6.1 and- 5.2% respectively. Individuals with a“fairly bad” perception of experts in government display a reduced level of punishment towards candidates with undemocratic positions. Their loss of vote share is –3%. Those who perceive expert government as“very bad” do not punish undemocratic behavior. These patterns show that support for“expert governance” in an autocratic context may not necessarily reflect dissatisfaction with democracy, but rather a critical attitude towards the incumbent regime. To some extent, the introduction of expert governance may be understood by many as the restoration of the independence of certain branches of government, such as independent courts, public prosecutors, competition authorities and other regulatory bodies. Attitudes towards an army ruling the country have a significant impact on voters’ reactions to undemocratic candidates. Those with negative perceptions –“very bad” or“fairly bad” opinion of army rule – show significant reductions in the vote share for undemocratic candidates, with losses of around-6.5 and-3.6% respectively. Conversely, respondents with a“very good” or“fairly good” opinion of military rule do not significantly penalize candidates with undemocratic positions. This shows that, in contrast to attitudes towards expert governance, opposition to military rule is associated with a democratically conscious vote. Perceptions regarding the quality of democracy in Hungary also significantly impact voting behavior towards candidates Punishment of democratic violations Rate of punishment according to the response to the question “ How democratically is your country being governed today?”. All figures in percent not democratic 12 rather nondemocratic 345 rather democratic 67 democratic 8 9 10 no answer 1-3 4-5 6-7 8-10 -1% -2% -3% -2,5 -2,7 -4% -3,7 -5% -4,8 -6% -6,5 Punishment of democratic violations Rate of punishment according to the response to the question “ How satisfied are you with the state of democracy in Hungary?”. All figures in percent not rather not satisfied satisfied 12345 rather satisfied satisfied 6 7 8 9 10 no answer 1-3 4-5 6-7 8-10 0,0 -1% -2% -3% -2,9 -2,4 -4% -3,9 -5% -6% -7% -7,2 38

How democratically elected politicians get away with autocratizing Hungary with undemocratic positions. Respondents who see Hungary as “not democratic” or“rather undemocratic” are significantly less likely to vote for undemocratic candidates, with losses of around-6.4 and-4.3%, respectively. These punishment rates are smaller,-3.7and-2.5%, for those who view Hungary as “rather democratic” or“democratic”. Similarly, the level of satisfaction with the state of democracy in Hungary significantly influences voters’ reaction to candidates with undemocratic positions. Respondents who express strong dissatisfaction with the quality of democracy under Fidesz show the highest reduction in the vote share for undemocratic candidates, with a loss of-7.1%. In the“rather dissatisfied” group, the punishment is much smaller(-3.9%. Those who express satisfaction with the state of democracy (“rather satisfied” or“satisfied”) also exhibit a significant but smaller reduction in the vote share for such candidates, of around-2.9% and-2.4% respectively. Respondents who say it is very important for them to live in a democracy show a significant reduction in their vote share for undemocratic candidates, with a loss of around 5.98%. Similarly, those who consider it“not very important” also show a significant drop in their vote share(- 5.1%). However, for those who consider living in a democracy“not at all important” or “rather important”, there is minimal to no statistically significant punishment for candidates with undemocratic positions. Democratic competence All figures in percent Democratically competent Democratically incompetent punishment -1% -2% -3% -3,8 -4% -5% -5,1 We also examined the extent to which knowledge of democracy influences the punishment of undemocratic candidates. Democratically competent respondents are significantly less likely to vote for undemocratic candidates(-5.1%). On the other hand, democratically incompetent respondents, who find it difficult to distinguish between objectively democratic and non-democratic events, also punish undemocratic practices, but to a lesser extent(-3.8%). Main Takeaways: This study section examined whether respondents’ views on democracy predicted their voting behavior regarding punishment for undemocratic candidates. Results indicate that those who view democracy positively are more likely to penalize undemocratic candidates. Attitudes towards expert governance and military rule also influence punishment rates. Those with positive views of expert governance tend to penalize undemocratic candidates more, while negative perceptions of army rule correlate with significant reductions in vote share for undemocratic candidates. Perceptions of Hungary’s democracy quality and satisfaction with its state further impact voting behavior. Dissatisfaction with the quality of democracy under Fidesz leads to the highest reduction in vote share for undemocratic candidates. Furthermore, the importance individuals place on living in a democracy also affects punishment rates. Those considering democracy very important exhibit a significant vote share reduction for undemocratic candidates, as do those who consider it not very important. However, minimal to no punishment is observed for those who view living in a democracy as not at all important. Lastly, knowledge of democracy influences punishment rates, with democratically competent respondents showing a higher likelihood of penalizing undemocratic candidates compared to those less knowledgeable in democratic principles. 39

IDENTITY, PARTISANSHIP, POLARIZATION SOCIO-ECONOMIC, DEMOGRAPHIC AND PSYCHOLOGICAL PROFILE OF DEMOCRATICALLY MINDED VOTERS Age Underlining the age-based trend, older demographic groups are more punitive towards undemocratic candidates than younger age groups, showing a consistent increase in punitive attitudes with increasing age. Specifically, the oldest age cohort(76 and over) shows a significant decrease in vote share(-7.8%), while the youngest age group(18-29) shows a notable but comparatively smaller decrease in vote share(-4.2%). One possible explanation for this trend is that Hungarian youth is politically very passive, disinterested and disillusioned by Central and Eastern European standards, and protests mobilize mainly members of older generations(with the exception of climate activism or university demonstrations). Regarding the violation of specific democratic norms, the only undemocratic act punished by the youth(exceeding the threshold of statistical significance) is the prosecution of journalists who accuse the government of corruption. Age groups and punishment Values for select European countries, all figures in percent 18-29 30-39 years old 40-49 50-65 66-75 76 or older punishment -2% -3,5 -4,2 -4% HUN -5,1 -5,3 -5,8 GER -6% ESP -8% -7,8 EST POL -10% -12% SWE Gender Similar to our findings in other countries studied earlier, women in Hungary, on average, demonstrate a greater propensity to penalize undemocratic behavior compared to men(-5.2% and -4.3% respectively). However, significant disparities between the punishment rates of women and men are evident within certain age groups. Notably, women aged 40-49 exhibit the highest degree of punishment towards undemocratic candidates(-7.3%), whereas men in the same age bracket penalize democratic violations to a lesser extent(-2.8%). Conversely, within the age group 50-65, we observe an inverse trend. While men within this age range impose a significant penalty for democratic violations (-7.0%), making them the second-highest group in penalizing such transgressions, women within the same cohort exhibit a lower degree of punishment towards undemocratic politicians (-4.0%). The observed differences may be attributed to historical context; women aged 40-49 were teenagers during Hungary’s democratic transition and the collapse of the Soviet Union, while men aged 50-65 were young adults. Religion Among the statistically significant findings, those identifying as Protestant show a notable punitive response, with a reduction in vote share of-9.5%, followed by atheists(-7.9%). In addition, those with an affiliation categorised as“something else”(-5%) and Roman Catholics(-4.7%) show significant but comparatively lower punitive tendencies. Rural vs. urban Our research also examined the correlations between the intensity of punishment for undemocratic behavior and the geographical residence of individuals. For certain types of undemocratic behavior, our results show significant differences in 40

How democratically elected politicians get away with autocratizing Hungary punishment rates, for instance between residents of Budapest Education levels and punishment and those living in rural areas. For example, if a political candiAll figures in percent date were to“encourage his supporters to violently disrupt campaign rallies of his political opponents,” inhabitants of Budapest and other larger cities would punish this democratic University/college degree Secondary education with completed leaving exam Lower education violation with-11.8 and-10.3%, respectively. In contrast, resipunishment dents of villages would only impose a punishment of-5.3% for -2% this democratic breach. In addition, if a politician were to advocate infrastructure funding for districts that support his or her party, Budapest residents would penalize this violation with -3.1 -4% -13.3%. In contrast, residents of villages would impose a penalty of-10.5%, while those in county towns and other towns would impose penalties of-7.1 and-5.5% respectively. -5.4 -6% Education The data show a clear educational pattern, with higher levels of education corresponding to more pronounced punitive responses. Individuals with a university/college degree show the most punitive attitudes, with a substantial reduction in their vote share(-8.5%), followed by those with secondary education and completed leaving exams(-5.4%). In contrast, individuals with a lower level of education show a comparatively lower, but still significant, punitive tendency(-3.1%). There are certain democratic violations that only intellectuals are concerned about in Hungary. If a politician supported passing laws without debate, disciplining judges critical of the government, monitoring social media posts critical of the government and banning foreign funding of trade unions, this led to a significant loss of votes among the group of respondents with a tertiary education, but not among those with a secondary education or less. Net income Higher net income groups show a tendency toward more puni tive attitudes. Households with an net income between HUF 650,001 and HUF 1,000,000 show a significant decrease in the -8.5 Net income levels and punishment All figures in percent Household net income HUF less than HUF 250,001 250,001 450,000 HUF 450,001 650,000 punishment -3.8 -3.8 -6.4 -8% HUF 650,001 1,000,000 -2% -4% -6% -8% -8.6 41

IDENTITY, PARTISANSHIP, POLARIZATION Punishment of democratic violations and future personal status Rate of punishment according to the response to the question “Now looking ahead— do you think that a year from now you (and your family) will be much worse off, worse off, just about the same, better off, or much better off?” All figures in percent Better Stay about the same Worse Much worse -1% punishment -2% -3% -4.3 -5.0 -3.5 -4% -5% -5.8 As we did not get statistically signifi cant results for the category“much better”, we are notreporting any results for it. vote share(-8.7%), followed by those in the net income range between HUF 450,001 and HUF 650,000(-6.4 percent). In contrast, lower net income brackets brackets show a comparatively smaller, but still noteworthy punitive tendency, with households earning less than HUF 250,001 and those within the range of HUF 250,001 to HUF 450,000 displaying a reduced vote share (-3.8 and-3.8% respectively). Households with an net income of more than HUF 1,000,000 do not exhibit a statistically significant punitive response – most likely due to the limited number of respondents from this net income group in our sample. Future economic prospects Punitive reactions to candidates with undemocratic positions reveal interesting patterns in the projection of future economic prospects, both personal and national. The most punitive attitudes are among respondents who expect their personal economic situation to be worse in a year’s time(-5.8%) and those who expect the economic situation in Hungary to be worse (-6.2%). However, the most optimistic respondents(both with regard to the country’s economy and their personal financial situation) do not punish undemocratic candidates significantly. This is probably related to the fact that they are more likely to vote for the government. For comparison, a look at the data on the same question in Poland shows that those respondents who think that the economic situation will get worse punish twice as much than those in Hungary, with a rate of – 10.9%. Punishment of democratic violations and the future economic situation of Hungary Rate of punishment according to the response to the question “What about the next year? Do you expect the economic situation in Hungary to get much worse, worse, stay about the same, get better, or get much better?” All figures in percent Better Stay about the same Worse Much worse Satisfaction with democracy and punishment Rate of punishment for democratic violations according to satisfaction with democracy(select countries). All figures in percent Hungary Germany Poland not statisfied -7.2 (1-3) -5.9-12.4 punishment -1% -2% (4-5) rather not statisfied -3.9 -3% -3.7 -4% -5% -4.9 -5.3 -6% -6.2 As we did not get statistically signifi cant results for the category“much better”, we are notreporting any results for it. rather statisfied -2.9 (6-7) statisfied -2.4 (8-10) -6.07-11.1 -7.7-5.0 -8.9-4.2 42

How democratically elected politicians get away with autocratizing Hungary Trust and authoritarian personality Examining the psychological profile of voters reveals interesting patterns in their punitive responses to candidates with undemocratic positions. In the case of trust, higher levels of trust correspond to more pronounced punitive reactions. Those who believe that“Most people can be trusted” show a significant reduction in their vote share(-6.0%), while those who are cautious in their interpersonal dealings display a smaller but still significant punitive tendency(-4.4%). Similarly, authoritarian personality traits show a distinct pattern, with lower authoritarian tendencies coinciding with more significant punitive attitudes. Respondents who identified as“Not authoritarian” demonstrate the largest decrease in their vote share(-8.5%), followed by those categorised as“Rather not authoritarian”(-6.9%). Conversely, higher authoritarian tendencies lead to comparatively lower punitive responses, suggesting a negative correlation between authoritarian tendencies and punitive attitudes towards candidates associated with undemocratic positions. Characteristics of respondents who showed more willingness to punish undemocratic politicians Women (especially those who are democratically competent) Higher educated ODEMACY CR Higher income Older Pessimistic economic prospects Main Takeaways: Our analysis of the voting behavior of Hungarians with different demographic, socio-economic and psychological characteristics revealed the main patterns of democratic consciousness. The groups most likely to punish politicians for undemocratic positions are women(especially those who are democratically competent), along with older respondents and those with higher education. In the Hungarian context, the seemingly contradictory result that respondents with higher net incomes and those with pessimistic economic prospects are less tolerant of undemocratic behavior holds simultaneously. This may be explained by the fact that supporters of the Orbán government are over-represented in the poorest social groups, while opposition voters are proportionally more numerous in the highly educated and wealthier social groups. Higher levels of interpersonal trust and lower levels of authoritarian traits can also be associated with higher punishment rates. 43

IDENTITY, PARTISANSHIP, POLARIZATION TOLERATING DEMOCRATIC VIOLATIONS Party Loyalty Most voters forgive politicians who violate democratic norms if they represent the voters’ preferred party. This subsection examines the impact of party loyalty on voters’ decisions to tolerate democratic violations and continue to support their preferred political candidates. By examining the role of partisanship in shaping voters’ responses, we can gain insight into the extent to which party affiliation may influence voters’ willingness to prioritize other considerations over democratic principles. THE INFLUENCE OF PARTISANSHIP THE ROLE OF PARTISANSHIP In line with the results of previous cross-national FES surveys, the Hungarian data also show that partisanship overrides democratic principles. The average increase in final vote share for a candidate affiliated with the respondent’s preferred party is 36 percent. This is more than seven times the average penalty for candidates who support undemocratic behavior. Comparing Hungary’s results with those of seven other European countries, it is clear that Hungarian respondents are not only the least democratically aware, but also the most loyal to their political groups. The general effect of co-partisanship Select European countries. All figures in percent 31.2 Final reward +30% compensation 36.0 +25% 21.7 19.9 18.0 +20% 15.9 15.6 15.4 +15% 11.5 +10% 27.6 22.9 23.5 23.7 25.1 29.1 +5% 21.5 0% punishment -4.8 Hungary -5.9 Spain -7.1 -7.8 -7.9 -7.5 -9.2 Poland Ukraine Estonia Serbia Germany Average punishment Compensation by voters’ party loyalty -5% -10.0 Sweden -10% In Hungary, it is not only party affiliation that matters, but also belonging to the same broad political camp. In the 2022 general election, the main opposition parties ran together against the ruling Fidesz-KDNP coalition. Although the alliance broke up after Fidesz’s historic landslide victory, the“opposition voter” seems to remain a relevant social and political group, as Hungarian society is politically polarized alongside the Fidesz vs. non-Fidesz divide. Therefore, we also 44

How democratically elected politicians get away with autocratizing Hungary tested the extent to which belonging to the same“broad political camp” as a particular candidate influences respondents’ vote choices. For these analyses, we limited our sample to those respondents who voted for the opposition alliance or Fidesz in the previous elections, and to scenarios where candidates belonged to either Fidesz or one of the mainstream opposition parties. Being in the same camp as a particular candidate resulted in an average gain of 28.8%. This is somewhat lower than the co-partisanship effect, but shows that the“broad camp identity” of a candidate also plays an influential role in Hungary, and that this factor also overrides other democratic considerations. Party support and compensation for co-partisan candidates All figures in percent 39.2 Final reward 36.0 31.3 29.4 28.6 21.6 44.0 compensation 34.0 44.0 25.0 36.0 39.0 -4.8 -2.7 -3.4 -8.0 -9.6 -7.4 +30% +20% +10% 0% -10% punishment JobbikConservatives Democratic Coalition (DK) FideszKDNP Average punishment Momentum Movement (MM) Hungarian Two-Tailed Dog Party (MKKP) Compensation by voters’ party loyalty As we did not get statistically significant results for the punishment of the Hungarian Socialist Party and"Politics can be different", we cannot calculate final rewards for these parties. Our Homeland Among the different partisan groups, the highthe government and the mainstream opposition est reward 14 for co-partisanship is among voters parties. It is noteworthy that Fidesz voters showof the Hungarian Socialist Party(+50% gain), case a below-average co-partisan reward, and the lowest among voters of Our Homeland (+34%), while it is above-average among supparty(25%). A possible explanation for the latporters of the mainstream opposition parties. ter is that Our Homeland’s base is a diverse group that includes not only far-right nationalists but also anti-vaccination activists, as well as voters who have become disillusioned with both 14 To enhance clarity and improve reader comprehension, in the following sections, we use the word“reward” to describe the effect that mitigates the initial loss of vote share of undemocratic candidates based on partisan or policy interests. Favorite party Hungarians are more loyal to their favorite parties than voters in the other countries we analyzed. 45

IDENTITY, PARTISANSHIP, POLARIZATION DETERMINANTS OF PARTY LOYALTY We also investigated the factors that influence the loyalty of partisan groups in Hungary. Party loyalty for certain groups varies across different settlement types, reflecting the importance of local contexts. In particular, Our Homeland voters exhibit an outstanding compensation for co-partisan candidates in Budapest(41%), while this effect is not as large in smaller towns and villages(12-27%). Both Jobbik and Momentum voters show the strongest co-partisan compensation in small towns(56 and 48%). In the case of the Democratic Coalition(DK), party loyalty is strongest in villages and small towns(49 and 45%). The Hungarian Two-Tailed Dog Party (MKKP) also records a significant co-partisan compensation in villages(49%). Among Fidesz voters, party loyalty is strongest in county seats and small towns(35 and 38%), but co-partisan compensation varies slightly in this group. Party support, rural vs. urban and compensation for voters’ party loyalty Rate of punishment and compensation according to partisanship(based on future vote) and rural vs. urban. All figures in percent Hungarian Two- Democratic FideszTailed Dog Party Jobbik- Politics Can be Coalition KDNP (MKKP) Conservatives Different(LMP) (DK) +39.5 pun. compensation +45.3 pun. compensation +39.7 pun. compensation How to read this graph: 1 The regular punishment of undemocratic behavior among voters of the MKKP party equals -9.2%. 2 In this example,+35.2 percentage points is the absolute compensation rate that candidates receive if they belong to the voter's preferred party. 3 Relative compensation(or sometimes punishment) is the evolution of vote shares or the electoral reward/punishment that candidates ultimately receive when both the original punishment and compensation are taken into account. 46 Budapest City +35.2 n/a* 3 26.0 n/a* 2 1 -9.2 49.2 n/a* n/a* -6,9 County 30.4 n/a* town 18.3 +35.0 +27.5 -4.6 -9.2 +56.1 n/a* n/a* 85.3 +9.9 +75.4 31.8 -7.7 34.6 -10.7 32.4 -7.3 42.9 Village 28.1 n/a* n/a* n/a* +31.4 -3.3 -5.9 *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. +48.8 pun. compensation

How democratically elected politicians get away with autocratizing Hungary The influence of party loyalty within different levels of education also shows some interesting patterns. A low level of education correlates with weaker party loyalty, but only within certain partisan groups(Our Homeland, Two-Tailed Dog Party, MSZP, Jobbik and LMP). It should be highlighted that pro-government voters in all education groups reward Fidesz’s candidates to a similar extent. Party support, education level and compensation for voters’ party loyalty Rate of punishment and compensation according to partisanship(based on future vote) and education level. All figures in percent Hungarian Two- Democratic Momentum Fidesz- Our Tailed Dog Party JobbikCoalition Movement KDNP Homeland (MKKP) Conservatives (DK)(MM) +34.4 punishment compensation 26.2 University/ college degree -7.9 16.4 -13.1 Secondary education with completed leaving exam 32.2 0.0 26.6 0.0 +32.2 +26.6 +59.0 +34.1 +29.5 +39.7 31.0 n/a* -8.7 51.9 n/a* -7.1 35.3 -8.7 33.1 -9.7 +42.8 +44.0 20.4 -14.0 n/a* punishment compensation punishment compensation 35.4 22.6 Lower education n/a* n/a* 39.5 n/a* +45.9 +35.4 +22.6 0.0 0.0 -6.4 *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. 47

IDENTITY, PARTISANSHIP, POLARIZATION Finally, we analyzed the effect of media consumption on the role of partisanship in respondents’ vote choice. The results show a consistent pattern. Exposure to counter-attitudinal media reduced party loyalty to a large extent, but only within the groups of voters of mainstream opposition parties. When a Jobbik, LMP, Democratic Coalition, or Momentum supporter consumed news from predominantly pro-government outlets, they did not reward those candidates who belonged to their preferred party. Among Fidesz respondents, we did not observe a similar effect of exposure to government-critical media consumption on party loyalty. Pro-government voters rewarded Fidesz candidates to a similar extent regardless of the political tone of their media environment. The only visible pattern was that Fidesz voters who did not consume mainstream media had the weakest party loyalty. In contrast, for the non-mainstream opposition parties(Two-Tailed Dog Party and Our Homeland), party loyalty was highest among those who do not regularly consume mainstream media. Party support, media consumption and compensation for voters’ party loyalty Rate of punishment and compensation according to partisanship(based on future vote) and media consumption. All figures in percent FideszKDNP Our Homeland Hungarian TwoTailed Dog Party (MKKP) DialogueGreens JobbikConservatives Hungarian Socialist Party(MSZP) Politics Can be Different(LMP) Democratic Coalition (DK) Momentum Movement (MM) punish. compensation +31.3 punish. compensation punish. compensation Mainly progovernment media n/a* n/a* n/a* n/a* n/a* n/a* consumer n/a* n/a* -0.7 -23.9 -24.6 -1.2 +42.2 Mixed30.8 media consumer n/a* n/a* 12.2 n/a* n/a* n/a* n/a* 14.1 +34.4 11.0 -3.6 -17.2 governmentMainly 33.0 37.1 critical media n/a* n/a* n/a* n/a* n/a* n/a* n/a* consumer +45.7 +39.7 -6.8 -8.6 80.2 No media consumer/ 25.0 29.7 27.3 No data n/a* n/a* +90.6 n/a* n/a* n/a* +43.3 +37.1 -12.1 -13.6 -14.9 -10.4 *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. 48 punish. compensation

How democratically elected politicians get away with autocratizing Hungary TRADE OFFS FOR SPECIFIC POLICY ISSUES Party loyalty has the greatest effect in Hungary when it comes to accepting the violation of democratic norms. Policy positions 15 play a much smaller role, but even here, undemocratic attitudes are always fully compensated for by the voters’ preferred policies. This means that voters in our experiment overlook the undemocratic attitudes of candidates and instead reward them with an increase in votes. In other words, depending on the issue at hand, they trade democratic standards for their preferred policies (trade-off). The“preferred” policy or position first refers to the entire policy area, and these can vary depending on the voter and the issue. Our results in the seven other countries we tested suggest that voters are most forgiving of undemocratic politicians when they appear in combination with identity-based policy preferences. In our experiment we consider the following issues to be primarily identity-based policy preferences: rights for same-sex couples and immigration. 16 Some would also consider the question of national vs. EU responsibility for defense policy to be identity related. Similarly, in countries such as Germany, the question of how to tax renewables and fossil fuels is seen as a social identity issue rather than a purely economical one. For Hungary, we do not see exactly this pattern. First and foremost, Hungarian respondents trade off democratic standards for all the policies offered; they make the largest trade-offs for their preferred policy position on the rights of same-sex couples. Thus, a politician who violates democratic norms would generally lose-4.8% of his vote share. However, he/she can make up for this initial loss by proposing the voters’ preferred policy on the rights of same-sex couples: this scenario leads to a+14.9% increase in vote share, which means that the candidate ends up with an overall reward of+10.1% in vote share. All other policy issues are similar: A politician who violates democratic principles would initially Policies Hungarian voters trade off democratic standards for all the policies offered. Trade-offs: what are democratic standards most likely to be exchanged for? All figures in percent punishment compensation Average 1 Punishment -4.8% Final reward Partisanship Same-sex couples' rights +14.9% +36.0% 2 10.1% 31.2% 3 Immigration +10.3% 5.5% Defense policy +10.0% 5.2% Education and income tax +9.6% 4.8% Energy taxation +9,0% 4.2% -5% 0%+5%+10%+15%+20%+25%+30% How to read this graph: 1 The regular punishment of undemocratic behavior among left-leaning voters equals-4.8%. 2 In this example,+36.0 percentage points is the absolute compensation rate that candidates receive if they belong to the voter's preferred party. 3 Relative compensation(or sometimes punishment) is the evolution of vote shares or the electoral reward/punishment that candidates ultimately receive when both the original punishment and compensation are taken into account. 15 An overview of all policy positions can be found on page 11. 16 For other countries, we also classified the use of minority language and the EU vs. Russia foreign policy-orientation as identity-related issues. 49

IDENTITY, PARTISANSHIP, POLARIZATION lose-4.8% of his/her vote share, could compensate by adopting the voter’s preferred policy on immigration by 10.3%. Offering the voters’ preferred position on defense policy would give him or her a compensation rate of+10.3%. The more socio-economic positions fare relatively high as well, with a compensation rate of+9.6% for the preferred position on education policy combined with net income taxes and a compensation rate of+9.0% for the preferred energy tax. WHAT INFLUENCE DO VOTERS’ CHARACTERISTICS HAVE ON THEIR WILLINGNESS TO TRADE OFF DEMOCRATIC STANDARDS FOR CERTAIN ISSUES? Age Do some age groups trade off more easily than others? When it comes to education policy or energy taxation, our respondents behave quite homogeneously across all age groups, with a difference in compensation of mostly 1-2%. The spread is wider for responsibility for defense policy: here the 30-39 year olds would compensate with an increase of+7.8% of the vote share, while the 66-75 year olds would compensate with an increase of+13.8%. This shows that the question of who is responsible for defense policy is more important for the older generation. On the question of immigration policy, one age group stands out: Voters aged 40-49 reward an undemocratic candidate at the ballot box with a +16.1% increase in their vote share. All other age groups reward their preferred immigration policy with a reward of between+8.6 and+9.8%. That same age-group(40-49) also stands out with regard to the question of rights for same-sex couples, as they reward their favorite policy offered with+20%. Individuals belonging to the group of 30-39 year olds reward with+15.3% and the 50-65 year olds with 16.2%. Surprisingly, both the youngest(18-29) and oldest (66-75) respondents are the least mobilized by this topic, with+11.6 and+9.5% respectively. Trading off democracy for favorite position on same-sex couples' rights By age group, all figures in percent punishment compensation 18-29 -4.2 years old +11.6 7.4 -3.5 30-39 +15.3 11.8 -5.1 40-49 +20.0 14.9 -5.3 50-65 +16.2 10.9 -5.8 66-75 +9.5 3.7 -5% 0%+5%+10%+15% 50

How democratically elected politicians get away with autocratizing Hungary Gender While women punish undemocratic behavior slightly more than men(by almost one percentage point), they also trade off democratic standards more against their preferred policy interests. The main difference between men and women is in the assessment of specific issues: women reward an undemocratic candidate for their preferred energy tax policy with+10.4% and for their preferred position on defense policy with+11.9%, while men would only reward the same with+7.3%(energy) and+7.9% (defense) respectively. Both genders are almost equally mobilized by the status of same-sex couples:+14.6%(men) and+15.1% (women). ucation levels is slightly higher than in Germany, for example, where education has little effect on the compensation rate. We find the biggest difference in compensation rates for defense policy, where people with lower education reward their favorite policy with+7.2%, those with secondary education with +12.0% and those with a university degree with 13.5%. The policy that moves the most evenly across education levels is energy taxation: respondents with a lower level of education compensate with+7.6%, followed by those with a university degree with+9.3% and those with a secondary education with +10.9%. Rural vs. urban We have seen that the rate of punishment decreases with the size of the settlement type. However, when it comes to trading democratic standards, we do not see this linearity. Instead, both capital city dwellers and village dwellers are comparatively equally mobilized by their respective preferred position on rights for same-sex couples(+16.2% and+16.6% respectively) and their preferred position on education/income tax. Respondents from towns and cities are least mobilized by their preferred position on immigration(+7.1%). Education As mentioned above, the rates of punishment vary considerably according to the level of education. Respondents with less education punish at-3.1%, those with secondary education at-5.4 %, and those with a university degree at-8.5%. These different starting points influence the final rate of punishment. In general, we see that people with a university degree compensate more than those with less education. And the difference between edNet income A peculiar picture emerges for respondents in the highest net income group: In the experiment, it is not even possible to derive a clearly quantifiable punishment rate from their behavior. If they did, it would be below average at-2.5%. However, when presented with undemocratic candidates on certain issues, a clear behavior can be observed. They make their highest compromise on education/income tax at+16.9%. Moreover, this issue does not play such a mobilizing role for any other net income group. The other issue that mobilizes those with the highest net in comes: Rights for same-sex couples. The findings for the middle-net income group(HUF 450,001 to HUF 650,000) are also notable: there is one issue that mobilizes them so little that respondents in this group would not tolerate undemocratic behavior for it, and that is education/income tax. Here, the compensation rate is+3.7%, but the general punishment rate for this group is-6.4% An undemocratic candidate with their preferred position would still be punished and would lose the election in our scenario. Trading off democracy for favorite immigration policy By Household Income. All figures in percent punishment compensation less than HUF 250,001 -3.8 +8.1 4.3 HUF 250,001 450,000 -3.8 +11.6 HUF 450,001 -6.4 650,000 +12.2 5.8 HUF 650,001 -8.6 1,000,000 +3.7 -4.9 -8%-6%-4%-2% 0%+2%+4%+6% 7.8 +8% 51

IDENTITY, PARTISANSHIP, POLARIZATION Trade offs In general, we find a trade-off behavior that is in line with the thematic priorities of the parties. Trading off democracy for favorite defense policy By Household Income. All figures in percent punishment compensation less than HUF 250,001 -3.8 +6.2 2.4 HUF 250,001 450,000 -3.8 +13.1 HUF 450,001 -6.4 650,000 +13.7 9.4 7.3 HUF 650,001 n/a* 1,000,000* -8%-6%-4%-2% 0%+2% *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. +4%+6%+8% When we look at the two clear identity issues, we see interesting differences in the extent to which voters in different net income groups would trade them for democratic standards The picture is more homogeneous when it comes to rights for same-sex couples. What is striking here is the unusually high level of compensation among the lowest net income group. At+16.5%, it is around twice as high as their compensation rates for all other issues. The three middle net income groups are equally concerned about immigration, with compensation rates ranging from just over+12% to just under+14%. The lowest net income group compensates less than half as much for an undemocratic candidate with their preferred position on immigration, at just+6.2%, and the highest net income group is indifferent. Our data thus suggest that LGBT rights mobilize people in the lowest net income group much more than immigration. Party supporters In general, we find a trade-off behavior that is in line with the thematic priorities of a party. For example, supporters of the right-wing parties Jobbik and MMNP would trade democratic standards for their preferred immigration policy with the highest compensation rates for this policy,+18.2% and+22.4% respectively. A closer look, however, paints a more ambiguous picture: Supporters of the ruling Fidesz party reward undemocratic candidates the most for their preferred policies on the rights of same-sex couples (+12.5% compensation) and responsibility for defense policy(+10.9% compensation). The issue that mobilizes their voters the least is immigration, with an offset of only+5.4%. Supporters of the right-wing Jobbik party compensate their rather average overall penalisation rate of-4.8% most strongly for their favorite policies on immigration(+18.1%) and the rights of same-sex couples(+16.3%). The far-right party Our Homeland is on a more moderate level in terms of democratic standards, with a compensation rate between+11.5 and 12.7% for three identity issues: immigration, same-sex couples’ rights and defense policy. The other right-wing party, the MMNP, is much more willing to trade off democratic standards, both for its preferred immigration policy(+22.4%) and for its preferred education/income tax policy(+20.1%). This is followed by rights for same-sex couples(+16.1%). And they do not really show any conclusive voting behavior when it comes to energy taxes or defense policy. Supporters of the center-left party DK would forgive an undemocratic candidate the most for their preferred policies on same-sex couples’ rights(+19.4%) and immigration(+13.6%). Supporters of the other center-left party, MSZP, in contrast, seem to compensate more for their preferred policy on energy tax and defense policy, but do not show a significantly conclusive enough behavior in our experiment. Voters of MM( Momentum Movement), who have the highest punishment rate, would probably not trade off democratic standards for 52

How democratically elected politicians get away with autocratizing Hungary punishment compensation 5.4% 11.5% 12.6% 18.2% 13.6% 15.2% 22.4% 16.5% 8.6% Party support and shifts in punishment when presented with favored policies Punishment and rewards for voter`s favorite policy position on immigration. Final reward 13.3% 2.7% 8.1% 5.2% 5.7% 5.6% 14.2% 10.1% -2.7% -3.4% -7.4% -4.8% -8.0% -9.6% -8.2% -6.4% +10% 4.3% +5% 0% -4.3% -5% -10% FideszKDNP Our Homeland Hungarian TwoTailed Dog Party (MKKP) JobbikConservatives Democratic Coalition (DK) Momentum Movement (MM) Everybody’s Hungary People’s Party(MMNP) Punishment and rewards for voter`s favorite policy position on same-sex couples` rights. 9.8% 8.1% 12.7% 11.4% 8.2% 6.1% 7.9% Other party/No answer Non-voters of the ´22 elections Average punishment Compensation by voters’ preferred policy 12.2% 10.4% +10% +5% punishment compensation 12.5% 11.5% 20.2% 16.3% 16.1% 15.7% 16.1% 18.6% 14.8% -2.7% -3.4% -7.4% -4.8% -8.0% FideszKDNP Our Homeland Hungarian TwoTailed Dog Party (MKKP) JobbikConservatives Democratic Coalition (DK) Due to rounding, totals may not correspond with the sum of the separate figures. -9.6% Momentum Movement (MM) -8.2% Everybody’s Hungary People’s Party(MMNP) -6.4% Other party/No answer -4.3% Non-voters of the ´22 elections 0% -5% -10% their preferred income tax/education policy or for their preferred energy tax policy, as they show inconclusive behavior in the experiment. They would trade the most for their preferred policy on rights for same-sex couples(+15.7%), immigration(+15.2%) and defense policy(+11.6%). Similarly, voters of the satirical Hungarian Two-Tailed Dog Party( MKKP) do not show consistent behavior on issues of income tax/education or defense policy. Their voters trade off other democratic standards more than any other voters for their preferred position on same-sex couples’ rights with a compensation rate of+20.2% and for their preferred position on energy tax with a compensation rate of+16.2. MKKP voters would also tolerate undemocratic behavior for their preferred policy on immigration(+12.6%). Non-voters are most willing to trade off democratic standards for their preferred position on the rights of same-sex couples (+14.8, if they vote). But socio-economic issues are also important to them: They would also tolerate undemocratic behavior for their preferred policies on education and income tax (+11.9%) and on energy taxes(+10.4%):+10.4%. They would reward an undemocratic candidate for their preferred immigration policy with+8.6%. The most political voters with clear policy preferences are voters of“other parties”. The question of who is responsible for defense policy mobilizes them the most and more than any other party:+19.5%. They would also tolerate undemocratic behavior by a candidate for their preferred position on the rights of same-sex couples(+18.6%), immigration(+16.5%), energy taxes(14.3%) and education/income taxes(13.7%). We cannot present conclusive results from LMP- and Dialogue Greens voters based on this experiment. 53

IDENTITY, PARTISANSHIP, POLARIZATION ARE SUPPORTERS OF A PARTICULAR POLICY MORE FORGIVING OF UNDEMOCRATIC BEHAVIOR THAN OTHERS? In the previous chapter we presented the average willingness of voters to forgive undemocratic behavior if the candidate stands for certain preferred policies. A question that remains to be answered, however, is whether this tendency can be observed across the political spectrum, i.e. whether there are differences between left- and right-leaning voters. In some cases, a policy issue may become so politicised that supporters of the policy feel very strongly about it. It is therefore useful to look at how supporters of a particular policy issue would forgive or not forgive a candidate’s undemocratic behavior. Same-sex couples’ rights Voters who support equal rights for same-sex couples(i.e. to marry and adopt children, a culturally left-leaning position) are generally less tolerant of undemocratic behavior than the country’s average voter, with a punishment rate of-7.3%. Right-wing voters, who would not allow same-sex couples to get married, are generally more tolerant of undemocratic behavior than the average, with a punishment rate of only-1.9%. Notably, this punitive behavior of socially left-wing voters shifts decisively when voters are presented with a candidate who(in addition to proposing an undemocratic policy) undemocratic policies) also proposed their preferred policy on rights for same-sex couples. same-sex couples: Left-leaning voters then became much more tolerant of undemocratic candidates. candidates(see graph to visualise the shift). However, the final reward for their preferred position remains significantly lower for left-leaning voters. Same-sex couples' rights: shift in punishment when presented favorite policy By political orientation. All figures in percent punishment compensation Left -7.2 +16.6 9.4 Center -4.4 +13.0 8.6 Right* -1.8 +13.5 11.7 -6%-4%-2% 0%+2%+4%+6%+8%+10% *Whenever either one of the punishment or compensation variable was not statistically significant we refrained from reporting it in this table. Immigration As in the other European countries in our previous survey, left-leaning voters value democracy more on the issue of immigration than right-leaning ones, as they punish it more than those on the right. In other words, the more left-leaning a voter is when it comes to immigration, the more likely they are to punish undemocratic behavior. But we can also observe that voters with a far-left position(“allow all immigration regardless of the country of origin”) also compensate for their preferred position to a much greater extent than the others. Voters with a centerleft position and a center-right position(“banning immigration from outside EU”) would both compensate for their respective preferred policy to a very similar extent. However, center-left voters are more likely to punish undemocratic behavior. Immigration: shift in punishment when presented with favorite policy By political orientation. All figures in percent punishment compensation Left -8.9 +14.5 5.6 Center -7.2 Left +9.3 2.1 Center Right -4.6 +9.9 5.3 -8%-6%-4%-2% 0%+2%+4% As we did not get statistically signifi cant results for the category"Right", we are not reporting any results for it. 54

How democratically elected politicians get away with autocratizing Hungary Energy tax Green voters are slightly more likely to penalize undemocratic behavior(-5.7%) than fossil fuel voters(-4.5%). But fossil fuel supporters are twice as likely as green energy supporters to trade off democratic standards for their energy preference(+10.9% vs.+5.2%). In fact, the latter would not punish an undemocratic candidate even if he offered their preferred energy policy. Income tax and education policy Libertarian respondents are the most forgiving of undemocratic candidates in our survey, punishing them by-4.2%. Undemocratic candidates lose-5.0% of the votes of people who want to keep income tax rates and the education budget as they are; and the same candidates lose-7.4% of the votes of people with a more left-wing economic position. However, when presented with their preferred policy on taxes and education, two groups become very forgiving of undemocratic behavior: people with a libertarian position compensate the candidate with+11.8%, thus rewarding them with+7.6%; people with a centrist position compensate with+6.6% the vote share, thus rewarding them+1.6%. Economically left-leaning people, however, do not tolerate undemocratic behavior for their preferred position and still punish it with a final-1.1% vote share loss(having traded off +6.3% of votes). Defense policy Respondents who prefer the EU to have sole responsibility for defense policy punish undemocratic candidates the least, at -3.5%, lower than the Hungarian average. Respondents who think that defense policy should be a joint responsibility of Hungary and the EU punish at a rate of-5.2% and those who think Hungary’s defense policy should be the sole responsibility of Hungarian institutions punish slightly less, at-4.0%. This group would trade off their preferred position at the highest rate, compensating such a candidate with+14.6%. Respondents with a cooperative view would tolerate undemocratic behavior by compensating a candidate with their preferred policy offer with +9.3% and respondents who would prefer the EU to determine defense policy would compensate their preferred candidate with +5.7%. Energy tax: shift in punishment when presented with favorite policy All figures in percent punishment compensation Green -5.7 leaning voters +5.2 -0.5 Fossil -4.5 fuel +10.9 6.4 supporters -4%-2% 0%+2%+4%+6% Income tax and education policy: shift in punishment when presented with favorite policy By political orientation. All figures in percent punishment compensation Left -7.4 +6.3 -1.1 Center -5.0 +6.6 1.6 Right -4.2 +11.8 7.6 -6%-4%-2% 0%+2%+4%+6% Defense policy: shift in punishment when presented with favorite policy All figures in percent punishment compensation EU& -5.2 National +9.3 4.1 National -4.0 only +14.6 10.6 -4%-2% 0%+2%+4%+6%+8%+10% Main Takeaways: Partisanship emerges as the most dominant factor influencing voter behavior in Hungary, surpassing concerns over anti-democratic action. Furthermore, the analysis above also reveals that in addition to party loyalty, Hungarian voters also trade off democracy for policy priorities. While voters in the other European countries we studied may prioritize identity-based policy preferences over democratic norms, Hungarian respondents exhibit a distinct pattern of prioritizing policy positions across all areas over democratic principles. This tendency is particularly pronounced in issues such as rights for same-sex couples and then- with some distance only- immigration policy. 55

IDENTITY, PARTISANSHIP, POLARIZATION POLARIZATION Is it becoming easier for politicians to establish autocratic governance in polarized societies? In our previous chapter, we examined which policy areas Hungarian voters prioritized over safeguarding democracy. This chapter examines whether polarization contributes to this tendency to prioritize other partisan or political concerns over democratic values. While a certain level of polarization, which indicates the ideological gap between opposing political factions, is typically indicative of a robust democratic system that provides voters with clear and diverse programmatic options, there is an emerging consensus among scholars that increased polarization is detrimental to de mocracy. Various scholars(Graham and Svolik 2020; Svolik 2019; McCoy and Sommer 2019; Nalepa, Vanberg, and Chiopris, 2018) argue that polarization fosters democratic erosion. They suggest that voters may tolerate undemocratic politicians when the opposition’s policy positions differ significantly from their own beliefs(Nalepa, Vanberg, and Chiopris, 2018). Therefore, in what follows, we examine whether there is increased polarization in Hungary with regard to the political issues on which Hungarian respondents are more inclined to tolerate undemocratic behavior. To assess the polarizing nature of the political issues used in our experiment, we asked respondents about their agreement(on a scale of 0 to 10) with a series of political statements. This approach allowed us to measure political polarization based on voters’ preferences and to identify specific divisive issues in Hungary. In addition to asking respondents about their agreement with a set of policy statements, we also asked them about the importance of these policy positions to them. This approach allowed us to examine whether political polarization is more pronounced for issues that are considered important by voters, or whether there is no correlation between perceived policy importance and political polarization. Our results show that, compared to the countries surveyed so far, Hungary and Ukraine have the highest levels of polarization among the countries surveyed, scoring 0.46 on a scale of 0 to 1. Conversely, Germany emerges as the least polarized country in our sample, with a score of 0.35. Regarding specific policy domains, our findings indicate that Hungarian respondents demonstrate the highest polarization concerning the rights of same-sex couples(0.55), followed by the policy area on taxes and education(0.49), immigration(0.46), energy taxes(0.42) and lastly, the area of defense policy(0.37). An intriguing observation emerges: while the rights of same-sex couples are the most polarizing policy area in HunWhich issues let politicians get away with undemocratic behavior? Issues displayed according to their importance to voters and degree of polarization Size of the bubble represents the average trade-off reward for each policy 8 more important (1 10) 7 Education/ Energy taxation income tax 6 Defense policy 5 Immigration 4 3 0,2 0,3 Same-sex couples’ rights 0,4 0,5 0,6 more polarizing(0,0- 1,0) gary, respondents perceive it as the least important. Conversely, Hungarian respondents consider taxes and education to be the most important policy areas. Moreover, unlike the other seven countries analyzed in our previous study, Hungary does not stand out as an outlier in terms of increased polarization over the rights of same-sex couples. We observed comparable results in the other Eastern European countries in our sample: Estonia, Poland, Serbia and Ukraine. Regarding same-sex couples’ rights policy, our findings in fact indicate different degrees of polarization among different social groups. While people with a higher university degree are slightly more polarized on this issue(0.60) than those with a secondary education(0.56) or lower education(0.48). Furthermore, the analysis by age group shows that the highest polarization(0.64) on the rights of same-sex couples is among respondents aged 66-75, while the lowest polarization (0.39) is among those aged 76 and over. With regard to the political issue of taxes and education, our results show differences in polarization levels between different age groups. In particular, the highest level of polarization is observed among respondents in the middle age groups(4056

How democratically elected politicians get away with autocratizing Hungary 49 and 50-65), with a polarization measure of 0.52. In contrast, the lowest level of polarization is observed among the oldest respondents(aged 76 and over), with a measure of 0.24. In addition, when we look at polarization levels among different education groups on taxes and education, we see a contrasting trend compared to the policy area of same-sex couples’ rights. In this case, people with a university or college degree are less polarized(0.35) than those with lower(0.45) or secondary(0.48) education. In terms of gender, there are only small differences between the polarization levels of female (0.48) and male(0.47) voters on this issue. POLARIZATION AND DEMOCRATIC TRADE-OFFS In the previous chapter, we noted a striking difference between Hungary and the other seven European countries in terms of the willingness to trade off other policy preferences against democracy. While respondents in the other countries tend to compromise democracy primarily for identity-based policy preferences, Hungarian respondents showed a unique pattern in their willingness to trade democracy for all policy areas. In Hungary, the issue of rights for same-sex couples’ emerges as the most polarizing policy area. Notably, it is also the policy area where Hungarians are most willing to compro mise democratic values. This correlation between increased polarization and voters’ willingness to trade democracy for their preferred policy on same-sex couples’ rights is not unique to Hungary. In our comparative analysis of the seven countries surveyed, similar patterns emerge in all but Spain. Furthermore, the second most polarizing policy area in Hungary revolves around taxes and education, which are also the most important policy areas for Hungarian voters. Interestingly, while Hungarian voters express a willingness to compromise democratic values for their preferred policies on taxes and education, this inclination is less pronounced compared to other issues. In contrast to the correlation observed in the area of rights for same-sex couples, here we see a weaker link between increased polarization and an increased willingness to trade off democracy. With regard to the policy area of immigration, which has been a prominent issue in Hungarian politics in recent years, our research indicates that it ranks as the third most polarizing issue. Although less polarizing than the policy areas of same-sex couples’ rights and taxes and education, it remains a significant concern for Hungarian voters, leading to a remarkable willingness to trade off democratic values for their preferred stance on immigration. It is the second policy area where Hungarian voters are most willing to compromise democratic principles, just after the issue of same-sex couples’ rights. Main Takeaways: Hungary and Ukraine exhibit the highest levels of polarization among the countries analyzed. Unlike other nations where polarization often centers on identity-related issues, Hungary’s debate on same-sex couples’ rights is most divisive, followed by taxes, education, and immigration policy discussions. This polarization correlates with voters’ readiness to compromise democratic norms for preferred stances on same-sex couples’ rights, a trend seen in most surveyed countries except Spain. Interestingly, although same-sex couples’ rights is most polarizing in Hungary, it’s perceived as least important by respondents, contrasting with taxes and education, deemed most crucial. These findings underscore politicians’ pivotal role in shaping public discourse and prioritizing issues on the political agenda. 57

IDENTITY, PARTISANSHIP, POLARIZATION CONCLUSION Have Hungarians moved away from democracy? Is Hungarian democracy resilient enough to withstand efforts towards autocratization? Where are its weaknesses and where are its strengths? With this study, we wanted to answer these and similar questions, thereby contributing to the understanding of how democratic decay functions and how democratic resilience can be strengthened. In doing so, we did not only rely on the usual form of direct questioning in a survey, but also included the results of an experiment. Through experiments on candidate selection, the respondents of our survey were presented with scenarios that closely resemble real election situations. This allowed us to investigate under what circumstances and in exchange for which political benefits certain voters tend to overlook violations of democratic principles. This enables us to draw the following conclusions. Hungarians consider democracy important, but are dissatisfied with its quality In Hungary, the vast majority of society considers it important to live in a democracy and believes that democracy is better than other forms of government. The popularity of democracy in Hungary is similar to other Eastern European countries(Serbia, Poland, Estonia), but lags behind Western European countries(Germany, Sweden). Pro-democracy attitudes are in the majority among supporters of all parties. However, there are significant differences: 60% of pro-government voters and almost 90% of opposition voters consider it important to live in a democracy. At the same time, our findings reflect widespread dissatisfaction with the functioning of democracy in Hungary. Approximately two-thirds of respondents expressed dissatisfaction with the state of democracy(67%) and perceived a lack of democratic governance (64%). When it comes to the quality of democracy, there are parallel realities between pro-government and opposition voters. While a significant majority of opposition voters express dissatisfaction with the state of Hungarian democracy, and see it as undemocratically governed, the majority of Fidesz voters hold the opposite view. Our study confirms that the majority of Hungarians view the current state of democracy through a partisan lens. Furthermore, our analysis also reveals gender differences on these topics. More female respondents are dissatisfied with the state of democracy than male respondents. Furthermore, more female respondents think that Hungary is not governed democratically. Democratic competence of Hungarians is low by international standards Compared to other European countries where this indicator was measured, democratic competence in Hungary was the lowest, with 72% of respondents demonstrating knowledge of democratic principles. The level of democratic literacy appears to be positively correlated with the level of democracy in a country. Those countries characterized by a lower proportion of democratically competent citizens also tend to have deficits in democratic governance. At the same time, democratic competence is higher on average in more democratic countries. 58

Strengthening young people's knowledge of how democracy works is crucial For the future development of Hungarian democracy, it is particularly important to raise young people’s awareness of the significance and functioning of democracy and democratic institutions. Not only do younger age groups consider it less important to live in a democracy than older age groups, but the youngest cohort also has the lowest level of democratic literacy. The central role of education is underlined by empirical evidence showing a significant gap in democratic competence between individuals with different levels of educational attainment. In particular, those with higher levels of education show a significantly greater ability to distinguish between actions that are democratic and those that are not. Strengthening civic education in public schools is therefore likely to reduce the prevalence of misconceptions such as the notion that opposition-led demonstrations, constitutional court rulings challenging government policies, or media discourse critical of the government are inherently antithetical to democratic principles. By international comparison, Hungarian voters are less likely to punish politicians for breaking democratic norms Our results show that the average Hungarian voter is willing to hold politicians accountable for violating democratic principles, although to a lesser extent than voters in the previous FES comparative study. In Hungary, the average loss of vote share that a politician suffers for violating democratic principles is-4.8%, which is the lowest of all the countries examined in our research. It is noteworthy that the penalty in Hungary is about half of that in Poland (-9.2%), despite the fact that both countries have experienced significant levels of autocratization in recent years. Undemocratic politicians are punished more by those who are generally dissatisfied with the state of Hungarian democracy Our findings also reveal correlations between punishment rates for undemocratic politicians and other factors, such as perceptions of the current state of Hungarian democracy or party affiliations. Respondents who are dissatisfied with the state of Hungarian democracy punish undemocratic politicians at a rate of-7.2%, while those who are satisfied with the current situation would impose a punishment rate of-2.4%. Voters with a better understanding of democracy are also more likely to punish undemocratic politicians. This is because, in general, those who consider democracy important in Hungary today and know how democracy works are more likely to have a poor opinion of the current state of democracy in the country and of Fidesz's activities regarding democracy. This is confirmed by the fact that supporters of the Fidesz-KDNP coalition have the lowest disapproval rating at-2.7%, while supporters of the liberal opposition party Momentum Movement have the highest disapproval rating at-9.7%. Of all the political parties studied in our comparative research, Fidesz voters are the most tolerant of undemocratic behavior It is important to emphasize that among all the partisan subgroups examined in our study, Fidesz-KDNP supporters clearly emerge as the most tolerant when it comes to democratic violations. They are the most lenient of all political groups, regardless of the specific national context. These results confirm that, both in Hungary and in the seven European countries surveyed, supporters of radical right-wing parties and non-voters are the least likely to punish undemocratic behavior. The notable difference between Poland and Hungary is that not only do PiS voters punish undemocratic behavior slightly more than Fidesz voters, but the former Polish opposition parties that came to power in the autumn elections in 2023 also have a higher punishment rate than the Hungarian opposition parties.

Women and older, better educated voters from Budapest more likely to punish undemocratic candidates Similarly to the other countries we investigated, on average, women punish undemocratic politicians more than men. Furthermore, highlighting the age divide, older demographic groups tend to be more critical of undemocratic candidates than their younger counterparts, with punitive attitudes increasing steadily with age. One possible explanation for this pattern is the relatively low level of political engagement, disinterest, and disillusionment among Hungarian youth, especially compared to their peers in Central and Eastern Europe. Protests, with a few exceptions such as climate activism or university demonstrations, are largely attended by older generations in Hungary. When it comes to violations of specific democratic principles, it is noteworthy that young people are particularly adamant about punishing the prosecution of journalists who expose government corruption. Our data underline the stark urban-rural divide in Hungary: Residents of Budapest, where support for the opposition is higher than the national average, show a robust-6.9% punitive attitude towards undemocratic behavior, in contrast to the more subdued-3.6% punitive rate observed among villagers, where Fidesz enjoys strong support. There is also a clear educational trend, with those with higher levels of education tending to be more emphatic in their punitive responses. In terms of specific transgressions, voters are most likely to punish the violation of electoral rules, followed by the crackdown on journalists and the violation of the independence of the judiciary. The lingering effects of past government campaigns against the media and civil society are evident in our findings: voters exposed to predominantly pro-government media are disinclined to punish politicians who advocate a ban on foreign NGO funding, instead favoring candidates who support such measures. Public policy positions more important for identification with politicians than violations of democratic norms In Hungary, party loyalty emerges as the most influential factor in determining acceptance of violations of democratic norms. Policy positions play a much smaller role, but even here undemocratic attitudes are always fully compensated for by voters' preferred policies. In contrast to respondents in the seven other countries surveyed, who tend to compromise democracy primarily for identity-based policy preferences, Hungarian respondents show a unique pattern in their willingness to trade democracy for all policy areas. In particular, they are most willing to compromise democracy for their preferred policy position on rights for same-sex couples, followed only with some distance by immigration policy. Partisanship outweighs anti-democratic behavior even more in Hungary than in other countries Consistent with the findings of previous FES research, our study shows that voters have a remarkable tendency to overlook undemocratic behavior, especially when such behavior is associated with their preferred political party. A comparative analysis between Hungary and seven other European countries shows that Hungarian respondents not only exhibit the lowest level of democratic competence, but also the highest level of loyalty to their respective political groups. Polarization over public policy issues in Hungary is high by international standards and contributes to tolerance of violations of democratic norms Our findings suggest that Hungary and Ukraine have the highest levels of polarization compared to the countries we have analyzed so far. Moreover, unlike in the other countries where we have mainly observed increased polarization along identity-related political issues, in Hungary the debate on the rights of same-sex couples emerges as the most divisive issue, followed by discussions on taxes, education and immigration policy. The correlation between increased polarization and voters’ willingness to compromise democratic principles for their preferred position on same-sex couples’ rights is not unique to Hungary. Our comparative analysis of the seven countries surveyed reveals similar trends in all but Spain. It is noteworthy that while same-sex couples' rights is the most polarizing policy area in Hungary, respondents perceive it as the least important. On the contrary, Hungarians see taxes and education as the most important areas. These findings underline the central role and responsibility of politicians in shaping public discourse and prioritizing issues on the political agenda. 60

How democratically elected politicians get away with autocratizing Hungary LITERATURE Ágh, A.(2016). The decline of democracy in East-Central Europe: Hungary as the worst-case scenario. Problems of Post-Communism, 63(5–6), 277–287. https://doi.org/10.1080/10758216.2015.1113383 Anghel V., Jones, E.(2024), What went wrong in Hungary. Journal of Democracy, 35(2), 52-64. Avramovska, E., Lutz, J., Milačić, F. and Svolik, S.(2023): Identity, Partisanship, Polarization. How democratically elected politicians get away with autocratizing their countries. https://library.fes.de/pdf-files/bueros/wien/19536-20220926.pdf Bermeo, N.(2016). On democratic backsliding. Journal of Democracy, 27(1), 5-19. Bernhard, M.(2021). Democratic backsliding in Poland and Hungary. Slavic Review, 80(3), 585–607. https://doi.org/10.1017/slr.2021.145 Bartels, L. M.(2023). Democracy erodes from the top: Leaders, citizens, and the challenge of populism in Europe. Princeton University Press. Cianetti, L., Dawson, J.,& Hanley, S.(2018). Rethinking„Democratic Backsliding“ in Central and Eastern Europe – Looking beyond Hungary and Poland. East European Politics, 34(3), 243–256. https://doi.org/10.1080/21599165.2018.1491401 Court of Justice of the European Union.(2020). Judgement in Case C-78/18 Commission vs. Hungary Dahl, R. 1971. Polyarchy: Participation and opposition. New Haven, CT: Yale University Press. Dahl, R. 1989. Democracy and its critics. New Haven: Yale University Press. Gherghina, S.,& Volintiru, C.(2023). Budgetary Clientelism and Decentralization in Hungary and Romania. Journal of Developing Societies, 39(1), 40–62. https://doi.org/10.1177/0169796X221148509 Graham, M.,& Svolik, M.(2020). Democracy in America? Partisanship, polarization, and the robustness of support for democracy in the United States. American Political Science Review, 114(2), 392–409. Grzymala-Busse, A., Kuo, D., Fukuyama, F.,& McFaul, M.(2020). Global populisms and their challenges. Freeman Spogli Institute for International Studies, Stanford University. Grzymala-Busse, A.(2019). How populists rule: The consequences for democratic governance. Polity, 51, 707–717. Lewandowsky, M.,& Jankowski, M.(2023). Sympathy for the devil? Voter support for illiberal politicians. European Political Science Review, 15(1), 39–56. https://doi.org/10.1017/S175577392200042X Linz, J. and Stepan, A. 1996. Toward Consolidated Democracies, Journal of Democracy 7: 14-33. McCoy, J.,& Somer, M.(2019). Toward a theory of pernicious polarization and how it harms democracies: Comparative evidence and possible remedies. The Annals of the American Academy of Political and Social Science, 681, 234–271. Nalepa, M., Vanberg, G.,& Chiopris, C.(2018). Authoritarian backsliding. https://doi.org/10.1177/00104140211060285 Nord, M., Lundstedt, M., Altman, D., Angiolillo, F., Borella, C., Fernandes, T., Gastaldi, L., Good God, A., Natsika, N.,& Lindberg, S. I.(2024). Democracy report 2024: Democracy winning and losing at the ballot. V-Dem Institute, University of Gothenburg. Policy Solutions(2022): Divided Hungary. Political Polarization of the Hungarian Society. https://policysolutions.hu/userfiles/elemzes/341/policy_solutions_divided_hungary_2022_en.pdf Saikkonen, I.,& Christensen, H. S.(2023). Guardians of democracy or passive bystanders? A conjoint experiment on elite transgressions of democratic norms. Political Research Quarterly, 76(1), 127-142. Scheppele, K. L.(2022). H ow Viktor Orbán wins. Journal of Democracy, 33(3), 45-61. Svolik, M.(2019). Polarization vs. democracy. Journal of Democracy, 30, 20–32. Svolik, M., Avramovska, E., Lutz, J.,& Milačić, F.(2023). In Europe, democracy erodes from the right. Journal of Democracy, 34(1), 5–20. https://doi.org/10.1353/jod.2023.0000 Vachudova, M.(2020). Ethnopopulism and democratic backsliding in Central Europe. East European Politics, 36, 318–340. https://doi.org/10.108 0/21599165.2020.1787163 Varieties-of-Democracy. Democracy Report 2022. Autocatisation Changing Nature? Visnovitz, P.,& Jenne, E.(2021). Populist argumentation in foreign policy: The case of Hungary under Viktor Orbán. Comparative European Politics, 19, 683–702. https://doi.org/10.1057/s41295-021-00256-3 Wunsch, N.,& Gessler, T.(2023). Who tolerates democratic backsliding? A mosaic approach to voters’ responses to authoritarian leadership in Hungary. Democratization, 30(5), 914–937. https://doi.org/10.1080/13510347.2023.2203918 61

IDENTITY, PARTISANSHIP, POLARIZATION Imprint FES Regional Office for International Cooperation Democracy of the Future Reichsratsstr. 13/5, A-1010 Vienna Responsible for Content: Johanna Lutz| Director, Democracy of the Future Tel.:+43 1 890 3811 301 @FES_Democracy democracy.fes.de Orders/ Contact: democracy.vienna@fes.de ISBN 978-3-98628-486-2 Commercial use of all media published by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES. The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung or of the organization for which the author works. 62

democracy.fes.de