STUDY UNEQUAL DEMOCRACIES WHO DOES (NOT) VOTE IN ROMANIA? This report examines, comparatively, the evolution of voter turnout in Romania since the 1990 elections, focusing on the significant decline in electoral parti­ cipation and on the most inportant inequalities iden­ tified in data from official sources, public opinion surveys, and exit-polls. Although most post-communist countries have re­gistered voter turnout decline since the founding elections, the decline was the highest in Romania. This can be explained, among others, by high emi­ gration and low levels of interest in politics, linked to low levels of trust in political parties. Mircea Com ş a and Claudiu Tufi ș November 2024 The report suggests different solutions to increase voter turnout: remove institutional barriers and im­prove procedures for diaspora voters, convince poli­tical parties to come up with a better offer for voters, and decrease the voting age to 16 for the European Parliament elections, at least.

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? UNEQUAL DEMOCRACIES WHO DOES (NOT) VOTE IN ROMANIA? B

Content INTRODUCTION 2 ROMANIA IN COMPARATIVE PERSPECTIVE 3 INEQUALITIES IN VOTER TURNOUT IN ROMANIA 6 INCREASING VOTER TURNOUT BY ­LEGISLATIVE MEASURES  15 CONCLUSIONS AND 18 REFERENCES 19

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? INTRODUCTION The evolution of democracy in some central and eastern European countries – such as Hungary, Poland or Romania – over the past decade has been characterised as“democratic backsliding”(International IDEA, 2023) or even(episodes of) autocratisation(V-Dem, 2024). Given this worrying develop­ ment, it makes sense to take a step back and re-evaluate the performance of the political system in the region, as well as people’s political participation. In this report, which is part of the FES Unequal Democracies series, we take a step back and focus our attention on political participation, in its purest form: voting in elections. Democracies, using the simplest definition, are built on com­ petition and participation(Dahl, 1971). Weakening either ele­ ment may have negative consequences for the quality of the democratic system and, unfortunately, Romania seems to be experiencing a significant decline in political participation. While in the early 1990s, in its founding elections, Romania registered a turnout of more than 85%, by the 2020s turnout had fallen below 40%, one of the most precipitous declines in the region. This means that the majority of the population are not taking part in the basic game of a democratic system, the electoral game. Starting from this observation, we are interested in under­ standing the people who have a say in the electoral process, the voters, and the people whose voices have been muted, for various reasons, the non-voters. Various population subgroups are absent from the electoral process and so political parties ignore their views and interests, with consequences ranging from complete withdrawal from the political scene to rejecting democracy and embracing authoritarian alternatives. We start the discussion by putting Romania in a compara­ tive context, using the Unequal Democracies Comparative Dataset(Wenker, 2024). Next, we identify and analyse ine­ qualities in voter turnout using both official turnout data and self-reported turnout data from public opinion surveys and exit polls. Finally, we identify and discuss ways in which voter turnout could be increased and offer some recommenda­ tions based on this discussion. 2

Romania in comparative perspective ROMANIA IN COMPARATIVE PERSPECTIVE VOTER TURNOUT Voter turnout in parliamentary elections in central and east­ ern European(CEE) countries is around 20 percentage points lower than in western European countries(Pet ričević and Stockemer, 2020). According to the Unequal Democracies (UD) Comparative Dataset, average turnout in Romania since 1990 is 55%, the lowest among all 30 countries in the data set(see Figure 1). But although official turnout in Romania is the lowest, reported turnout is similar to turnout in other countries in the UD dataset. As a result, the average turnout overestimation(the difference between official and reported turnout) is one of the highest in Romania. What factors could explain these differences? Most post-election surveys, regardless of country and election type, show that reported turnout is significantly larger than official turnout(DeBell et al., 2020). Turnout overestimation in surveys is mainly the result of two factors: nonresponse bias (the cooperation rate is higher for voters than for non-voters) and social desirability bias(people are ashamed to admit that they did not vote). Other, minor factors are memory failures and health conditions(Brenner, 2021; Com ș a& Postelnicu, 2013; McAllister& Quinlan, 2022). In the case of Romania, the emigration rate also plays a role in turnout overestimation. A significant part of the Romanian population(around 2.6 mil­ lion, equivalent to 14% of registered voters) is working or living abroad. For several reasons, the cost of voting is much higher for emigrants, so they vote at lower rates. Figure 1 Average official and reported voter turnout(%) 100 90 84.3 80 70.1 70 83.5 72.0 82.3 73.6 60 50 40 30 20 10 0 Northern America Northern Europe Western Europe Official voter turnout Data source: UD Comparative Dataset(general election level). 3 74.3 58.3 80.1 63.4 Eastern Europe Southern Europe Reported voter turnout 78.3 54.7 Romania

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? Figure 2 Reported voter turnout by gender, age and education(%) 100 90 83 80 81 80 77 75 74 72 70 85 88 80 80 82 78 79 80 72 71 70 88 80 77 75 73 82 84 66 62 60 50 40 30 20 10 0 Male Female GENDER 18–29 30–44 45-60 60–99 Low AGE Western Europe Eastern Europe Romania Data source: UD Comparative Data Set(general election level). Medium EDUCATION High INEQUALITIES IN REPORTED TURNOUT For the reasons already mentioned, reported turnout exceeds real turnout. It should be noted, however, that the UD Comparative Dataset controls for this higher reported turnout, which means that despite the overes­ timation, differences in reported turnout among various population groups reflect real differences. In other words, the finding reported below that the turnout difference between men and women in Romania is very small is statistically robust, even though the actual turnout rate might be lower than the reported one. Whenever we refer to reported turnout inequality, we mean the voting differences that are found in survey-based voter research such as the UD Dataset. Because overestimation of turnout depends on the charac­ teristics of the survey respondents, the data presented in this section should be regarded as indicative. Based on the results presented here, we can argue that reported turnout is gen­ erally similar for women and men, but also that it increases with age and formal education(see Figure 2). Moreover, these findings are relatively similar across groups of coun­ tries, and Romania does not seem to differ significantly from other countries in the region. The level of inequality in reporting turnout varies across dif­ ferent groups of countries(see Figure 3). As a general trend for all countries, inequalities in reporting turnout are lower for gender and higher for age and education. In the case of Romania, differences in reporting turnout are a bit higher for gender and a bit lower for age and education, meaning that in Romania men tend to overestimate their turnout more than women, while the differences in overestimating turnout among various age and education groups are lower than in other European countries. It should also be noted that differences in reported turnout do not change very much over time(1990–2009 vs 2010–2023 elections). Inequality by gender is quite stable, while inequality by age or education increases a little, but only in some countries. Summing up, the data in this section show that Romania does not differ very much from the other countries in terms of inequality of reported turnout by gender, age or education. Differences in terms of general overreporting of turnout and of the neg­ ative effect of high levels of income inequality on voter turn­ out are discussed in the next section. 4

Romania in comparative perspective Figure 3 Inequalities in reported voter turnout by gender, age and education 14 12 11.7 12.3 11.7 10 9.5 8 6 8.6 5.9 4 3.6 2.4 2.4 2 0 Western Europe Eastern Europe Gender Age Education Note: A larger value indicates higher variation. Data source: UD Comparative Data Set(general election level). Romania 5

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? INEQUALITIES IN VOTER TURNOUT IN ROMANIA WHAT VOTER TURNOUT? OFFICIAL, VOTING-AGE POPULATION OR VOTING-ELIGIBLE POPULATION? Voter turnout has decreased in most democracies over the past 50 years(Elsässer et al., 2022; Hooghe and Kern, 2017). Studies offer a variety of explanations for this, including generational change, the rise in the number of elective insti­ tutions, the democratic context, the abolition of compul­ sory voting, concurrent elections, inflation and economic globalisation(Frank and Martínez i Coma, 2023; Kostelka and Blais, 2021). Romania is not exceptional on this matter. Regardless of the measure we use, the conclusion is clear: the turnout rate is declining in Romania(see Figure 4). What we are ques­ tioning is the steepness of the decline across time periods. According to official data(BEC, Central Electoral Bureau), the turnout rate in general elections has declined in Roma­ nia by about 40 percentage points(from 86% in the 1990 founding elections to 32–48% in the past two elections). The Voting-Age Population(VAP) turnout follows the offi­ cial turnout closely. Accounting for voting eligibility and the difficulty(cost) of voting gives us a partially different pic­ ture: from the founding elections to the 2008 Parliamentary Figure 4 Voter turnout in Romania: BEC(REG, official) vs INS(VAP) vs VEP 100 89.9 90 88.9 80 86.2 76.6 78.5 75.3 76.3 68.6 70 75.0 76.0 66.3 66.0 60 65.8 65.3 54.4 60.5 50 58.5 54.3 50.6 48.0 40 42.0 39.6 41.8 30 39.2 11.711.6 20 7.8 8.3 8.8 5.8 10 3.7 2.2 2.5 2.8 3.3 0.9 1.6 0.3 0 1990 1992 1996 2000 2004 2008 8.6 8.8 2009 2012 61.2 53.1 53.2 8.1 8.0 2014 46.4 40.1 39.8 6.3 6.6 2016 56.1 47.7 47.6 8.5 8.4 2019 39.2 33.2 31.9 7.3 6.0 2020 Voter turnout(%): VAP REG VEP Voter turnout underestimation(pp): VAP REG Note:(1) Data source: authors’ computations based on data from BEC(Central Electoral Bureau), ROAEP(Romanian Electoral Authority), INS(Na­ tional Institute of Statistics), ANPDPD(National Authority for the Protection of the Rights of Persons with Disabilities), UN(International Migrant Stock), and Eurostat(EU and EFTA citizens who are residents in another EU/EFTA country. DOI: 10.2908/migr_pop9ctz).(2) Parliamentary elec­ tions in 2008, 2012, 2016, 2020; Presidential elections 2009, 2014, 2019; Parliamentary and Presidential elections 1990, 1992, 1996, 2000, 2004. (3) REG= Registered Population(BEC), VAP= Voting Age Population(INS), VEP= Voting Eligible Population(own estimation). 6

Inequalities in voter turnout in Romania Box 1 How do we measure turnout? The literature uses three different measures of voter turnout. They all have in common the number of people that showed up to vote for a particular election, but they differ in the denominator, the reference population, they use. The three measures are:(1) Official turnout is the percentage of registered voters who actually vote. The registered voter category does not count people who are eligible to vote but are not registered, for various reasons, ranging from not having official ID papers to living abroad.(2) Voting-age population turnout(VAP) is the percentage of voting age people who vote. The voting-age category might include people of voting age but who are not eligible to vote and it might exclude people who are of voting age but that are living abroad and are not registered to vote. (3) Voting-eligible population turnout(VEP) is the percentage of people who are eligible to vote and who do vote. The voting-eligible category is an attempt to solve the measurement problems of the other two turnout measures by counting only people who have the right to vote, regardless of their residence or migration status. For more details, see Stockemer(2017). elections, the decline was rather abrupt, on all indicators. Since 2008, the Voting-Eligible Population(VEP) turnout has continued to decrease, but at a lower rate. In many former communist countries in central and eastern Europe, Romania being the most preeminent case, the eli­ gible voter population is much smaller than both the vot­ ing-registered population and the voting-age population. This is explained by a higher emigration rate in Romania (Com ș a, 2015). The various costs of voting(distance to the closest polling station, travel cost, time spent waiting in line, replacing expired IDs and passports) are much higher for emigrants than for people voting in their own countries (Szulecki et al., 2021). The negative impact of emigration on voter turnout is larger in Romania but similar effects can be identified in most of the former Communist countries in the region(Com ș a, 2017; Kostelka, 2017). WHAT DOES THE OFFICIAL DATA SAY ABOUT TURNOUT INEQUALITY IN ROMANIA? Reported turnout computed on the basis of survey data is usually higher than official turnout(DeBell et al., 2020). This can be a problem when, due to misreporting of voting and nonresponse, reported turnout is biased across different socio-demographics. Previous studies show that respondents with a higher edu­ cational attainment and more interest in politics are more likely to overreport turnout(Enamorado and Imai, 2019). Other studies, using more complex data, concluded that socioeconomic inequality might play a more important role in electoral participation than previously thought(Lahtinen et al., 2019). Unfortunately, although collected, individual level turnout data is not publicly available in Romania. Similarly, no offi­ cial data can be accessed regarding voters’ individual level education, occupation or income. The publicly available official data report only the number of voters according to gender and age at polling station level. Turnout differences estimated on the basis of such official data are sometimes larger or in a different direction by comparison with similar estimates based on survey data. In this section, using official data, we compute and compare turnout across sex, age and emigration status. Figure 5 shows that regardless of the type of election, Roma­ nians from abroad are voting at a lower rate. Part of the explanation lies in the costs of voting, which are higher for emigrants. The number of polling stations abroad increased from only 111 in 1992 to 835 in 2019, falling back to 748 in 2020. The visibility of national elections, especially presiden­ tial ones, has also increased. Both these factors contributed to a significant increase in the turnout rate of Romanian emi­ grants from 2009 to 2019. While studies based on survey data suggest that men vote at a slightly higher rate than women(Kostelka et al., 2019), according to official data, in most democratic countries the gender gap either does not exist, or it shows women voting in a higher proportion than men(Cox and Morales Quiroga, 2022; Dahlgaard et al., 2019). The difference is explained by the fact that men tend to overreport voting more than women(Stockemer and Sundstrom, 2023). Studies also sug­ gest that sometimes women tend to vote less than men in second-order contests, such as the election for the European Parliament(Kostelka et al., 2019). Age plays a significant role in shaping people’s likelihood to vote in elections. Research indicates that turnout tends to be the highest among middle aged people, while younger or older people tend to exhibit lower voting turnout(Bhatti et al., 2012; Frank and Martínez i Coma, 2023). In the case of Romania, we have reached similar conclusions. While surveys report that the turnout rate is a little higher among men, our 7

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? Figure 5 Official voter turnout rate by migration status 70 64.1 60 58.0 54.4 53.2 50 47.7 49.9 41.8 40 39.2 39.8 32.0 30 36.6 26.2 20 17.0 10.2 10 7.3 2.5 4.6 3.9 1.2 0 2008 2012 2016 2020 2009 2014 2019 PARLIAMENTARY PRESIDENTIAL ROUND 1 6.1 2009 2014 2019 PRESIDENTIAL ROUND 2 Romania Abroad Note: Romanian turnout data comes from BEC(Central Electoral Bureau) and includes all voters(Romania and abroad). Turnout abroad is computed based on number of voters from abroad(BEC) and estimated number of Romanian emigrants(United Nations University World Institute for Develop­ ment Economics Research, WIID, https://doi.org/10.35188/UNU-WIDER/WIID-281123, and Eurostat data, DOI: 10.2908/migr_pop9ctz). Figure 6 Official voter turnout by gender, age and migration status – 2019 Presidential elections, round 2 80 70 60 50 40 30 20 10 0 18–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 50–64 65+ female RO female MIG male RO male MIG Note: Romanian emigrants’ data from Eurostat(DOI: 10.2908/migr_pop9ctz). Emigrants to non-European countries are not included, so the number is underestimated. Data on voters from abroad(any country) are from BEC(https://prezenta.roaep.ro/prezidentiale10112019/). Voters from Republic of Moldova are excluded, because many have Romanian citizenship. 8

Inequalities in voter turnout in Romania Figure 7 Official voter turnout by gender, age and migration status – 2020 Parliamentary elections 60 50 40 30 20 10 0 18–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 50–64 65+ female RO female MIG male RO male MIG Note: Romanian emigrants’ data from Eurostat(DOI: 10.2908/migr_pop9ctz). Emigrants to non-European countries are not included, so the num­ ber is underestimated. Data on voters from abroad(any country) are from BEC(https://prezenta.roaep.ro/prezidentiale10112019/). Voters from Re­ public of Moldova are excluded because many have Romanian citizenship. analyses of official data show that a voter turnout gender gap is still present at times in Romania, but not always in the same direction: – At lower ages(18–34), women vote in a higher propor­ tion than men in presidential elections(see Figure 6), while men vote in a slightly higher proportion in the case of parliamentary elections(see Figure 7). – Among the middle aged(35–54) women still vote more than men in presidential elections, but the gap is dimi­ nishing, while men still vote slightly more than women in parliamentary elections. – For seniors(55+), the gender gap is increasing, with men voting more than women in both presidential and parliamentary elections. The gender gap across almost all age categories is wider in the case of Romanians living abroad(emigrants) than among voters in Romania. While we do not have data to offer us additional insights into this interesting finding, it is probably linked to differences between men and women on other relevant socio-economic characteristics, such as education and occupation, as suggested by the high per­ centage of votes obtained in the last European Parliament elections by radical right parties among Romanian voters from abroad. Figure 8 presents the difference between turnout among men and turnout among women for ten age categories for the lat­ est Presidential elections(in 2019) and the latest parliamentary elections(in 2020). Positive values indicate that the percentage of men who voted was higher than the percentage of women. Negative values indicate that more women voted than men, in relative terms. The figure shows that in parliamentary elec­ tions turnout among men is higher for all age groups and it increases from a 0.4% difference among voters aged 25–29 to a 14% difference among voters aged 65 and over. In the presidential elections, by contrast, turnout among voters under 50 years of age is higher among women than among men, but among voters aged 55 and over this is reversed, as turnout among men becomes higher than among women. This is another interesting finding, probably linked to the particular characteristics of the two types of elections: par­ liamentary elections require from voters more interest in politics and more information, and we know from existing studies that in Romania men exhibit more political interest. In the case of presidential elections, it is easier for voters to choose among a few individual candidates than from several lists containing dozens of candidates. This might increase the turnout rate among women because it lowers the cultural barriers raised in the past by the notion that politics is for men, not for women. Needless to say, this gendered framing of politics and voting is a matter of enculturation. Women from the older generation in particular have been socialised 9

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? Figure 8 Official voter turnout rate gap by gender across age categories and elections 15 10 Turnout rate gap (male-female) 5 0 –5 18–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 50–64 65+ 2020 Parliamentary 2019 Presidential Round 1 2019 Presidential Round 2 not to get involved in such activities. The results pertaining to the younger generation show that this pattern of low voter turnout among women is slowly changing. effects of age on voting turnout we need to distinguish among three different ways in which time can influence our behaviour and attitudes. VOTERS, OCCASIONAL VOTERS AND NON-VOTERS Researchers with access to sufficient data have developed various typologies of voters and non-voters(see, for instance, Güllner, 2013 or Hagemeyer et al., 2023). Although we do not have access to such detailed datasets, we are able to use CSES(Comparative Study of Electoral Systems) data about voting turnout in the past two elections to distinguish between three types of citizens: non-voters, who did not vote in either election, occasional voters, who voted in only one round of elections, and voters, who voted in both elec­ tions. Based on this typology we can produce a profile of Romanian non-voters in order to better understand who they are and, if possible, to identify their reasons for not voting. NON-VOTERS’ CHARACTERISTICS: SOCIO-DEMOGRAPHICS Most turnout studies show that turnout increases from youth to middle age then, after a plateau, it slowly decreases (Deželan, 2023). It should be noted that when analysing the First, we have the effect of age: as people age, they go through different life stages, have different experiences and react accordingly. Second, we have generational effects: different socialisation experiences in a more affluent and secure environment make younger generations increasingly oriented toward post-materialist values and less interested in politics and conventional forms of political participation such as voting(Norris and Inglehart, 2019). Third, there are also period effects, which are significant political and socio-cul­ tural events that have a significant impact on voting turnout (Lisi et al., 2021). From this perspective, we can argue that the elections in 2020 took place at a confluence of young people who“par­ ticipate less in institutional politics than other age groups and also less than cohorts of young people decades ago” (Deželan, 2023), superimposed on the Covid-10 pandemic, which reduced voter turnout even more than in regular times. Based on CSES survey data(Figure 9), the largest share of non-voters and occasional voters is observed among young people, under 35 years of age. Alternatively, they could be referred to as Generation Y, those born after 1981. It should be noted that the largest share of non-voters and occasional 10

Inequalities in voter turnout in Romania Figure 9 Voters and non-voters by demographic features(predicted probabilities) 11 19 a few times a month Religios services attendance a few times a year never/ once a year 12 18 19 20 14 20 large town 17 17 Residence small/middle town 10 19 rural Married 12 yes 15 no 18 22 10 17 high Education 14 17 medium 14 21 low 14 28 gen Y(1981–1996) 15 20 gen X(1965–1980) 8 18 Generation boomers(1946–1964) 19 10 silent(1928–1945) 32 18 greatest(–1927) 20 18 18–24 14 25 25–34 10 18 35–44 Age 15 18 45–54 16 15 55–64 11 19 65+ Gender 13 21 female 13 16 male 70 70 62 66 66 71 71 63 73 70 65 58 65 74 71 50 63 61 72 68 69 70 66 71 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Non-Voters Occasional Voters Voters Note: Data are weighted by political weight; values are predicted probabilities(multiple logistic regression model). Data source: CSES Romania 11

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? voters(50% in total) is registered among the so-called “Greatest generation”(people born before 1927), but in their case the explanation for not voting probably has more to do with their health than with their interest in political participation. Using the same dataset, we found that the declared voter turnout is lower for unmarried people. They participate less because the social pressure to vote is lower in their case and because they are younger. This result is in line with findings from other countries(Frödin Gruneau, 2018). Generally, the higher a person’s socioeconomic status(SES), the higher the probability that they are a habitual voter. Because socioeconomic status is determined by educa­ tion, occupation and income, we observe higher turnout rates among voters with a higher education, higher income and better jobs. Survey data from the 2004 parliamentary and presidential elections show that Romanian non-voters are overrepresented among the less educated, with lower incomes(Com ș a and Bobîrsc, 2005). Exit-poll data from the 2009 presidential elections show that turnout rate increases with education(Com ș a, 2012). Similar data for other elec­ tions(2014 and 2019 presidential elections, 2016 parlia­ mentary elections) lead to the same conclusion: turnout rate increases with education(see Figure 10). According to our analysis of CSES data, we also found turnout differences across occupational categories: those with lower status occu­ pations are more likely to be non-voters; the others are more likely to be occasional voters or voters. It should be noted that the findings we see in the Roma­ nian context have also been observed in other democracies. Comparative electoral studies have shown that the com­ position of turnout in terms of people with different levels of resources in terms of education, occupation and income varies in accordance with the general level of turnout in a particular round of elections. When turnout is high, that generally means that more people from all categories are voting and thus the differences in turnout between rich and poor, or those with higher education and those with lower education tend to be smaller. When turnout is low it is generally because a higher proportion of people with fewer resources have stayed at home. We do not yet know whether economic inequality is a direct cause of turnout inequality (Schäfer and Schwander, 2019) or whether they reinforce each other(Schafer et al., 2022), but we do know that the two phenomena are connected. As a study of turnout in OECD countries showed, the turnout difference between people with low and people with high socioeconomic status has increased over time, mainly because of demobilisation among the former(Bos ă ncianu, 2021). Figure 10 Voter turnout rate by education 100 90 85.4 82.3 80 70 72.1 60 61.7 50 51.2 40 37.2 30 20 10 0 53.3 43.6 31.2 65.5 46.7 40.9 69.0 47.5 43.9 71.1 52.5 42.8 2014(round 1) 2014(round 2) low education 2016 2019(round 1) 2019(round 2) secondary school college / master / phd Data sources: IRES exit-poll data and population structure by education according to the Census(2011 and 2021). 12 mean

Inequalities in voter turnout in Romania NON-VOTER CHARACTERISTICS: POLITICAL VARIABLES Previous studies have shown that non-voters are not inter­ ested in politics(Blais and Daoust, 2020), do not have a strong party identification, have lower levels of political effi­ cacy(Denver and Johns, 2022), and are more disappointed by what they perceive as the way democracy is working in their country(Koch et al., 2023). Most of these findings are visible in the Romanian case as well – the largest turnout differences are associated with party identification(16%) and political efficacy(19%). Smaller differences(of about 5–6%) can be observed for political information and satisfaction with democracy(see Figure 11). REASONS FOR NOT VOTING If inequality of turnout has increased, as we have argued, it is important to attempt to understand why certain categories of citizens choose not to vote. A sense of civic duty has tra­ ditionally been considered to be one of the most important underpinnings of voting turnout(Blais and Daoust, 2020). One of the most important factors associated with turnout decline therefore is a generational weakening of the sense that it is a citizen’s duty to vote(Denver and Johns, 2022). With regard to Romania, a multivariate analysis of turnout in the 2009 Romanian presidential elections found that the main two factors contributing to non-voting are the habit of voting and a sense of the duty of voting(Com ș a, 2012). In a study of the 2004 Romanian parliamentary elections, which used an open-ended question, the main reasons for non-voting mentioned by respondents included: sickness, absence from the locality, a lack of interest in politics and elections, a lack of trust in politicians and parties, a percep­ tion that voting does not make any difference and a lack of time(Com ș a and Bobîrsc, 2005). An interview-based qual­ itative study of the 2020 Romanian parliamentary elections mentioned that the three main reasons for absenteeism were: low trust in parliament and politicians, the vagueness of political parties’ electoral promises, and a feeling that the result was locked in before the elections took place. Unexpectedly, the pandemic played only a marginal role in decisions not to vote, functioning mainly as a pretext (Gherghina et al., 2023). Figure 11 Voters and non-voters by political variables(predicted probabilities) Ideological identification Party identification Satisfaction with democracy Political efficacy Political information yes no yes no very/fairly satisfied not very satisfied not at all satisfied high medium low high medium low 12 19 14 19 5 16 16 21 14 16 11 17 20 18 8 18 13 18 22 22 10 10 16 20 19 19 69 67 79 63 70 69 65 74 69 56 70 71 66 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Non-Voters Occasional Voters Voters Data source: CSES Romania; data are weighted by political weight; values are predicted probabilities(multiple logistic regression model). 13

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? WHY BOTHER? THE CONSEQUENCES OF LOW AND UNEQUAL TURNOUT The literature generally agrees that lower turnout and turnout decline are detrimental to democracies. But it also points out that in some cases they can have a positive effect: for instance, declining turnout in one round of elec­ tions can lead to increasing party responsiveness to public opinion in the following period(Ezrow and Krause, 2023). To assess the consequences of low and unequal turnout three questions might be useful(Gallego, 2015):(i) Would the election results and policy outcomes be different if voter turnout were 100%?(ii) Are the political preferences of vot­ ers, occasional voters and non-voters similar?(iii) Do govern­ ments strategically target voters with public resources? In the absence of data, we can offer some suggestions from the literature in response to these questions. VOTE SHARES Sometimes low turnout has little to no impact on election results, sometimes the impact is greater. Some older stud­ ies argue that in western European societies the effect is small and inconsistent over time(Pettersen and Rose, 2007; Fisher, 2007). In Romania, this was a particularly important issue in the 2024 elections for the European Parliament. The government decided to modify the election calendar and have both European Parliament elections and local elections on the same day. The government justified its decision on the ground that in this way turnout would be higher than if the European Parliament elections were organised by themselves, and in turn that a high turnout would mean that right-wing populist parties(especially the Alliance for the Union of Romanians and SOS Romania) would not win a high share of the vote. Given that AUR and SOS Romania managed to obtain, together, about 20% of the votes, we are not convinced that the government’s plan was particu­ larly successful. But here we are more interested in the gov­ ernment’s rationale. not necessarily be a problem for occasional voters, it might affect non-voters differently and it has the potential, in the long term, to create categories of citizens who become dis­ satisfied with democracy. TARGET RESOURCES The tendency of political actors to resort to accusations of malfeasance(vote buying, voter intimidation or vote stealing) both before and after elections is a constant in Romania. Although the number of convictions for elec­ tion-related illegality is small enough to indicate that such accusations are groundless or electoral folklore, the gener­ alised use of such accusations may influence voter turnout. Previous studies have identified multiple ways in which turn­ out is affected by such behaviours. For example, economic intimidation is used as an instrument of mobilisation, espe­ cially in localities with few large employers(Mare ș , Muntean and Petrova, 2018). Different clientelist strategies have been developed in rural and urban settings to maximise results (Volintiru, 2012). Loyal voters are more likely to accept clien­ telism and interpret it as a reward for their long-term com­ mitment(Gherghina and Ț ap, 2022). On the other hand, some believe that information campaigns explaining the illegality of such behaviour may encourage voters to take a negative view of it(Mare ș and Visconti, 2020), and also that targeted public spending can increase electoral par­ ticipation in communities with low socio-economic status (Pop-Eleche ș and Pop-Eleche ș , 2012). OPINIONS, VALUES AND POLICIES The positions of voters and non-voters differ on some issues, but not by much(Gant and Lyons, 1993). New evidence suggests that in countries with pronounced disparities in voter turnout there are also notable divergences in terms of political values. For example, voters tend to be more con­ servative than non-voters on economic issues, they exhibit a lower propensity for redistribution and are more oriented towards liberal cultural values than non-voters. This implies that unequal participation goes hand in hand with nega­ tive outcomes pertaining to representation and democratic processes(Gallego, 2015). This expectation is sustained by a relatively recent meta-analysis: legal changes to the voting regime benefit right-wing parties(Terry, 2016). This issue is particularly worrying for the health of democratic regimes because it suggests that the interests of non-voters are less likely to be represented by elected officials. While this may 14

Increasing voter turnout by legislative measures INCREASING VOTER TURNOUT BY LEGISLATIVE MEASURES The low turnout issue is not new and most democracies have to deal with it, taking diverse approaches. Given these circumstances, one might ask whether legislative measures introduced to boost turnout actually work(given the costs). After identifying several legislative attempts to increase turnout, one paper argues that“attempts to reduce the costs of voting have been largely a waste of time when it comes to improving turnout”(Denver and Johns, 2022). In this section we focus on what we consider to be the most important legislative changes: improving regulations on vot­ ing abroad, postal voting, internet voting and lowering the voting age to 16. VOTING ABROAD: BETTER REGULATION AND ORGANISATION Most Romanians residing abroad choose to vote at a poll­ ing station. The number of voters abroad has increased over time, and sometimes it is quite large both as an absolute value and as a share of the electorate, especially in presi­ dential elections. In the 2019 presidential election, 675,000 people voted abroad in the first round, and 944,000 in the second, representing 7.8% and 10.4% of total votes, respec­ tively. Even though the number of polling stations had been increased, in some cases voters had to wait in line for hours, and some could not vote at all because of poor organisa­ tion and regulations. Solving these problems could increase turnout. There are many possible changes(Expert Forum, 2019; Guzun, Mogîldea and Pârvu, 2021): better targeting of Romanians abroad, along with an increase in the number of polling stations in areas with many voters; allocating more resources to specific polling stations in order to increase the number of personnel, voting ballots, voting booths and vot­ ing stamps; extending the duration of voting from one to two or more days; increasing the number of hours for voting (at present polling stations close at 9 pm); and extending postal voting to European parliamentary elections. At the same time, better regulations and organisation can only go so far in bringing out the vote. In the end, turnout will depend on many other factors. The 2024 European Par­ liament and local elections prove this point: although there were 915 polling stations abroad, the largest number in recent years, only 216,000 people showed up to vote, less than a quarter of the voters in the 2019 presidential elec­ tions. The low turnout among the diaspora was expected, however, given that they are second-order elections(people abroad could vote in the European Parliament elections, but not in local ones). VOTING FROM ABROAD: POSTAL VOTING Romanians residing abroad can vote by mail in both par­ liamentary elections(since 2015) and presidential elections (since 2019). Those choosing this method must register at least 45 days beforehand on the website www.votstrainatate.ro. Subsequently, voters receive by mail an envelope containing the ballot and other necessary materials. The voters send their completed ballot by mail, free of charge, to the electoral office for postal voting and can check online when the enve­ lope reaches its destination. The Permanent Electoral Author­ ity(AEP) informs voters whose envelopes have not arrived by the deadline so that they can still vote at a polling station if they so wish. The whole process is sometimes difficult; mailing the envelopes can be subject to errors and delays. Official data collected on postal voting show a significant difference between the number of envelopes dispatched by the electoral office and the number of envelopes(votes) received(see Figure 12). This difference is caused by the fact that the envelopes arrived too late, so the votes were nulli­ fied. Other causes mentioned by the AEP include envelopes/ votes annulled for various legal reasons, incorrect addresses and deficient postal services(loss of envelopes). Romanian emigrants mentioned as the main causes of low registration a lack of information about registration and practical diffi­ culties(Gherghina and Toma, 2016). The data show that not many Romanians residing abroad choose postal voting, but also that their number is increasing (see Figure 12). It should be noted that not all envelopes or votes sent by registered voters are received and accounted for. Nonetheless, there has been a noticeable improvement in this process, rising from 52% in 2016 to about 60–70% in 2019–2020. During the second round of the 2019 presiden­ tial elections, mass media–fuelled concerns about envelopes not reaching their destination in time led a considerable share of voters to change their minds and to vote at the 15

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? Figure 12 Postal voting in Romania 40,000 30,000 20,000 35,808 21,329 35,917 25,189 35,917 17,503 10,000 8,889 4,591 0 2016 Parliamentary Data Source: Central Electoral Buerau(BEC) 2020 Parliamentary envelopes 2019 Presidential Round 1 valid postal votes 2019 Presidential Round 2 polling stations instead, which explains the decline to 49%. Regardless of that, the 35,917 envelopes sent to the electoral bureau did not change the turnout rate by much(turnout increased by just 0.2 percentage points). VOTING ABROAD: INTERNET VOTING Internet voting is relatively new. In a recent study of voting in Geneva(Petitpas et al., 2021), the authors arrived at two main conclusions:(i) offering internet voting has increased turnout among abstainers and occasional voters, and(ii) the effects of the availability of internet voting on equality of participation are mixed with regard to age cohorts and gender. Internet voting has been used in Estonia since 2005. In sub­ sequent years, the share of internet voters in total voters has increased from 1.9% at the first election to 51% at the 2023 parliamentary elections(Piirmets, 2023). While this had a small positive impact on turnout at national level, it has significantly increased turnout among Estonians abroad. The number of elderly voters using the system has also increased over time. Other studies show that the youngest voters, the well educated, those on higher income and voters in urban areas were early adopters of internet voting, but also that after only three elections internet voting became widespread across different societal groups(Vassil et al., 2016), to the extent that socio-demographic characteristics no longer pre­ dict its usage(Ehin et al., 2022). On the other hand, internet voting did not increase turnout in places where voting was already easily accessible. Based on these findings, we can anticipate that the intro­ duction of internet voting in Romania could increase turnout among the young and(temporary) emigrants. This could be particularly important because, as we have seen in Figure 5, the highest turnout recorded to date among Romanian emigrants is only 37%, lower than the 50% turnout recorded in Romania itself for the second round of the 2019 presidential elections. To give an idea of the magnitude of the effect, assuming that four million Romanians are eligible to vote abroad, an increase in turnout among Romanian emigrants of 10 percentage points(from 37 to 47%) would mean about 400,000 more people voting(the population of an average Romanian county). Unfortunately, implementing internet voting is a complex process that requires significant protections, for example, in relation to voter identification and maintaining the safety and secrecy of the vote. Internet voting works well in Estonia, where it has been implemented as part of a comprehensive IT system that covers almost all interactions between citi­ zens and state. Implementing internet voting in Romania will require a combination of technical know-how, political will and cultural acceptance which might be difficult to achieve. LOWERING THE VOTING AGE The previous wave of decreasing the voting age, from 21 to 18, which ran from the 1970s to the 2000s, was not as suc­ cessful as had been hoped, as a lower proportion of peo­ ple aged 18–20 voted than those who were 21 when they first voted(Franklin, 2020). Even so, there are arguments 16

Increasing voter turnout by legislative measures in favour of reducing the voting age even further, to 16, in order to encourage young people to vote. According to the literature, because most young people aged 16–17 live with their parents, they could be influenced to acquire the habit of voting, leading to a larger turnout(Franklin, 2020). Alterna­ tively, extending the electorate to include more young people would incentivise political parties to address the issues that are important to that particular group, which would itself result in higher turnout among the young(Cammaerts et al., 2016). An additional advantage of extending the voting age is that it would establish habitual voting early in the socialisation process(Schäfer et al., 2020). Previous studies show that low­ ering the voting age has certainly not decreased voter turnout (Leininger and Faas, 2020), and a comparative study identified a significant turnout boost(5–6 percentage points) over the 20 years following the lowering of the voting age, suggesting that the effect is long term(Franklin, 2020). Opponents of lowering the voting age argue that 16 and 17 year-olds lack the maturity required to vote and some studies support this argument(Maheó and Bélanger, 2020; Zeglo­ vits, 2013). Other studies, however, suggest otherwise(Oos­ terhoff et al., 2022). Among other things, their voter turnout is higher than among older people voting for the first time (Aichholzer and Kritzinger, 2020); they engage with electoral campaigns in a similar way to other voters and demonstrate notably high levels of external confidence and satisfaction with democracy. Even more important, there is evidence that lowering the voting age can positively impact educa­ tional outcomes by enhancing political engagement, learning about politics, influencing career aspirations and fostering mature attitudes(Breeze et al., 2023) and by increasing the likelihood of discussing politics and using voting advice appli­ cations. In short,‘the right to vote changes behaviour’ in a positive direction(Leininger et al., 2024). The issue of protest voting represents an additional concern in the context of this discussion, as this type of voting is asso­ ciated with young citizens with medium-level education sup­ porting marginal parties and candidates, sometimes populist or extremist ones. As a direct consequence of this type of voting, the share of seats won by such parties and candidates has been increasing, bringing them into parliament. Roma­ nia has experienced several situations of this type: the 2000 national elections(Vadim Tudor and PRM – Great Romanian Party), the 2012 parliamentary elections(PP-DD – The Peo­ ple’s Party – Dan Diaconescu), and the 2020 parliamentary elections(AUR – The Alliance for the Union of Romanians). Such cases suggest that lowering the voting age should be implemented only after significant structural economic and political issues that might push young voters towards extrem­ ist political parties have been properly addressed. Such devel­ opment requires changes in school curricula, civic education classes, awareness-raising campaigns or projects and mock elections in schools, as has been done in Austria(Zeglovits, 2013), or as impact evaluation research is proposing(Ribeiro et al., 2023). These preparatory steps are even more impor­ tant given that“the turnout gap among the young is due to rising‘start-up’ costs of voting, which affect mainly those who are resource poor”(Schäfer et al., 2020). DIFFICULT CHANGES In addition to legislative changes, which have their own char­ acteristics, turnout could also be increased by what we call “difficult changes”. They are difficult because they require political actors to go against three decades of socialisation in a system that attaches the highest value to maximising the number of votes that can be obtained in the next round of elections. Every scholar who studies voting in Romania has heard at least once, in interviews, focus groups or question­ naires that“all politicians are the same” and that“we need to vote for the lesser evil”. Both arguments have often been brought up in discussion when people were asked to explain why turnout is low in Romania. Romania represents an interesting case, in which most of the important political parties are fighting to win over what they consider to be the median voter, thus ignoring a significant part of the population which as a consequence is almost com­ pletely unrepresented on the political arena. Most Romanian political parties locate themselves as centre-right or right on economic issues and to the right on social issues. People who have more liberal positions on social issues or who prefer more regulation of the economy and the market are hard pressed to find a political party that might be willing to repre­ sent their interests. From this perspective, it is natural to have a low turnout in Romanian elections because the existing political parties mainly ignore such issues while also protect­ ing themselves from the emergence of significant competitors with an agenda that currently unrepresented citizens might support. The few attempts made to date to create a viable political party to address the interests of the unrepresented citizens on the left(USR, Demos) have failed. An additional reason that might explain some voters’ reluc­ tance to go to the polls on election day is related to their perception that political parties are selecting their leaders and their candidates based only on the interests of the parties and politicians, ignoring meritocratic concerns and or any need to match candidates’ abilities to the requirements of the posi­ tions for which they are running. In short, there is a general perception of so-called“negative selection” among political parties, which has led to a significant decline in the quality of top level politicians. Witnessing this type of behaviour for more than three dec­ ades has disillusioned a significant part of the population in relation to voting, elections and democracy and has reduced their trust in political institutions. The solution is relatively simple, but also poses one of the most difficult challenges facing political parties in Romania: parties need to change their organisational culture and their strategy to move closer to the citizens and to represent voters’ interests. As long as they insist on politics as usual, they will continue to lose the trust of ordinary Romanians(currently a mere one in ten pro­ fess such trust) and face high abstention rates. 17

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? CONCLUSIONS AND RECOMMENDATIONS The data presented in this report show that voter turnout in Romania has a number of characteristics in common with other countries in the region, but also some characteristics of its own. First, as in many other post-communist countries in central and eastern Europe, Romania has seen a decline in voter turnout from the 1990 founding elections to the present. What sets Romania apart from the other countries in the region is the magnitude of this decline. Second, we have shown that emigration is a key element in explaining the fall in voter turnout. A significant part of the Romanian population is working and living(for shorter or longer periods of time) in other EU countries or elsewhere in the world. Given the higher costs of voting for Romanian migrants, they are less likely to vote, which means that an decrease in voter turnout is correlated with an increase in the Romanian diaspora. Third, inequality of voter turnout in Romania seems to be similar to that in other countries in the Unequal Democracies project. Gender-related inequalities are minimal, while ageand education- related inequalities are larger but similar to what was observed in the other countries: people with a higher socioeconomic status(better education, better job, higher income) are more likely to vote. In terms of age, turn­ out increases until about the age of 50, plateaus for about a decade, and starts to decline after the age of 60. Addition­ ally, people living in rural areas have a higher turnout than people living in large cities. This is usually explained by the efforts of local mayors to get out the vote in favour of their own political parties. Finally, frequent churchgoers also have a slightly higher turnout rate. This may be explained by peer pressure but there is also anecdotal evidence of priests acting as electoral agents in election campaigns. Fourth, survey data show that non-voters in Romania have a political values profile that suggests that not voting is rather a personal decision and not the result of electoral or institutional barriers. Non-voters are less informed about politics, have low levels of political efficacy, are less likely to have a party identification or attachment, and are more likely to be dissatisfied with Romanian democracy. Other factors that increase the probability of not voting are related to the quality of political parties in Romania, which offer vague electoral promises, do not make convincing attempts to differentiate themselves from their competitors and, as a results, are among the least trusted political institutions in Romania. Because one function of elections is to establish and main­ tain the legitimacy of the political system, underpinning successive electoral cycles, low voter turnout is a particular problem that needs to be solved. There are a variety of ways of approaching this, but we preferred recommendations that we believe address critical points, namely low voter turnout among the Romanian diaspora and among young people. We also favoured solutions that that should not be very dif­ ficult to implement. When it comes to improving voter turnout among the Roma­ nian diaspora, the first step is to improve the relevant regula­ tions and better organise the process of voting abroad. Given that online voting for Romanian emigrants is not possible in the short or medium term, postal voting and casting a ballot at polling stations need to be improved. Postal voting needs to be streamlined and simplified, made available for a longer period in order to overcome possible postal delays and, most importantly, possible voters need to be informed about the advantages of using this method. Currently, many Romanians abroad are not aware that postal voting is possible, or what they need to do. Voting at polling stations can be improved by opening up more locations on election day, in particular where there are concentrations of Romanian expats. At the same time, Romanian voters need to be taught to be more proactive in demanding new polling stations where they live (there is a procedure for this, but it is not very well known). Our second key recommendation is aimed at lowering the voting age to 16 years. One option is to lower the voting age for the European Parliament elections first, following the examples of Austria, Germany and Malta. These countries’ experiences could be used to demonstrate the advantages of lowering the voting age. Linking this issue to the European Parliament elections has the advantage that the EU could be used as an instrument and/or partner to promote the idea. Moreover, handling it as a supranational issue could help in the development of a pan-European movement in support of low­ ering the voting age across the EU. A second option is to lower the voting age for local elections. As we have seen, some have criticised this idea on the grounds that 16 and 17 year olds are not ready for the complexities of voting. But while this has some plausibility in the case of parliamentary or presidential elections, the argument is less convincing with regard to local elections, in which awareness of the local context suffices. 18

References REFERENCES Aichholzer, J. and Kritzinger, S.(2020): Voting at 16 in Practice: A Review of the Austrian Case, in: Eichhorn, J. and Bergh, J.(eds): Lowering the Voting Age to 16. Palgrave Studies in Young People and Politics. Palgrave Macmillan; available at: https://doi.org/ 10.1007/978-3-030-32541-1_5. Bhatti, Y., Hansen, K.M. and Wass, H.(2012): The relationship be­ tween age and turnout: A roller- coaster ride, in: Electoral Studies, 31: 588–593; available at: https://doi.org/10.1016/j.electstud.2012.05.007. Blais, A. and Daoust, J.-F.(2020): The Motivation to Vote. Explaining Electoral Participation. UBC Press.. Bo să ncianu, C.M.(2021): Determinants of the Socio-Economic Turnout Gap in OECD Countries; available at: https://doi.org/10.31219/osf.io/ v2m46 . Deželan, T.(2023): Young people’s participation in European democratic processes. How to improve and facilitate youth involvement. European Parliament: Directorate-General for Internal Policies of the Union; availa­ ble at: https://doi.org/10.2861/945934. Ehin, P., Solvak, M., Willemson, J. and Vinkel, P.(2022): Internet voting in Estonia 2005–2019: Evidence from eleven elections, in: Government Information Quarterly, 39(4): 101718; available at: https://doi.org/10.1016/J.GIQ.2022.101718. Elsässer, L., Schäfer, A. and Wenker, J.(2022): Ungleiche Demokratien: Wer sitzt(nicht) im Parlament? Die soziale Zusammensetzung der Parla­ mente in fünf OSZE-Ländern[Unequal democracies: Who does(not) vote. Voter turnout trends in the OSCE region since 1970]. Friedrich-Ebert-Stiftung. Breeze, M., Gorringe, H., Jamieson, L. and Rosie, M.(2023): Educational outcomes of political participation? Young first-time voters three years after the Scottish Independence Referendum, in: Journal of Youth Studies, 26(1): 61–79; available at: https://doi.org/10.1080/136762 61.2021.1980517 . Enamorado, T. and Imai, K.(2019): Validating Self-Reported Turnout by Linking Public Opinion Surveys with Administrative Records, in: Public Opinion Quarterly, 83(4): 723–748; available at: https://doi.org/10.1093/ poq/nfz051. Brenner, P.S.(2021): E ff ects of Nonresponse, Measurement, and Coverage Bias in Survey Estimates of Voting, in: Social Science Quarterly, 102(2): 939–954; available at: https://doi.org/10.1111/SSQU.12935. Cammaerts, B., Bruter, M., Banaji, S., Harrison, S. and Anstead, N. (2016): Youth Participation in Democratic Life. Palgrave Macmillan; available at: https://doi.org/10.1057/9781137540218. Com ș a, M.(2012): Participarea la vot. De ce(nu) voteaz ă oamenii?[Voter turnout. Why do people(not) vote], in: Com ş a, M., Gheorghi ţă , A. and Tufi ş , C.(eds): Alegerile preziden ț iale din România, 2009[2009 Romanian Presidential Election]. Polirom: 149–200. Com ș a, M.(2015): Turnout decline in Romanian national elections: Is it that big?, in: Studia Universitatis Babes-Bolyai Sociologia, 60(2); available at: https://doi.org/10.1515/subbs-2015-0010. Com ș a, M.(2017): Explaining turnout decline in post-communist countries: the impact of migration, in: Studia Universitatis Babes-Bolyai Sociologia, 62(2); available at: https://doi.org/10.1515/subbs-2017-0010. Com ș a, M. and Bobîrsc, D.(2005): Votan ț i versus non-votan ț i. Cine ș i de ce?[Voters versus non-voters. Who and why?], in: Rotariu, T. and Com ș a, M.(eds): Alegerile generale 2004. O persp ectivă sociologică [The 2004 National Elections. A sociological view]. Eikon: 73–98. Expert Forum(2019): Policy Brief 74; available at: https://expertforum. ro/wp-content/uploads/2019/06/Solutii- vot-strainatate-EFOR.pdf . Ezrow, L. and Krause, W.(2023): Voter Turnout Decline and Party Responsiveness, in: British Journal of Political Science, 53(1): 85–103; available at: https://doi.org/10.1017/S0007123421000673. Fisher, S.D.(2007):(Change in) turnout and(change in) the left share of the vote, in: Electoral Studies, 26(3): 598–611; available at: https://doi. org/10.1016/J.ELECTSTUD.2006.10.006. Frank, R.W. and Martínez i Coma, F.(2023): Correlates of Voter Turnout, in: Political Behavior, 45: 607– 633; available at: https://doi. org/10.1007/s11109-021-09720-y . Franklin, M.N.(2020): Consequences of Lowering the Voting Age to 16: Lessons from Comparative Research, in: Lowering the Voting Age to 16. Springer International Publishing: 13–41; available at: https://doi.org/10. 1007/978-3-030-32541-1_2 . Frödin Gruneau, M.(2018): Reconsidering the partner e ff ect on voting, in: Electoral Studies, 53: 48–56; available at: https://doi.org/10.1016/J. ELECTSTUD.2018.03.003. Com ș a, M. and Postelnicu, C.(2013): Measuring Social Desirability E ff ects on Self-Reported Turnout Using the Item Count Technique, in: International Journal of Public Opinion Research, 25(2): 153–172; available at: https://doi.org/10.1093/ijpor/eds019. Cox, P. and Morales Quiroga, M.(2022): Gender Gaps in Electoral Turnout: Surveys versus Administrative Censuses, in: Political Studies Review, 20(2): 304–313; https://doi.org/10.1177/14789299211019562. Dahl, R.(1971): Polyarchy. New Haven, CT: Yale University Press. Dahlgaard, J.O., Hansen, J.H., Hansen, K.M. and Bhatti, Y.(2019): Bias in Self-reported Voting and How it Distorts Turnout Models: Disentangling Nonresponse Bias and Overreporting Among Danish Voters, in: Political Analysis, 27(4): 590–598; available at: https://doi.org/10.1017/PAN.2019.9. DeBell, M., Krosnick, J.A., Gera, K., Yeager, D.S. and McDonald, M.P.(2020): The Turnout Gap in Surveys: Explanations and Solutions, in: Sociological Methods and Research, 49(4): 1133–1162; available at: https://doi.org/10.1177/0049124118769085. Denver, D. and Johns, R.(2022): Turnout: Why People Vote(or Don’t). Elections and Voters in Britain, 29–59. https://doi.org/10.1007/978-3-03086492-7_2 . Denver, D. and Johns, R.(2022): Turnout: Why People Vote(or Don’t), in: Elections and Voters in Britain. Palgrave Macmillan; available at: https://doi.org/10.1007/978-3-030-86492-7_2. Gallego, A.(2015): Unequal political participation worldwide. Cambridge University Press. Gant, M.M. and Lyons, W.(1993): Democratic Theory, Nonvoting, and Public Policy, in: American Politics Quarterly, 21(2): 185–204; available at: https://doi.org/10.1177/1532673X9302100202. Gherghina, S., Mi ș coiu, S. and Tap, P.(2023): Using the Pandemic as a Pretext. Voter Absenteeism in the 2020 Elections in Romania, in: Communist and Post-Communist Studies, 56(4): 81–99; available at: https://doi.org/10.1525/CPCS.2023.1823167. Gherghina, S. and Toma, I.(2016): Lipsa de informa ț ie ș i procedurile dificile. Înregistrarea redus ă pentru votul prin coresponden ță în România [Lack of information and di ffi cult procedures. Reduced registration for postal voting in Romania], in: Sfera Politicii, 189–190; available at: https://revistasferapoliticii.ro/sfera/189-190/art02-Gherghina-Toma.php. Gherghina, S. and Tap, P.(2022): Buying loyalty: volatile voters and electoral clientelism, in: Politics, 1–18; available at: DOI:10.1177/02633957221132707. Guzun, V., Mogîldea, M. and Pârvu, S.(2021): Metode alternative de exprimare a votului: Lecții învățate pentru Republica Moldova din experiența României și Estoniei . Institutul pentru Politici ș i Reforme Europene. 19

UNEQUAL DEMOCRACIES – WHO DOES(NOT) VOTE IN ROMANIA? Güllner, M.(2013): Nichtwähler in Deutschland[Non-voters in Germany]. Friedrich-Ebert-Stiftung. Hagemeyer, L., Faltas, C. and Faus, J.(2023): Gelegenheits­ wähler*innen auf der Spur. Motive der Nichtwahl und Wege zur Stärkung der Wahlbeteiligung[On the trail of occasional voters. Motives for not voting and ways to increase voter turnout]. Friedrich-Ebert-Stiftung. Hooghe, M. and Kern, A.(2017): The tipping point between stability and decline: trends in voter turnout, 1950–1980-2012, in: European Political Science, 16(4): 535–552; available at: https://doi.org/10.1057/ s41304-016-0021-7 . International IDEA(2023): The Global State of Democracy 2023: The New Checks and Balances; available at: https://doi.org/10.31752/ idea.2023.78 . Koch, C.M., Meléndez, C. and Rovira Kaltwasser, C.(2023): Mainstream Voters, Non-Voters and Populist Voters: What Sets Them Apart?, in: Political Studies, 71(3): 893–913; available at: https://doi. org/10.1177/00323217211049298 . Kostelka, F.(2017): Does democratic consolidation lead to a decline in voter turnout? Global evidence since 1939, in: American Political Science Review, 111(4): 653–667; available at: https://doi.org/10.1017/S0003 055417000259 . Kostelka, F. and Blais, A.(2021): The generational and institutional sources of the global decline in Voter Turnout, in: World Politics, 73(4): 629–667; available at: https://doi.org/10.1017/S0043887121000149. Kostelka, F., Blais, A. and Gidengil, E.(2019): Has the gender gap in voter turnout really disappeared?, in: West European Politics, 42(3): 437–463; available at: https://doi.org/10.1080/01402382.2018.1504486. Lahtinen, H., Martikainen, P., Mattila, M., Wass, H. and Rapeli, L. (2019): Do Surveys Overestimate or Underestimate Socioeconomic Di ff erences in Voter Turnout? Evidence from Administrative Registers, in: Public Opinion Quarterly, 83(2): 363–385; available at: https://doi.org/ 10.1093/poq/nfz022 . Leininger, A. and Faas, T.(2020): Votes at 16 in Germany: Examining Subnational Variation, in: Eichhorn, J. and Bergh, J.(eds): Lowering the Voting Age to 16. Palgrave Studies in Young People and Politics. Palgrave Macmillan; available at: https://doi.org/10.1007/978-3-030-32541-1_8. Leininger, A., Schäfer, A., Faas, T. and Roßteutscher, S.(2024): Coming of voting age. Evidence from a natural experiment on the e ff ects of electoral eligibility, in: Electoral Studies, 88: 102751; available at: https://doi.org/10.1016/J.ELECTSTUD.2024.102751. Lisi, M., Quaranta, M., Real-Dato, J. and Tsatsanis, E.(2021): Assessing the Impact of Age, Cohort and Period E ff ects on Partisanship and Support for Mainstream Parties: Evidence from Southern Europe, in: South European Society and Politics, 26(2): 239–270; available at: https://doi.org/10.1080/13608746.2021.1965322. Maheó, V.A. and Bélanger, É.(2020): Lowering the Voting Age to 16? A Comparative Study on the Political Competence and Engagement of Underage and Adult Youth, in: Canadian Journal of Political Science/ Revue Canadienne de Science Politique, 53(3): 596–617; available at: https://doi.org/10.1017/S0008423920000232. Mar eș , I., Muntean, A. and Petrova, T.(2018): Economic Intimidation in Contemporary Elections: Evidence from Romania and Bulgaria, in: Government and Opposition, 53(3): 486-517; available at: doi:10.1017/ gov.2016.39. Mare ș , I. and Visconti, G.(2020): Voting for the lesser evil: evidence from a conjoint experiment in Romania, in: Political Science Research& Methods, 8: 315–328. McAllister, I. and Quinlan, S.(2022): Vote overreporting in national election surveys: a 55-nation exploratory study, in: Acta Politica, 57(3): 529–547; available at: https://doi.org/10.1057/S41269-021-00207- 6/ FIGURES/3 . Norris, P. and Inglehart, R.(2019): Cultural Backlash: Trump, Brexit, and Authoritarian Populism. Cambridge University Press; available at: https:// doi.org/DOI: 10.1017/9781108595841 . Oosterh off , B., Wray-Lake, L. and Hart, D.(2022): Reconsidering the Minimum Voting Age in the United States, in: Perspectives on Psychologi­ cal Science, 17(2): 442–451; available at: https://doi.org/10.1177/174569 1621994221 . Petitpas, A., Jaquet, J. M. and Sciarini, P.(2021): Does E-Voting matter for turnout, and to whom?, in: Electoral Studies, 71: 102245; available at: https://doi.org/10.1016/J.ELECTSTUD.2020.102245. Petr ičević , V. and Stockemer, D.(2020): Why Do Citizens Not Turn Out? The E ff ect of Election-Specific Knowledge on Turnout in European Elections in Eastern Europe, in: East European Politics and Societies, 34(3): 591–610; available at: https://doi.org/10.1177/0888325419870228. Pettersen, P.A. and Rose, L.E.(2007): The dog that didn’t bark: Would increased electoral turnout make a di ff erence?, in: Electoral Studies, 26(3): 574–588; available at: https://doi.org/10.1016/J.ELECTSTUD.2006.10.004. Piirmets, E.(2023): How did Estonia carry out the world’s first mostly online national elections; available at: https://e-estonia.com/how-didestonia-carry-out-the-worlds-first-mostly-online-national- elections/ . Pop-Eleche ș, C. and Pop-Eleche ș , G.(2012): Targeted Government Spending and Political Preferences, in: Quarterly Journal of Political Science, 7(3): 285–320; available at: http://dx.doi.org/10.1561/100.00011017. Ribeiro, N., Malafaia, C. and Ferreira, T.(2023): Lowering the voting age to 16: Young people making a case for political education in fostering voting competencies, in: Education, Citizenship and Social Justice, 18(3): 327–343; available at: https://doi.org/10.1177/17461979221097072. Schäfer, A., Roßteutscher, S. and Abendschön, S.(2020): Rising start-up costs of voting: political inequality among first-time voters, in: West European Politics, 43(4): 819–844; available at: https://doi.org/10. 1080/01402382.2019.1638141 . Schäfer, A. and Schwander, H.(2019):‘Don’t play if you can’t win’: does economic inequality undermine political equality?, in: European Political Science Review, 11(3): 395–413; available at: https://doi.org/10.1017/ S1755773919000201 . Schafer, J., Cantoni, E., Bellettini, G. and Berti Ceroni, C.(2022): Making Unequal Democracy Work? The E ff ects of Income on Voter Turnout in Northern Italy, in: American Journal of Political Science, 66(3): 745–761; available at: https://doi.org/10.1111/AJPS.12605. Stockemer, D.(2017): Electoral participation: How to measure voter turnout?, in: Social Indicators Research, 133(3): 943–962. Stockemer, D. and Sundstrom, A.(2023): The gender gap in voter turnout: An artefact of men’s over-reporting in survey research?, in: British Journal of Politics and International Relations, 25(1): 21–41. https://doi. org/10.1177/13691481211056850 . Szulecki, K., Bertelli, D., Erdal, M.B., C oşc iug, A., Kussy, A., Mikiewicz, G. and Tulbure, C.(2021): To vote or not to vote? Migrant electoral(dis)engagement in an enlarged Europe, in: Migration Studies, 9(3): 989–1010. https://doi.org/10.1093/migration/mnab025. Terry, W.C.(2016): Yes, structurally low turnout favors the right, in: Politics, Groups, and Identities, 4(3): 444–461; available at: https://doi.org/ 10.1080/21565503.2015.1124789 . V-Dem(2024): Democracy Report 2024: Democracy Winning and Losing at the Ballot. University of Gothenburg: V-Dem Institute. Vassil, K., Solvak, M., Vinkel, P., Trechsel, A.H. and Alvarez, R.M. (2016): The di ff usion of internet voting. Usage patterns of internet voting in Estonia between 2005 and 2015, in: Government Information Quarterly, 33(3): 453–459; available at: https://doi.org/10.1016/J.GIQ.2016.06.007. Volintiru, C.(2012): Clientelism: Electoral Forms and Functions in the Romanian Case Study, in: Romanian Journal of Political Sciences, 1: 35–66. Wenker, J.(2024): Who(does) not vote? Friedrich-Ebert-Stiftung – Democracy of the Future. Zeglovits, E.(2013): Voting at 16? Youth su ff rage is up for debate, in: European View, 12(2): 249–254; available at: https://doi.org/10.1007/ S12290-013-0273-3. 20

Imprint ABOUT THE AUTHORS IMPRINT Mircea Com ş a is Professor at Babe ș -Bolyai University, Department of Sociology. He is or was involved in several research teams: Public Opinion Barometer(BOP), Romanian Electoral Studies(RES), World Values Survey(WVS), Euro­ pean values Study(EVS), and Comparative Study of Electoral Systems(CSES). Claudiu Tufi ș is Associate Professor of Political Science at the University of Bucharest. His research focuses on politi­ cal culture, political behavior, electoral studies, social move­ ments, and civil society. He is a founding member of the Romanian Quantitative Studies Association and the editor of the Romanian Journal of Political Science. Publisher: FES Regional Office for International Cooperation Democracy of the Future Reichsratsstr. 13/5 A-1010 Vienna Responsibility for content: Johanna Lutz| Director, Democracy of the Future Phone:+43 1 890 3811 301 X:@FES_Democracy democracy.fes.de Contact / Orders: democracy.vienna@fes.de Design: pertext, Berlin| www.pertext.de The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung(FES) or of the organ­ ization for which the author works. Commercial use of media published by the FES is not permitted without the written consent of the FES. Publications by the FES may not be used for electioneering purposes. ISBN 978-3-98628-625-5 © 2024 ABOUT UNEQUAL DEMOCRACIES Unequal Democracies is a project by FES Democracy of the Future. The main goal is to promote comparative under­ standing of why inequality in voting, political representation and other democratic processes hurt our democracies. In the series Who does(not) have a seat in Parliament? we analyse the social representation of European parliaments. In the series Who does(not) vote? we investigate election turnout levels across the parameters gender, age, social class and education in European democracies. Both series contain comparative studies and selective coun­ try reports. The comparative studies lay out general trends while the country reports provide country-specific analy­ ses about the state of particular national contexts with the aim to develop and discuss political recommendations for decision-makers. More information at: https://democracy.fes.de/topics/inequality-democracy 21

WHO DOES(NOT) VOTE IN ROMANIA? Why should the Romanians Who are the Romanian care about turnout? non-­voters? What should be done? This report examines, comparatively, the evolution of voter turnout in Roma­ nia since the 1990 elections, focusing on the significant decline in electoral participation and on the most inpor­ tant inequalities identified in data from official sources, public opinion surveys, and exit-polls. Although most post-communist coun­ tries have registered voter turnout decline since the founding elections, the decline was the highest in Roma­ nia. This can be explained, among oth­ ers, by high emigration and low levels of interest in politics, linked to low lev­ els of trust in political parties. The report suggests different solu­ tions to increase voter turnout: remove institutional barriers and improve pro­ cedures for diaspora voters, convince political parties to come up with a bet­ ter offer for voters, and decrease the voting age to 16 for the European Par­ liament elections, at least.