Compulsory Voting and Voter Turnout in the Low Countries

• 1 Introduction

  In recent decades research on voter turnout has vastly increased, especially because of the structural trend towards lower turnout in western societies. Some scholars consider the introduction of a system of compulsory voting as an effective tool to counter this trend. The academic debate on voter turnout and compulsory voting, however, points to a trade-off between equality of electoral participation, on the one hand, and the quality of the vote, on the other. The first argument considers a high turnout at elections as a condition for the effective functioning of representative democracy. Higher turnout levels are assumed to lead to more equality with regard to electoral participation (Birch, 2008; Lijphart, 1997). However, another strand in the literature argues that compulsory voting negatively impacts the quality of the vote since compulsory rules force the uninformed and uninterested as well. It is feared that they will cast blank, invalid or unsubstantiated votes (Rosema, 2007; Selb & Lachat, 2009).
  The debate on compulsory voting and voter turnout is relevant for several academic debates as it touches on politically relevant phenomena such as the representation of social groups (Guntermann et al., 2020), the effectiveness of voter turnout campaigning (van Ostaaijen et al., 2019), voting behaviour (Wauters & Devroe, 2017), government spending (Hoffman et al., 2017) and parties’ vote-seeking strategies (Singh, 2019).
  Thus, this article aims to make a state of affairs of the existing literature on compulsory voting and voter turnout in Belgium and the Netherlands¹x The original aim of the article was to include data and literature on Belgium, the Netherlands and the Grand Duchy of Luxembourg. Our search on the official website of the elections in the Grand Duchy of Luxembourg (elections.public.lu) revealed, however, that election results (including absenteeism rates) are not available over time, either for parliamentary or for municipal elections. In addition, our search for English-language literature on compulsory voting and turnout in the Grand Duchy of Luxembourg yielded hardly any relevant contributions. We therefore decided not to include the Grand Duchy of Luxembourg in our analysis. and its impact on both equal electoral participation and the quality of the vote. For this purpose, we collected and analysed previously published research on (1) the (simulated) consequences of abolishing compulsory voting in either of both or both countries and (2) (intended) voter turnout. We executed both for Belgium and the Netherlands a search for ‘compulsory voting’, ‘voter turnout’ and ‘electoral participation’ in the Web of Science database. In addition, we included relevant Dutch articles and book chapters building on election surveys from ISPO/PIOP, PartiRep and SCV in our overview for Belgium. For the Netherlands, we included relevant contributions from Sociaal en Cultureel Planbureau²x www.scp.nl. and Stichting Kiezersonderzoek Nederland.³x www.dpes.nl. On the basis of the snowballing method we identified some standard works on the topic of voter turnout in Belgium and the Netherlands. Records were eligible when they discussed determinants explaining individuals’ willingness to vote, individual-level turnout or the quality of the vote during local, regional and national elections in Belgium and/or the Netherlands.
  Following Smet and Van Ham (2013), our analysis is limited to individual-level studies for the following three reasons. First, personally held characteristics matter when it comes to voting. Singh (2011: 108), for example, concludes that “individual-level factors hold considerable weight in relation to turnout”, and we agree with him that only micro analyses fully capture the decision calculus individuals undertake before deciding whether or not to vote. Secondly, given the system of compulsory voting in Belgium, studies on voter turnout measure respondents’ willingness to vote under the condition that voting would no longer be compulsory. By focusing on individual-level studies, we are able to include these simulation studies within our review. Thirdly, multiple extensive literature reviews and meta-analyses concerning voter turnout examined the literature on aggregate-level variables on voter turnout recently (see i.e. Cancela & Geys, 2016; Geys, 2006; Kouba et al., 2020). Therefore, we consider the added value of another review of aggregate-level studies to be limited.
  We have limited the period of our review to studies published between 1970 and 2021. We chose 1970 as the starting date for the literature review since the first Dutch elections without compulsory voting were organised in this year.
  We divide this contribution into three sections. First, we briefly discuss the historical pathways and actual meaning of compulsory voting in Belgium and the Netherlands. Second, we focus on the impact of compulsory voting on equal participation. Third, we look into the correlation between compulsory voting and multiple indicators of the quality of the vote. Finally, we offer some conclusions and suggest avenues for future research.

• 2 An Historical Overview of Compulsory Voting in the Low Countries

  In 1893, Belgium adopted a system of universal male suffrage and plural voting under pressure from the socialist movement. As such, Belgium was the first country in the world to introduce compulsory voting. The introduction of this political duty was a concession to conservative (catholic) politicians fearing (leftist) radicalisation among the electorate due to its growth (Caluwaerts et al., 2022).
  Over the years, compulsory voting was regularly questioned, first by nationalist and green politicians, from the 1990s onwards – nearly one hundred years after its introduction – and also by liberal ones. However, the topic seldom dominated the political agenda. Because the obligation to vote for the general election is included in the Constitution, it can only be amended by broad majorities. Given the differences in the attitudes of political parties on this issue, these majorities are not immediately realistic.
  The situation concerning the local elections is different. The Lambermont agreements of 2001 transferred the responsibility for local government legislation from the Belgian federal level to the regions. Since January 2002 the Flemish, Walloon and Brussels regions have been granted the power to settle the structure of local government, including the local electoral system (Ackaert, 2005). A first attempt to abolish compulsory voting in Flanders at the local level in 2003 found no majority among the regional MP’s since this attempt was the initiative of a single MP without the support of the parties belonging to the government coalition.⁴x VOORSTEL VAN DECREET – van de heer André-Emiel Bogaert – houdende wijziging van de wet van 19 oktober 1921 tot regeling van de provincieraadsverkiezingen, met het oog op de afschaffing van de stemplicht.. About 15 years later, in 2019, the abolition of compulsory voting for local elections was included in the coalition agreement that led to the formation of the Flemish government, as part of a broader package of political-electoral measures. On 14 July 2021, the Flemish Parliament gave the green light for the abolition of compulsory voting in local elections in the Flemish Region of Belgium, which will apply from the next elections in 2024 onwards. A majority supported by liberals, Christian-democrats, nationalists and the far-right voted in favour of the abolition.⁵x ONTWERP VAN DECREET tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie. The regional minister of domestic affairs argued that compulsory voting is a flawed tool to strengthen local democracy. Under a system of voluntary voting, he argued, political parties would be obliged to explain the value and importance of participating in local elections. He also referred to the Netherlands, where compulsory voting was abandoned in 1970 for elections at all policy levels.⁶x 790 (2020-2021) – Nr. 7 23 juni 2021 (2020-2021) Verslag namens de Commissie voor Binnenlands Bestuur, Gelijke Kansen en Inburgering uitgebracht door Tom Ongena en Yves Buysse over het ontwerp van decreet tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie.
  In the Netherlands, compulsory voting was introduced in 1917 as part of a broader package of political-electoral measures that included universal male suffrage and the system of proportional representation. In particular, the Catholics were in favour of compulsory voting, being apprehensive that without compulsory voting some of their voters would stay at home, as they were assumed to do so traditionally in constituencies that were won or lost in advance. Half a century after its introduction, various initiatives were undertaken to abolish compulsory voting, and this finally took place in 1970 (Irwin & van Holsteyn, 2010). The main argument was that citizens should be treated as ‘political adults’. Compulsory voting was perceived as an undemocratic encroachment on individual freedom. Therefore, voting should be considered a right and not a duty. The final decision, although not taken unanimously, was hardly controversial. ­Lijphart (2010) noticed the absence of a great national debate on the issue.
  To some extent, the argument in the Netherlands and in Flanders (50 years later) for abolishing compulsory voting ran quite similar. In essence, the decisions were motivated by legal/philosophical frames concerning rights and duties of citizens versus the state and the formal/normative design and operation of institutions. In the Netherlands little attention was paid to attitudes and behaviour of both politicians and citizens and the consequences of the functioning of institutions for those attitudes and behaviour (Irwin & van Holsteyn, 2010). Faced with the decline in turnout after the abolition of compulsory attendance in the Netherlands, Lijphart raised the question of whether it could have been foreseen by politicians or political scientists. Albeit dating from before World War II, literature on the relationship between compulsory attendance and turnout was already available at the time of the abolition (Lijphart, 2010). Irwin and van Holsteyn (2010) additionally argued that empirical data from the Netherlands not only predicted the decline in turnout but also warned of a social bias in the turnout.
  Unlike the debate in the Dutch parliament, the Flemish parliament organised expert hearings with political scientists. A number of them warned about the consequences of the abolition of compulsory voting with regard to the expected decline in electoral participation and the social bias of the electorate.⁷x 790 (2020-2021) – Nr. 6 ingediend op 16 juni 2021 (2020-2021) Verslag van de hoorzitting namens de Commissie voor Binnenlands Bestuur, Gelijke Kansen en Inburgering uitgebracht door Tom Ongena en Yves Buysse over het ontwerp van decreet tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie. This was also repeated before the final vote in parliament in a newspaper opinion piece signed by more than twenty Belgian political scientists (Ackaert et al., 2021). Nevertheless, the parliament agreed to abolish compulsory voting. By consequence, Lijphart’s and Irwin & van Holsteyn’s assumption that the input of political science and social science research might have resulted in a different outcome of decision-making does not seem to apply to Flanders.
  The limited weight of political scientists’ input into the debate in the Flemish Parliament is a derivative of the marginal role of parliament in a partocracy with no scientific and political traditions of policy analysis in Belgium. De Winter and Wolfs (2017) emphasise the governmental coalition policy programme as the most central ex ante control mechanism of policy. They conclude that this coalition agreement

  not only ties coalition parties to each other, but also enhances discipline within each coalition party and its parliamentary group, and therefore undermines the relevance of policy analysis by MP’s and parliament as a collective actor.

• 3 Compulsory Voting and Turnout

  Compulsory voting rules have proven to be a powerful tool for increasing levels of electoral participation, as it is confirmed by a vast comparative literature examining the impact of compulsory voting on turnout (Birch, 2008; Blais & Aarts, 2006; Blais & Dobrzynska, 1998; Cancela & Geys, 2016; Dassonneville et al., 2019; Quintelier et al., 2011; Singh, 2011; van der Meer & van Deth, 2010).
  Figure 1 shows the evolution of electoral absenteeism in local and general elections, for both Belgium and the Netherlands from the 1960s onwards.⁸x For Belgium, only data from 1976 onwards provides a valid picture of the evolution of absenteeism. Until then, the members of the college of mayor and aldermen reviewed the list of voters in an election year between 1 July and 20 September. Because of the large time frame between the closing of the voter register and the election date, it regularly happened that voters who passed away during this time span, were considered absent in the elections and included in the official abstention numbers. Following the introduction of the elections law of 5 July 1976, the registers of voters are permanently kept up to date by the college of mayor aldermen. As such, the time span between the establishment of the voters’ list and the election date has been shortened (De Winter, 1978). Furthermore, until 1970 it regularly happened that no elections were organised in a significant number of municipalities owing to a lack of competition. In 1970 it concerned 12.6% of municipalities, in 1964 even 15.5% (Dewachter, 1973). This also hinders the mapping of reliable data regarding absenteeism at Belgian local elections before 1976. Therefore, only the figures from 1976 onwards to evaluate electoral absenteeism in Belgium are presented.
  Figure 1 confirms the frequently recurring observation that compulsory voting is a powerful tool to boost turnout levels in the Low Countries. Both at the national and at the local level, electoral absenteeism tripled in the Netherlands at the first elections after the abolition of compulsory voting. Since then, electoral absenteeism increased up to 21% in the 2021 parliamentary elections and even 50% in the 2022 local elections. However, the evolution of absenteeism did not increase linearly over time, either for the national or for the local elections.

  The evolution of electoral absenteeism in local and general elections, for Belgium and the Netherlands.For Belgium, this data is based on the calculations of De Winter and Ackaert. For the Netherlands, we relied on the data of van Ostaaijen.
  

  In Belgium as well, absenteeism increased in the period under review, both at the national and at the local levels. However, the increase was not nearly as spectacular as in the Netherlands. During the parliamentary elections of 2019, 11.6% of the voters abstained from voting, while in 2018 9.4% of the voters did not cast a vote during the local elections.
  Another difference between Belgium and the Netherlands concerns the differing levels of absenteeism during national and local elections. In general, absenteeism during Belgian elections remained lower in local elections than in the general elections. However, the differences in turnout between both types of elections remain rather limited. On the contrary, at the end of the period under review, absenteeism in the Netherlands was twice as high in the local elections compared with the national elections. This finding is in line with the theory of second-order elections, which posits that participation in second-order elections (local, regional and European elections) is lower because less is at stake in second-order elections compared with national elections (Reif & Schmitt, 1980).
  On the basis of the experiences in the Netherlands, one would expect that the abolition of compulsory voting at the Flemish local elections would lead to a significant decline in voter turnout. This expectation is also confirmed by various surveys that estimated the impact of abolishing compulsory voting. Based on these simulations, Hooghe and Pelleriaux (1998) estimated that the abolition of compulsory voting in Belgium would lead to a drop in turnout of about 20%. According to more recent studies, the decline in turnout would fluctuate between 30% and 35%. Verlet et al. (2010) measured intentions to show up at local elections without compulsory voting and recorded 67.7% positive responses compared with 31.7% negative responses. Five years later, Reuchamps et al. (2015) counted 56.6% respondents who indicated they would ‘always’ vote without compulsory voting versus 23% who were certain to ‘never vote again’. For about 20% of respondents, the answer varied between ‘often’ and ‘sometimes’. Although these simulation studies offer a first glimpse of the voter turnout at local elections in the Flemish region of Belgium after the abolition of compulsory voting, previous research indicates that the effect of abolishing compulsory voting is often underestimated. Studies in the Netherlands that gauge the willingness to vote traditionally underestimate the percentage of non-voters compared with the actual absenteeism at the polls (Boelhouwer, 2002; Krouwel et al., 2009).

• 4 Compulsory Voting and ­(Un)equal Participation

  A vast strand in literature considers high voter turnout as an effective instrument for equal participation since it ensures that voters wield equal influence in the democratic process, regardless of their resources (Dahl, 1989; Lijphart, 1997). When turnout rates are high, little room is left for disparities and inequalities in turnout (Dassonneville & Hooghe, 2017; Jaitman, 2013). When participation levels are low, on the contrary, the less well off are the ones to be less likely to turn out to vote. According to this view, compulsory voting systems guarantee that parties pay attention to these groups of citizens and contribute to their political integration.
  In what follows, we investigate this claim and develop a comprehensive overview of individual-level determinants of voter turnout in Belgium and the Netherlands. For this purpose, we collected and analysed previously published research on determinants of voter turnout in either of both or both countries. Given the system of compulsory voting in Belgium, existing research consists of studies investigating the (simulated) consequences of abolishing compulsory voting. For the Netherlands both studies investigating the impact of the abolishment of compulsory voting on voter turnout and more general studies on voter turnout are included in the analysis. We included empirical studies published between 1970 and 2021. We chose 1970 as the starting date for the literature review since in this year the first Dutch elections without compulsory voting were organised. This sample selection results in a total of 27 articles.
  In general, many hypotheses have been advanced on the reasons for citizens’ participation in elections, and a large variety of empirical support has been found. Following Smets and Van Ham (2013), we bring unity into this wealth of individual-level explanations by accommodating them into five broad theoretical models of individual-level voter turnout that reflect the main theoretical approaches in the literature: the resource model, the mobilisation model, the socialisation model, the rational choice model and the psychological model.⁹x The political institutional model is not considered in the remainder of the article. The political institutional model considers the decision to vote to be related by the political and institutional context. In the third paragraph we already elaborated on the most striking institutional difference between Belgium and the Netherlands, namely the (non-) compulsory voting system and its impact on voter turnout. In what follows, we explain the theoretical models consecutively and categorise all independent variables into one of the models.

  4.1 The Resource Model

  According to the resource model, voters’ propensity to vote is related to their (political) resources such as time, money and skills (Verba & Nie, 1972; Verba et al., 1995) and to some socio-demographic characteristics that usually correlate with these resources (gender, age, socio-economic status and level of education).

  4.1.1 Income and Occupation

  Income appears to be a factor positively related to the likelihood to vote, with higher income groups generally being more likely to vote than lower income groups under a voluntary voting system (Smets & Van Ham, 2013). When voting is compulsory, the over-representation of the rich relative to the poor decreases as those with a lower socio-economic status are compelled to vote (Dassonneville et al., 2017; Guntermann et al., 2020).
  For the Netherlands, various studies confirm that lower income groups are less likely to vote when they are not obliged to. After the abolition of compulsory voting in the Netherlands, Irwin (1974) perceived a slightly increasing turnout with increases in income. More recent studies confirm that high-income groups are more likely to participate in elections (Aarts, 2001; Steenvoorden & Van der Waal, 2016; van Ostaaijen et al., 2016). Studying the socio-demographic background of people not having the intention to vote, Boelhouwer (2002) concludes that the number of non-voters decreases as the income level increases. Similarly, low-income groups were over-represented among non-voters when compared with voters (Jansen & Boogers, 2018).
  For Belgium, the correlation between willingness to vote and income level is less clear-cut. On the one hand, neither Ackaert and De Winter (1993) nor Verlet et al. (2010) finds significant evidence that higher-income groups would be more inclined to vote under a voluntary voting system. Reuchamps et al. (2015), on the other hand, do find a negative correlation between income level and the probability of a respondent turning out to vote under a voluntary voting regime. As income levels increase, people are more willing to vote voluntarily during elections.
  A possible explanation for these inconclusive results may be the close relationship between income and other socio-economic and socio-demographic variables such as educational level, ­occupational status and occupational type (Ackaert & De Winter, 1993; Verlet et al., 2010). Simulating the potential absenteeism after abolishing compulsory voting in Belgium, neither Ackaert and De Winter (1993) nor Verlet et al. (2010) finds a relationship between occupational status and voter turnout in a multivariate analysis. ­Occupational type, however, seems to be positively related to respondents’ willingness to vote: people in professions that require more social and intellectual skills are more likely to vote under a voluntary voting system (Ackaert & De Winter, 1993; Hooghe & Pelleriaux, 1998). Similarly, Verlet et al. (2010) concluded that people who perceive themselves to be more important in society show a higher willingness to vote.

  4.1.2 Age

  Based on an extensive literature review, Smets and Van Ham (2013) identify age as one of the three most common independent variables in research on individual voter turnout. Young people in western democracies more often refrain from taking part in elections than older age groups (Kimberlee, 2002). In general, the propensity to vote increases steadily with age and typically declines in old age when people start to withdraw from social life (Dassonne­ville & Hooghe, 2017; Fieldhouse et al., 2007; Smets & Van Ham, 2013; van der Meer & van Deth, 2010). A possible explanation may be that older people are more likely to have been socialised into the habit of voting and have a stronger sense of civic duty (Blais et al., 2004; Franklin et al., 2004). When focusing on countries with compulsory voting, the aforementioned association between age and voting turns out to be much weaker. When voting is compulsory, younger people are more inclined to vote than under voluntary voting systems. However, Quintelier et al. (2011) conclude that, also in countries with compulsory voting, older citizens often outnumber younger citizens at the polling stations.
  These conclusions with regard to compulsory voting seem to be partially confirmed for the Belgian situation. Fieldhouse et al. (2007) show the turnout gap between age groups to be more limited in Belgium than in countries without compulsory voting. Similarly, in Belgium, voting turnout does not differ significantly between age groups (Quintelier et al., 2011).
  Studies simulating the effects of abolishing compulsory voting in Belgium show no conclusive results with regard to the relationship between respondents’ age and their willingness to vote. Neither in multivariate nor in bivariate analyses were significant differences found between respondents’ age and their self-reported willingness to vote during elections under a voluntary voting regime (Ackaert & De Winter, 1993; Hooghe & Pelleriaux, 1998; Reuchamps et al., 2015).
  The relationship seems to differ, however, according to the policy level and the type of analysis. On the one hand, bivariate tests point to a non-linear relationship between respondents’ age and their willingness to vote, with the age group between 55 and 64 being most willing to voluntarily take part in elections. The results of the multivariate analyses, on the other hand, indicate that the willingness to vote increases with age, only for the regional and federal elections. Seemingly contradictory to these results, Reuchamps et al. (2015) found a significant linear positive relationship between age and voting turnout only at the municipal level.
  In the Netherlands, age differences with regard to voting in elections increased significantly after the abolition of compulsory voting. In 1967, under a system of compulsory voting, turnout figures differed only slightly between age groups. The abolition of compulsory voting in 1970, however, led to a substantial age gap of about 20% in voter turnout: while 60.2% of the voters between 21 and 24 participated in the 1970 elections, this percentage increased to 80% for respondents over 65. In line with the expectations in 1970, a curvilinear relationship was found between age and voter turnout (Irwin, 1974).
  More recent studies confirm the relationship between age and voter turnout in Dutch elections. In general, younger citizens are least inclined to cast a vote. Numerous studies find a linear relationship between age and voter turnout, with voter turnout increasing with age (Aarts, 2001; den Ridder & Dekker, 2016; Howe, 2006; Jansen & Boogers, 2018; Steenvoorden & Van der Waal, 2016). Other studies, however, find a curvilinear relationship, with both the youngest and oldest age cohorts being least inclined to vote. Similarly, the middle group has the lowest percentage of non-voting respondents (Boelhouwer, 2002).

  4.1.3 Gender

  Traditionally, men have long been considered to be more likely to participate in elections than women. However, no conclusive findings have emerged on this topic. On the one hand, some studies do not find significant gender differences with regard to electoral turnout (Inglehart & Norris, 2003; Smets & Van Ham, 2013; Van der Meer & van Deth, 2010) or find the gender effect in turnout to be incorporated by other socio-demographic variables such as educational and professional status (Campbell, 2006). Studying the impact of compulsory voting on stratification patterns, Quintelier et al. (2011) find no significant differences between men and women with regard to participation in elections under a system of compulsory voting.
  On the other hand, other studies report higher rates of turnout among women than men (Coffé & Bolzendahl, 2010; Smets & Van Ham, 2013).
  In Belgium and the Netherlands as well, studies are inconclusive with regard to the relationship between gender and voter turnout. In Belgium, a first strand of literature points to an over-representation of men in turnout levels. Quintelier et al. (2011) conclude that even under a system of compulsory voting, women participate less during Belgian elections. Similarly, simulation studies by Hooghe and Pelleriaux (1998) and Verlet et al. (2010) demonstrate that abolishing compulsory voting in Belgium would lead to an over-representation of men at the ballot box. Under a voluntary system, women would be less likely to participate in elections.
  Yet other studies conclude that women do not show up less because of their gender than because of a lower educational and professional status or political variables (Ackaert & De Winter, 1993; Billiet, 2002; Reuchamps et al., 2015).
  The available empirical evidence for the Netherlands demonstrates that the introduction of a non-compulsory system of voting in 1970 increased the gender gap, men being more likely to vote voluntarily during elections than women (Irwin, 1974). More recent studies, however, point to no or only small gender differences in actual voter turnout and voting intentions, both at the national and at the local levels (Jansen & Boogers, 2018; Steenvoorden & Van der Waal, 2016; van der Meer & van Deth, 2010). If significant differences are found between men and women, they disappear under the control of variables such as education or income (Boelhouwer, 2002).

  4.1.4 Education

  Various studies conclude that educational level serves as a stratification mechanism between those who participate in elections and those who do not (Smets & Van Ham, 2013). Highly educated respondents are found to vote more often than less educated citizens (Dassonneville et al., 2017; Quintelier et al., 2011; Van der Meer & Van Deth, 2010). The question, remains however, whether the introduction of compulsory voting would reduce inequality. On the one hand, Dassonneville et al. (2017) argue that a system of compulsory voting would be effective in reducing inequalities with regard to education. Quintelier et al. (2011), on the other hand, state that a system of compulsory voting is ineffective in reducing stratification, since it simply raises the turnout level for all groups within society, without levelling out any differences between groups.
  The transition from compulsory to voluntary voting in the Netherlands has resulted in a turnout gap according to educational attainment. The lowest educational groups, in particular, have dropped out of electoral participation (Boelhouwer, 2002; Dassonneville & Hooghe, 2017; Hakhverdian et al., 2012; Irwin, 1974; Steenvoorden & Van der Waal, 2016). Bovens and Wille (2010) confirm that after abolishing compulsory voting the turnout of the more highly educated groups remained high, whereas the turnout of the less educated fluctuated at a much lower level. The increasing weight of the well-educated voters in the ballot is most outspoken in second-order elections (municipal, provincial and European). Over the years, both Bovens and Wille (2010, 2011) and Dassonneville and Hooghe (2017) find growing educational inequalities in turnout. Hakhverdian et al. (2012), however, do not find any evidence of growing disparities in turnout over time. These possibly contradictory findings may be related to the different time frames of the studies and the different political arenas and participation forms that were included in the studies.
  Studies concerning the abolition of compulsory voting in Belgium have shown mixed results. In line with the findings of Topf (1995), Dassonneville et al. (2017) do not find significant differences in voter turnout between lower and higher educated voters under the compulsory voting system in Belgium. Studies simulating the impact of abolishing compulsory voting in Belgium are less conclusive. While Ackaert and De Winter (1993) conclude that education has little autonomous effect on citizens’ willingness to vote, Hooghe and Pelleriaux (1998), Billiet (2002) and Verlet et al. (2010) identify education as a very important independent variable explaining the willingness to vote. According to the authors, the abolition of compulsory voting in Belgium would lead to an over-representation of citizens with a high educational level. The results of Reuchamps et al. (2015) confirm these findings, but only for the federal level. The authors suggest that this might be because the effect of education is already incorporated by another variable in their model, namely political interest.

  4.1.5 Marital Status

  Lastly, Smets and van Ham (2013) find a statistically significant relationship between marital status and voter turnout. A possible explanation is that married couples are more likely to conform to the idea of ‘good citizens’ and consider political engagement a civic duty (Denver, 2008).
  In Belgium and the Netherlands, the impact of marital status on voter turnout is not frequently investigated. Whenever it is included in analyses explaining voter turnout or absenteeism, no statistically significant relationship can be found (Ackaert & De Winter, 1993).

  4.2 The Mobilisation Model

  The mobilisation model of voter turnout starts from the idea that parties, candidates, interest groups and social movements do mobilise citizens to vote. As such, one may expect that under systems of compulsory voting, people without societal engagement would be more inclined to vote compared with voluntary voting regimes. Contrary to the findings of Smets and Van Ham (2013), the mobilisation model is less common in studies investigating determinants of voter turnout in Belgium and/or the Netherlands. In both these countries four studies cover at least one mobilisation variable.

  4.2.1 Attendance of Religious Services

  The meta-analysis of Smets and Van Ham (2013) shows that a religious denomination is in most instances not significantly related to turnout. Attendance of religious services is found to affect individual turnout roughly half of the time.
  In Belgium, attendance of religious services is found to affect voter turnout positively. One may assume that persons inclined to fulfil their religious duty are equally inclined to fulfil their civic duty (Ackaert & De Winter, 1993). Self-reported identification with religion does not show any clear differen­ces with regard to respondents’ willingness to vote (Hooghe & Pelleriaux, 1998). These seemingly contradictory results may be explained by the fact that, contrary to the attendance of religious services, mere identification with religion does not require any mobilisation.
  In the Netherlands as well, both religious denomination and frequent church attendance are positively related to turnout (Boelhouwer, 2002; Howe, 2006). Among Dutch-Muslim respondents as well, there is a strong and statistically significant positive association between regular praying and voting in the national elections (Moutselos, 2020).

  4.2.2 Organisational Membership

  People can be members of various associations and organisations. In general, membership of formal and informal groups broadens interest, interpersonal contacts and social integration. This stimulates political participation (Putnam, 2000). Therefore, membership of all kinds of organisations is expected to be positively related to individual voter turnout.
  In general, this expectation seems to be confirmed for the Belgian situation (Verlet et al., 2010) Similarly, citizens who perform unpaid voluntary work seem to be significantly more inclined to vote under a voluntary voting regime as well (Verlet et al., 2010).
  In contrast to the aforementioned studies, Ackaert and De Winter (1993) do not find significant differences with regard to voting turnout according to the respondents’ self-reported degree of social isolation. This may be related to the construction of the scale, which focusses exclusively on loneliness and informal social capital and not on associational membership or formal social capital.
  In the Netherlands, research is in line with the expectation. Both volunteers and members of organisations are more inclined to vote under a voluntary voting regime. Being an organisation member and being in contact with all kinds of groups through it affects voting positively, compared with not being a member of an organisation. In addition, the frequency of participation in sociocultural activities is also positively related to the chances of voting (Boelhouwer, 2002; Kranendonk & Vermeulen, 2019).
  Social isolation, on the other hand, is negatively correlated with voting behaviour. Similarly, having a relatively large number of friends increases the chance of voting (Boelhouwer, 2002; Kranendonk & Vermeulen, 2019).

  4.2.3 Media Exposure and Campaign Advertisement

  On the basis of their literature review, Smets and van Ham (2013) conclude that exposure to (political) news in the media has a positive impact on voter turnout among citizens. Baekgaard et al. (2014) confirm that local news about elections generally leads to an increase in voting turnout if politically relevant news is being communicated. Campaign advertisements do not seem to have a significant effect on voter turnout. This is possibly because although political campaigns aim to mobilise people to vote, negative campaigns may have a demobilising effect (Smets & van Ham, 2013).
  For Belgium, Ackaert and De Winter (1993) conclude that those who closely follow the election campaign are more likely to vote effectively. According to the authors, it may be assumed that people who pay close attention to the campaigns are more open to the stimuli from the political system.

  4.3 The Rational Choice Model

  The rational choice model considers the act of voting as the result of a cost-benefit analysis. According to this model, citizen will vote when they are convinced that the benefits of voting outweigh its costs. All else being equal, an individual’s expected likelihood of voting is higher when the cost of voting is low and when he or she derives some benefit from the act of voting (Downs, 1957). In other words, the higher the stakes in the elections, the more inclined citizens will be to turn out to vote. One may expect compulsory voting rules to impact citizens’ cost-benefit analysis by compelling people who would be inclined to abstain under voluntary rules. Yet compulsory voting disincentivises the exit option and compels citizens to vote, even without severe or enforced sanctions (Birch, 2008; Blais & Aarts, 2006; Singh, 2015).
  In the Low Countries, rational choice studies on voter turnout are not very common. This is in line with the results of the meta-analysis of Smets and van Ham (2013). Out of the 111 investigated studies on voter turnout published in ten top journals in political science and political behaviour between 2000 and 2010, only 10% included variables related to the rational choice model. According to the authors, this may be the consequence of the time period assessed in this study. Our results seem to nuance this explanation, at least for Belgium and the Netherlands.
  In Belgium, Verlet et al. (2010) find that citizens with a stronger utilitarian individualistic attitude, who strive for their individual aims without taking care of others, are more likely to refrain from voting under a voluntary system. Caring about others may increase turnout during elections. Secondly, perceptions of a high policy impact on a person’s daily life encourage and stimulate citizens to turn out to vote. Someone who assesses the policy influence to be larger is more likely to vote voluntarily during elections (Reuchamps et al., 2015). In this instance, the benefits of voting outweigh the costs. Similarly, Dutch research on national, European and local elections finds a relationship between individual voter turnout and the perceived importance of a policy level. Voters who consider that less is at stake are less likely to go to the polls during both national and local elections. In other words, citizens who turned out rated the importance of the policy level higher compared with abstainers (Lefevere & Van Aelst, 2014). In line with these findings, Jansen and Boogers (2018) conclude that at the local level abstainers more often than voters assess the impact of municipalities on their daily lives to be low. Furthermore, a sense of civic duty seems to convince Dutch voters to cast a vote (Howe, 2006).

  4.4 The Socialisation Model

  The socialisation model stresses the importance of socialisation processes on the decision of whether or not to take part in elections. The years between childhood and adulthood are generally considered a formative period for the development of political attitudes and behaviours (Plutzer, 2002). Various socialising agents (family, friends, peers, school, etc.) play an important role in the developments of these attitudes and norms that influence behaviour. For example, being raised to believe that voting is an important aspect of good citizenship should encourage voting when a person comes of age (Alwin et al., 1991).
  In line with the meta-analyses of Smets and van Ham (2013), socialisation studies are somewhat less common in the Netherlands and, especially, in Belgium. In conversations with Dutch non-voters, De Greef (2002) notes that political interest usually comes from regular contacts with politics during childhood and adolescence.
  Secondly, regular talks with friends, family members or neighbours have a positive impact on citizens’ willingness to vote under the Dutch voluntary voting regime (Boelhouwer, 2002). Similarly, people who do not listen during political conversations are more strongly represented in the group of non-voters than people who indicate that they participate actively or passively in political conversations (Boelhouwer, 2002).

  4.5 The Psychological Model

  Psychological models of turnout represent a wide variety of approaches to explain voter turnout. Explanatory factors range from cognitive characteristics, such as political knowledge, political interest and political efficacy, to more personal characteristics and attitudes such as party identification and ideological preferences (Smets & Van Ham, 2013).

  4.5.1 Political Interest and Political Knowledge

  Political interest and political knowledge are important determinants of citizens’ likelihood to vote during elections, less interested and less knowledgeable citizens being less likely to cast a ballot under voluntary voting (Smets & van Ham, 2013). Under compulsory voting less knowledgeable and less interested citizens participate more often than they do under voluntary voting rules (Selb & Lachat, 2009; Singh, 2015). As such, the impact of political knowledge on the probability of turning out to vote is weaker in contexts where voting is mandatory (Dassonneville et al., 2017), and political compulsory voting rules significantly reduce political knowledge-based stratification in turnout. Discussion remains, however, on the causality of this correlation. ­Lijphart claims compulsory voting to be an important motivation for citizens to acquire more political information. According to this argument, the obligation to vote stimulates citizens to acquire more political information in order to cast a meaningful vote. Studies, however, do not find unequivocal evidence for this claim (Leon & Rizzi, 2014; Quintelier et al., 2011; Sheppard, 2015; Shineman, 2018).
  Simulation studies investigating potential absenteeism after the abolition of compulsory voting in Belgium confirm that citizens with a stronger interest in politics are more likely to participate in elections (Ackaert & De Winter, 1993, Billiet, 2002; Reuchamps et al., 2015). The results are less clear with regard to political knowledge, some studies having found a positive significant effect (Ackaert & De Winter, 1993; Billiet, 2002) and others not having found any (Reuchamps et al., 2015). These divergent conclusions may be related to the operationalisation of the independent variable and/or other variables included in the model.
  Similarly, Irwin (1974) concludes for the first election after the abolition of compulsory voting in the Netherlands that the drop in turnout among non-newspaper readers and the generally uninterested citizens is two to three times higher than the drop among regular readers and citizens with interest in political matters. More recent studies confirm the positive relationship between likelihood to vote and political interest. Citizens who express more interest in politics in general and actively follow the political news more frequently vote in elections (Aarts, 2001; Boelhouwer, 2002; den Ridder & Dekker, 2016; Irwin & Van Holsteyn, 2010; Steenvoorden & Van der Waal, 2016).
  In Dutch national and local elections, respectively, Howe (2006) and Steenvoorden and Van der Waal (2016) perceive a decline in participation across levels of political knowledge, the most knowledgeable respondents being most inclined to vote during elections. Similarly, people who perceive politics to be too complicated are less inclined to vote than people who do not share this opinion (den Ridder et al., 2012).

  4.5.2 Political Efficacy

  Internal efficacy usually refers to the degree to which people think they can influence the government or policy outcomes, while external efficacy indicates the extent to which citizens perceive a government to be responsive to their interests. Both forms of efficacy are expected to increase turnout under a voluntary voting regime.
  For Belgium, the empirical findings with regard to political efficacy are mixed. On the one hand, our overview of Belgian studies confirms that citizens who feel less politically powerful are less inclined to vote under a voluntary system (Ackaert & De Winter, 1993; Billiet, 2002; Verlet et al., 2010). Reuchamps et al. (2015), however, only find significant differences in voting turnout according to the level of internal political efficacy on the basis of bivariate testing, not on the basis of multivariate analyses. External political efficacy seems to be positively related to voter turnout at all policy levels in Belgium (Reuchamps et al., 2015).
  In the Netherlands as well, both internal and external efficacy are positively related to voter turnout. At the municipal level, den Ridder and Dekker (2016) conclude that people who are convinced that local councillors are not concerned about people like them are less likely to vote. Furthermore, people who are convinced of the added value of political participation vote more often.

  4.5.3 Party Identification and Ideological Self-placement

  Thirdly, the analysis of Smets and Van Ham (2013) shows that citizens’ party identification is positively related to their propensity to vote.
  For Belgium, only Reuchamps et al. (2015) tested this relationship and found a positive significant effect. As citizens identify more strongly with a party, they indicate more often that they would vote under a voluntary voting system. However, party identification did not have a significant effect on voter’s likelihood to vote for municipal elections. This does not come as a surprise since the policy level closest to the citizen has own (sometimes less politicised) local dynamics. A lot of local lists are, for example, not related to a national party (Buelens & Deschouwer, 1996). As such, voting turnout at the local level may be less dependent on political dynamics and identification with a political party.
  For the Netherlands, Irwin (1974) identifies party identification as one of the variables with the strongest impact on going to the polls. More specifically, after compulsion was removed in the Netherlands, the difference in voting turnout between strong supporters and those with no identification with a party increased from 9% to 36%. The study of Howe (2006) confirms the positive correlation between party identification and voter turnout. In addition, Boelhouwer (2002) concludes that nearly all citizens who are members of a political party go to the polls under a voluntary voting system.
  With respect to ideological self-placement, a common hypothesis is that right-wing or conservative voters vote more often because they tend to perceive voting as a civic duty more often than do left-wing or liberal voters. On the basis of their literature analysis on voter turnout, Smets and van Ham (2013), however, do not find a correlation between voter turnout and ideological self-placement.
  In the Netherlands, only one study investigated the relationship between voter turnout and political views. The hypothesis is only partially confirmed by Aarts (2001), who concludes that people with more extreme political views are more likely to go to the ballots during national elections.

  4.5.4 Satisfaction with Democracy and Policy

  Furthermore, satisfaction with democracy is hypothesised to increase turnout. The general picture in Belgium is that satisfaction with the functioning of democracy is not related to respondents’ willingness to vote (Reuchamps et al., 2015; Verlet et al., 2010). Nevertheless, Ackaert and De Winter (1993) conclude that negative feelings towards politics are higher among potential abstainers than among people who indicate to vote under a voluntary voting regime.
  Similarly, in the Netherlands political cynicism is negatively related to citizens’ willingness to vote (Aarts, 2001; Boogers et al., 2010). Furthermore, the impact of satisfaction with government policy found to have a significant positive effect, both at the local (den Ridder & Dekker, 2016) and at the national (Aarts, 2001; Boelhouwer, 2002) levels.

  4.5.5 Trust

  In general, people that have more confidence in the political institutions and in others will more likely have a positive outlook on the workings of the electoral process (Bélanger & Nadeau, 2005)., and one may therefore expect them to show a higher willingness to vote when voting is not or no longer compulsory.
  In Belgium and the Netherlands, many studies have investigated the impact of both political and social trust on potential absenteeism and voter turnout. A majority of studies in the Netherlands do find a significant positive relationship between citizens’ level of political trust and their willingness to vote (Aarts, 2001; den Ridder & Dekker, 2016; Steenvoorden & Van der Waal, 2016). The results are somewhat different in Belgium, where most of the studies conclude that willingness to vote is not related to the level of political trust (Reuchamps et al., 2015; Verlet et al., 2010). In both studies, however, bivariate analyses do show political trust to be significantly lower among potential non-voters. The disappearance of this relationship in multivariate analyses implies that other variables, closely related to political trust, are responsible for the bivariate results. Secondly, Verlet et al. (2010) conclude that more specific measures with regard to political trust on different policy levels do correlate significantly with voter turnout at the respective elections.

  4.5.6 Health and Well-being

  Lastly, people’s willingness to vote may also be related to more personal variables. On the basis of their meta-analysis, Smets and van Ham (2013) find that having a hard-working personality boosts turnout.
  Boelhouwer (2002) concludes for the Netherlands that general satisfaction affects turnout significantly. The number of citizens who describe themselves as unhappy is significantly higher among abstainers than among voters. The same holds true for people with bad health. However, the impact of these personal characteristics disappears when other, more ‘political’ variables are included in the analysis.
  Table 1 summarises the impact of the various variables on individuals’ intention to vote under a non-compulsory voting system in Belgium. Table 2 includes the results of the analysis for the Netherlands.¹⁰x For the sake of parsimony and clarity, only quantitative studies that contained at least three variables explaining voter turnout were included in the tables. The variables are divided according to the different theoretical models that have been used to structure the findings. For each variable, the direction of the effect is denoted by ‘+’ (positive relationship), ‘−’ (negative relationship) or ‘n.s.’ (non-significant relationship) after the variable designated in the table. Non-inclusion of variables in the concerned research is denoted by cells marked grey.
  In general, we do notice that there are some similarities between Belgian and Dutch studies. Regardless of institutional rules, education, political interest, political knowledge and feelings of external political efficacy seem to boost the likelihood of turnout.

• 5 Compulsory Voting and the Quality of the Vote

  The aforementioned studies on turnout under a non-compulsory voting regime focus on the equality of participation. The foregoing analysis identifies political interest and knowledge as important drivers of voter turnout. Building on this argument, a second strand of literature contends that compulsory voting harms the quality of the vote. Instead of focussing on the realised turnout levels, the focus is on the systematically different decision calculus of potential voters in compulsory and voluntary systems (Singh, 2015). In general terms, scholars are worried that the obligation to vote has a negative impact on the representative function of elections since low involved and low interested voters – who are forced to vote under a compulsory voting regime – will cast blank, invalid or non-informed votes. By doing so, they undermine the fundamental mechanisms of voting, such as accountability or ideological proximity.
  The general argument is that voters who lack political sophistication are less successful in correctly identifying the political party that is best able to represent their preferences (Rosema, 2007; Selb & Lachat, 2009). Similarly, Selb and Lachat (2009) conclude that voters in Belgium who are less likely to vote under voluntary rules are also less likely to have voted spatially. As such, compulsory voting reduces the accuracy of the electoral signal. Rosema (2007), however, has shown that even those with low levels of political sophistication still manage to achieve a fairly good match between their preferences and the party they vote for.
  In what follows we give an overview of studies investigating the impact of compulsory voting on the quality of the vote. Consecutively, we discuss studies concerning the number of blank and invalid votes, ideological proximity and accountability mechanisms under both voluntary and compulsory voting systems. Given the scope of this article, we limit our overview to empirical studies, executed in the Netherlands and/or Belgium. Because of our focus on the impact of (abolishing) compulsory voting on the quality of the vote, studies from the Netherlands are rather limited, mostly to studies that were conducted shortly after the abolishment of compulsory voting in 1970. For Belgium, we include studies simulating the potential impact of the abolishment of compulsory voting.

  5.1 Blank and Invalid Voting

  Available comparative studies univocally find higher invalid voting rates in countries with a compulsory voting system (Kouba & Lysek, 2019; Power & Garand, 2007; Reynolds & Steenbergen, 2006). This is presumably because voters who would abstain under voluntary voting express their disinterest or discontent by casting a blank or invalid vote when compelled to vote (Kouba & Mysicka, 2019).
  Data for Belgium and the Netherlands confirm this pattern, with the amount of blank and invalid votes being substantially higher in Belgium compared with the Netherlands.
  In Belgium the total number of blank and invalid votes fluctuates between 6% and 9%. Traditionally, these numbers are lower during municipal elections (own calculations; Dejaeghere & Vanhoutte, 2016).¹¹x During the most recent federal elections the total number of blank and invalid votes was 6%. For the regional elections, 5% of the Flemish voters voted blank or invalid, 8.3% in Wallonia and 6.7% in Brussels. In 2018, 3.8% of the voters voted blank or invalid during the Flemish municipal elections. However, when looking at the Belgian municipal level, Dejaeghere and Vanhoutte (2016) find that low turnout is associated with a higher proportion of blank and invalid voting.
  After the abolition of compulsory voting in the Netherlands, the amount of invalid and blank ballots dropped from 2.8% in 1967 to 0.7% in 1971. Today, 0.55% and 0.038% of the votes were blank or invalid during, respectively, the Dutch municipal and national elections of 2022.¹²x www.verkiezingsuitslagen.nl.

  5.2 Ideological Proximity

  Many studies that examine the impact of turnout levels and compulsory voting on the quality of the vote focus on the ideological proximity between voters and the parties they vote for (Dassonneville et al., 2017; Rosema, 2007; Selb & Lachat, 2009). According to the theoretical literature on representation, voters need to vote in line with their ideological positions and issue preferences in order to ensure that their interests are considered by elected representatives (Przeworski et al., 1999; Thomassen & van Ham, 2014). Many empirical studies have examined the impact of compulsory voting rules on the ideological distance between voters and the parties they vote for.
  On the basis of cross-national data from countries with and without compulsory voting rules, Dassonneville et al. (2017) find a negative impact of compulsory voting rules on proximity voting. Under compulsory voting, voters are less likely to vote for the most proximate party. Studies that explore variation in proximity voting within compulsory voting settings conclude that those who are more reluctant to vote under voluntary voting rules are less likely to vote for the most proximate party. However, the effect appears to be rather weak: those who are least willing to turn out to vote are about 10% less likely to vote for the most proximate party compared with those being most willing to vote (Dassonne­ville et al., 2019).

  5.3 Accountability Mechanisms

  Next to ideological preferences, voters can also consider incumbents’ performances in their own vote choices. Holding incumbents accountable for their actions and outcomes contributes to democratic representation since the fear of being held responsible for bad performance is an important incentive for incumbents to govern in the interests of their voters (Przeworski et al., 1999). Investigating the impact of compulsory voting on individuals’ vote choices, Dassoneville et al. (2017) conclude that accountability mechanisms are not weakened significantly under compulsory voting regimes. In other words, in countries with compulsory voting, incumbents are being held accountable for their actions to the same extent as in countries with voluntary voting regimes.

• 6 Conclusion and Discussion

  The debate on compulsory voting and voter turnout is fed by a trade-off, balancing between a high and equal turnout, on the one hand, and the quality of the vote, on the other. There is global unanimity among scholars that a mandatory system not only increases turnout levels but also contributes to more equal participation in elections. Emphasising the importance of equal turnout, a first strand in literature considers compulsory voting systems to be a powerful tool for increasing levels of participation among different social groups. A second strand in the literature contends, however, that compulsory voting harms the quality of the vote and leads to an increase in blank, invalid and/or non-informed votes.
  In this article, we have taken a step back and assessed where we stand with regard to research on compulsory voting and voter turnout in Belgium and the Netherlands. Whereas the Netherlands abolished the system of compulsory voting in the 1970s, Belgium has been one of the few remaining countries where turning up to vote is mandatory. From 2024 onwards, however, voting will no longer be compulsory for the local elections in the Flemish Region of Belgium.
  The literature review shows conclusively that the compulsory voting regime in Belgium increases voter turnout compared with the Netherlands. The other side of the coin is that there are a higher proportion of blank and invalid votes in Belgium compared with the case in the Netherlands. Nevertheless, the proportion of valid votes in the voting population is substantially higher in Belgium. In addition, research on the quality of the vote under compulsory voting regimes nuance the impact of compulsory voting rules on accountability mechanisms and – to a lesser extent – proximity voting.
  This review was aimed at shedding light on those factors that are consistently linked to individual level (intended) voter turnout. In the literature a wide variety of indicators of voter turnout have been explored in Belgium and the Netherlands. In general, especially variables related to the resources model and the psychological model are included in the existing literature. Only a handful of studies consider the decision to vote as the result of a personal cost-benefit calculation or use sociological explanations of turnout such as socialisation or learning.
  The variables that we have found to have a consistent effect on turnout are mainly variables related to the psychological model (i.e. political knowledge, political interest, satisfaction with government and policy). Furthermore, individual resources matter as well when it comes to voter turnout. Education and income, in particular, seem to have a (direct or indirect) effect on turnout.
  In addition, the impact of certain variables (i.e. gender, age) on voter turnout in Belgium and the Netherlands is not conclusive. This may be because analyses are often limited to bivariate analyses, especially in the Netherlands. As such, too little attention is paid to the net impact of the variable of interest and to other factors that may confound the impact of the variable of interest on voter turnout. Furthermore, the findings of each independent variable are highly dependent on the other variables that are included in the different models. This is especially the case because many of the variables correlated with voter turnout are interconnected since they measure closely related concepts (such as education, income and political interest).
  Thirdly (and possibly related to the former point), although the inclusion of this wide array of variables is theoretically well developed, more encompassing theoretical models are often underspecified – or sometimes completely lacking – in the current state of research.
  Lastly, the literature review points to a number of lacunas that could provide inspiration for future research. First, only a few studies have investigated voter turnout and the possibly varying impact of abolishing compulsory voting at different policy levels (Reuchamps et al., 2015; Verlet et al., 2010). Nevertheless, one may expect to find considerable differences in the propensity to vote between local and other elections, with Belgian voters being traditionally more likely to vote in local elections (in contrast to the Netherlands). Furthermore, both Verlet et al. (2010) and Reuchamps et al. (2015) find a different impact of party identification, age and degree of involvement on respondents’ willingness to vote at the local and other levels.
  Secondly, future research could further explore the efforts of (local and central) governments, parties and civil society to encourage people to cast their vote under a non-compulsory system. In doing so, specific attention should be paid to hard-to-reach audiences. This approach could yield new interesting insights. van Ostaaijen et al. (2019) have taken a first step in this regard by focusing on the (success of) efforts undertaken by municipal governments to improve turnout in local elections.

References

• Aarts, K. (2001). Dertig jaar later. De empirische houdbaarheid van argumenten voor en tegen opkomstplicht. Beleid en Maatschappij, 28(2), 80-94.

• Ackaert, J., (2005). Transformation of the Political Executive in Belgian Local Government. In R. Berg & N. Rao (Eds.), Transforming Local Political Leadership (pp. 168-169). Palgrave Macmillan.

• Ackaert, J., Bouteca, N., Celis, K., Claes, E., Dassonneville, R., Delheye, P., Devos, C., De Ceuninck, K., De Winter, L., Erzeel, S., Hennau, S., Hooghe, M., Kavadias, D., Kern, A., Meier, P., Reynaert, H., Severs, E., Sinardet, D., Thijssen, P., … Wauters, B. (2021). ‘De opkomstplicht afschaffen is een historische vergissing’, De Standaard, 15 juni 2021, 26.

• Ackaert, J. & De Winter, L. (1993). De afwezigen hebben andermaal ongelijk. De stemverzaking in Vlaanderen op 24 november 1991. In M. Swyngedouw, J. Billiet, A. Carton & R. Beerten (Eds.), Kiezen is verliezen. Onderzoek naar de politieke opvattingen van de Vlamingen (pp. 67-82). Acco.

• Alwin, D. F., Cohen, R. L. & Newcomb, T. M. (1991). Political Attitudes Over the Life Span: The Bennington Women after Fifty Years (464 pp.). University of Wisconsin Press.

• Baekgaard, M., Jensen, C., Mortensen, P. B. & Serritzlew, S. (2014). Local News Media and Voter Turnout. Local Government Studies, 40(4), 518-532. doi: 10.1080/03003930.2013.834253.

• Bélanger, E. & Nadeau, R. (2005). Political Trust and the Vote in Multiparty Elections: The Canadian Case. European Journal of Political Research, 44, 121-146. doi: 10.1111/j.1475-6765.2005.00221.x.

• Billiet, J. (2002). Opkomstplicht of vrijheid om al of niet te gaan stemmen in Vlaanderen: Het verkiezingsonderzoek van 1999. In M. Swyngedouw & J. Billiet (Eds.), De kiezer heeft zijn redenen. 13 juni en de politieke opvattingen van Vlamingen (pp. 129-146). Acco.

• Birch, S. (2008). Full Participation: A Comparative Study of Compulsory Voting (186pp.). Manchester University Press.

• Blais, A. & Aarts, K. (2006). Electoral Systems and Turnout. Acta Politica, 41(2), 180-196. doi: 10.1057/palgrave.ap.5500148.

• Blais, A. & Dobrzynska, A. (1998). Turnout in Electoral Democracies. European Journal of Political Research, 33(2), 239-261. doi: 10.1111/1475-6765.00382.

• Blais, A., Gidengil, E. & Nevitte, N. (2004). Where Does Turnout Decline Come From? European Journal of Political Research, 43(2), 221-236. doi: 10.1111/j.1475-6765.2004.00152.x.

• Boelhouwer, J. (2002). Achtergronden van niet-stemmers. In P. Dekker (Ed.), Niet-stemmers: Een onderzoek naar achtergronden en motieven in enquêtes, interviews en focusgroepen (pp. 21-42). Sociaal en Cultureel Planbureau.

• Boogers, M., van Ostaaijen, J. J. C. & Slagter, L. (2010). Lokale kiezers: lokale keuzes? Onderzoek naar de achtergronden en de betekenis van het stemgedrag bij de gemeenteraadsverkiezingen van 2010 (98p.) Universiteit van Tilburg.

• Bovens, M. & Wille, A. (2010). The Education Gap in Participation: A Rejoinder. Acta Politica, 47(3), 259-271. doi: 10.1057/ap.2012.11.

• Bovens, M. & Wille, A. (2011). Diplomademocratie: Over de spanning tussen meritocratie en democratie (173p.). Bert Bakker.

• Buelens, J. & Deschouwer, K. (Eds.). (1996). De dorpsstraat is de wetstraat niet (122p.). VUBPress

• Caluwaerts, D., Devillers, S., Junios, N., Matthieu, J. & Pauwels, S. (2022). Compulsory Voting: Anachronism or Avant-Garde. In D. Caluwaerts & M. Reuchamps (Eds.), Belgium Exceptionalism. Belgian Politics Between Realism and Surrealism (p. 14).

• Campbell, D. E. (2006). Why We Vote: How Schools and Communities Shape Our Civic Life (280p.). Princeton University Press.

• Cancela, J. & Geys, B. (2016). Explaining Voter Turnout: A Meta-Analysis of National and Subnational Elections. Electoral Studies, 42, 264-275. doi: 10.1016/j.electstud.2016.03.005.

• Coffé, H. & Bolzendahl, C. (2010). Same Game, Different Rules? Gender Differences in Political Participation. Sex Roles, 62(5-6), 318-333. doi: 10.1007/s11199-009-9729-y.

• Dahl, R. A. (1989). Democracy and Its Critics (397p.). Yale University Press.

• Dassonneville, R., Feitosa, F., Hooghe, M., Lau, R. & Stiers, D. (2019). Compulsory Voting Rules, Reluctant Voters and Ideological Proximity Voting. Political Behaviour, 41, 209-230. doi: 10.1007/s11109-018-9448-6.

• Dassonneville, R. & Hooghe, M. (2017). Voter Turnout Decline and Stratification: Quasi-Experimental and Comparative Evidence of a Growing Educational Gap. Politics, 37(2), 184-200. doi: 10.1177/0263395716674370.

• Dassonneville, R., Hooghe, M. & Miller, P. (2017). The Impact of Compulsory Voting on Inequality and the Quality of the Vote. West European Politics, 40(3), 621-644. doi: 10.1080/01402382.2016.1266187.

• De Greef, P. (2002). Niet-stemmers gehoord: de interviews. In P. Dekker (Ed.), Niet-stemmers: Een onderzoek naar achtergronden en motieven in enquêtes, interviews en focusgroepen (pp. 63-108). Sociaal en Cultureel Planbureau.

• Dejaeghere, Y. & Vanhoutte, B. (2016). Virtuous Villages and Sinful Cities? A Spatial Analysis into the Effects of Community Characteristics on Turnout and Blank/Invalid Voting in Local Elections in Belgium 2006–2012. Acta Politica, 51(1), 80-101. doi: 10.1057/ap.2014.38.

• den Ridder, J. & Dekker, P. (2016). Politieke participatie. In T. van der Meer & H. van der Kolk (Eds.), Democratie dichtbij: Lokaal kiezersonderzoek 2016 (pp. 30-39). Stichting Kiezersonderzoek Nederland.

• den Ridder, J., Dekker, P. & van Ditmars, M. (2012). Burgerpectieven 2012/4 (53p.). Sociaal en Cultureel Planbureau.

• Denver, D. (2008). Another Reason to Support Marriage? Turnout and the Decline of Marriage in Britain. British Journal of Politics and International Relations, 10(4), 666-680. doi: 10.1111/j.1467-856x.2008.00329.x.

• Dewachter, W. (1973). Rustige gemeenteraadsverkiezingen op 11 oktober 1970: ongelijke politieke participatie en besluitvorming. Res Publica, 15(5), 842. doi: 10.5553/RP/048647001973015005839.

• De Winter, L. (1978). Officiële en reële omvang van het absenteïsme bij de na-oorlogse parlementsverkiezingen. Res Publica, 20(4), 691. doi: 10.21825/rp.v20i4.19439.

• De Winter, L. & Wolfs, W. (2017). Policy Analysis in the Belgian Legislatures: The Marginal Role of a Structurally Weak Parliament in a Partitocracy with No Scientific and Political Tradition of Policy Analysis. In M. Brans & D. Aubin (Eds.), Policy Analysis in Belgium (pp. 129-149). Policy Press.

• Downs, A. (1957). An Economic Theory of Political Action in a Democracy. Journal of Political Economy, 65(2), 135-150. doi: 10.1086/257897.

• Fieldhouse, E., Tranmer, M. & Russell, A. (2007). Something about Young People or Something about Elections? Electoral Participation of Young People in Europe: Evidence from a Multilevel Analysis of the European Social Survey. European Journal of Political Research, 46, 797-822. doi: 10.1111/j.1475-6765.2007.00713.x.

• Franklin, M., Lyons, P. & Marsh, M. (2004). Generational Basis of Turnout Decline in Established Democracies. Acta Politica, 39, 115-151. doi: 10.1057/palgrave.ap.5500060.

• Geys, B. (2006). Explaining Voter Turnout: A Review of Aggregate-Level Research. Electoral Studies, 25(4), 637-663. doi: 10.1016/j.electstud.2005.09.002.

• Guntermann, E., Dassonneville, R. & Miller, P. (2020). Are Inequalities in Representation Lower Under Compulsory Voting? Policy Studies, 41(2-3), 151-171. doi: 10.1080/01442872.2019.1694649.

• Hakhverdian, A., van der Brug, W. & de Vries, C. (2012). The Emergence of a ‘Diploma Democracy’? The Political Education Gap in the Netherlands 1971-2010. Acta Politica, 47(3), 229-247. doi: 10.1057/ap.2011.27.

• Hoffman, M., Léon, G. & Lombardi, M. (2017). Compulsory Voting, Turnout and Government Spending: Evidence from Austria. Journal of Public Economics, 145, 103-115. doi: 10.1016/j.jpubeco.2016.10.002.

• Hooghe, M. & Pelleriaux, K. (1998). Compulsory Voting in Belgium: An Application of the Lijphart Thesis. Electoral Studies, 17(4), 419-424. doi: 10.1016/S0261-3794(98)00021-3.

• Howe, P. (2006). Political Knowledge and Electoral Participation in the Netherlands: Comparisons with the Canadian Case. International Political Science Review, 27(2), 137-166. doi: 10.1177/0192512106061424.

• Inglehart, R. & Norris, P. (2003). Rising Tide: Gender Equality and Cultural Change Around the World (244p.). Cambridge University Press.

• Irwin, G. (1974). Compulsory Voting Legislation. Impact on Voter Turnout in the Netherlands. Comparative Political Studies, 7(3), 292-315. doi: 10.1177/001041407400700302.

• Irwin, G. & Van Holsteyn, J. (2010). Nieuwe vragen, oude antwoorden. Res Publica, 52(1), 19-43. doi: 10.5553/RP/048647002010052001003.

• Jaitman, L. (2013). The Causal Effect of Compulsory Voting Laws on Turnout: Does Skill Matter? Journal of Economic Behavior & Organization, 92(C), 79-93. doi: 10.1016/j.jebo.2013.05.008.

• Jansen, G. & Boogers, M. (2018). Opkomst en stemgedrag. In G. Jansen & B. Denters (Eds.), Democratie dichterbij. Lokaal kiezersonderzoek 2018 (pp. 7-17). Stichting Kiezersonderzoek.

• Kimberlee, R. H. (2002). Why Don’t British Young People Vote at General Elections? Journal of Youth Studies, 5(1), 85-98. doi: 10.1080/13676260120111788.

• Kouba, K. & Lysek, J. (2019). What Affects Invalid Voting? A Review and Meta-analysis. Government and Opposition, 54(4), 745-775. doi: 10.1017/gov.2018.33.

• Kouba, K. & Mysicka, S. (2019). Should and Does Compulsory Voting Reduce Inequality? Sage Open, 10p. doi: 10.1177/2158244018817141.

• Kouba, K., Novak, J. & Strnad, M. (2020). Explaining Voter Turnout in Local Elections: A Global Comparative Perspective. Contemporary Politics, 27(1), 1-21. doi: 10.1080/13569775.2020.1831764.

• Kranendonk, M. & Vermeulen, F. (2019). Group Identity, Group Networks, and Political Participation: Moroccan and Turkish Immigrants in the Netherlands. Acta Politica, 54, 625-666. doi: 10.1057/s41269-018-0094-0.

• Krouwel, A., Kokx, W. & Pol, M. (2009). Onderzoek opkomstbevordering. Eindrapport (34p).

• Lefevere, J. & Van Aelst, P. (2014). First-Order, Second-Order or Third-Rate? A Comparison of Turnout in European, Local and National Elections in the Netherlands. Electoral Studies, 35, 159-170. doi: 10.1016/j.electstud.2014.06.005.

• Leon, F. & Rizzi, R. (2014). Does Forced Voting Result in Political Polarization? Public Choice, 166, 143-160. doi: 10.1007/s11127-016-0318-7.

• Lijphart, A. (1997). Unequal Participation. Democracy’s Unresolved Dilemma. American Political Science Review, 91, 1-14. doi: 10.2307/2952255.

• Lijphart, A. (2019). Stemrecht, stemplicht, opkomstplicht: inleiding tot het debat. Res Publica, 52(1), 9-17. doi: 10.5553/RP/048647002010052001002.

• Moutselos, M. (2020). Praying on Friday, Voting on Sunday? Mosque Attendance and Voter Turnout in Three West European Democracies. Journal of Ethnic and Migration Studies, 46(11), 2275-2292. doi: 10.1080/1369183X.2019.1662283.

• Plutzer, E. (2002). Becoming a Habitual Voter: Inertia, Resources, and Growth in Young Adulthood. American Political Science Review, 96(1), 41-56. doi: 10.1017/S0003055402004227.

• Power, T. J. & Garand, J. C. (2007). Determinants of Invalid Voting in Latin America. Electoral Studies, 26(2), 432-444. doi: 10.1016/j.electstud.2006.11.001.

• Przeworski, A., Stokes, S. C. & Manin, B. (Eds.). (1999). Democracy, Accountability & Representation (351p.). Cambridge University Press.

• Putnam, R. (2000). Bowling Alone. The Collapse and Revival of American Community (544p.). Simon & Shuster Ltd.

• Quintelier, E., Hooghe, M. & Marien, S. (2011). The Effect of Compulsory Voting on Turnout Stratification Patterns: A Cross-national Analysis. International Political Science Review, 32(4), 396-416. doi: 10.1177/0192512110382016.

• Reif, K. & Schmitt, H. (1980). Nine Second-order National Elections e a Conceptual Framework for the Analysis of European Election Results. European Journal of Political Research, 8(1), 3-44. doi: 10.1111/j.1475-6765.1980.tb00737.x.

• Reuchamps, M., Caluwaerts, D., De Winter, L., Jacquet, V. & Meulewaeter, C. (2015). Stemplicht en absenteïsme in een multilevel perspectief. In K. Deschouwers, P. Delwit, M. Hooghe, P. Baudewyns & S. Walgrave (Eds.), De kiezer ontcijfers: Over stemgedrag en stemmotivaties (pp. 168-184). LannooCampus.

• Reynolds, A. & Steenbergen, M. (2006). How the World Votes: The Political Consequences of Ballot Design, Innovation and Manipulation. Election Studies, 25(3), 570-598. doi: 10.1016/j.electstud.2005.06.009.

• Rosema, M. (2007). Low Turnout: Threat to Democracy or Blessing in Disguise? Consequences of Citizens’ Varying Tendencies to Vote. Electoral Studies, 36(3), 612-623. doi: 10.1016/j.electstud.2006.10.007.

• Selb, P. & Lachat, R. (2009). The More, the Better? Counterfactual Evidence on the Effect of Compulsory Voting on the Consistency of Party Choice. European Journal of Political Research, 48(5), 573-597. doi: 10.1111/j.1475-6765.2009.01834.x.

• Sheppard, J. (2015). Compulsory Voting and Political Knowledge: Testing a “Compelled Engagement” Hypothesis. Electoral Studies, 40, 300-307. doi: 10.1016/j.electstud.2015.10.005.

• Shineman, V. A. (2018). If You Mobilize Them, They Will Become Informed: Experimental Evidence that Information Acquisition is Endogenous to Costs and Incentives to Participate. British Journal of Political Science, 48(1), 189-211. doi: 10.1017/S0007123416000168.

• Singh, S. (2011). How Compelling is Compulsory Voting? A Multilevel Analysis of Turnout. Political Behaviour, No. 33, 95-111. doi: 10.1007/s11109-010-9107-z.

• Singh, S. (2015). Compulsory Voting and the Turnout Decision Calculus. Political Studies, 3, 548-568. doi: 10.1111/1467-9248.12117.

• Singh, S. (2019). Politically Unengaged, Distrusting, and Disaffected Individuals Drive the Link Between Compulsory Voting and Invalid Balloting. Political Science Research and Methods, 7(1), 107-123. doi: 10.1017/psrm.2017.11.

• Smets, K. & van Ham, C. (2013). The Embarrassment of Riches? A Meta-Analysis of Individual-Level Research on Voter Turnout. Electoral Studies, 1-16. doi: 10.1016/j.electstud.2012.12.006.

• Steenvoorden, E. & Van der Waal, J. (2016). Stemgedrag bij gemeenteraadsverkiezingen. In T. van der Meer & H. van der Kolk (Eds.), Democratie dichterbij: Lokaal Kiezersonderzoek 2016 (pp. 56-63). Stichting KiezersOnderzoek Nederland.

• Thomassen, J. & van Ham, C. (2014). Failing Political Representation or a Change in Kind? Models of Representation and Empirical Trends in Europe. West European Politics, 37(2), 400-419. doi: 10.1080/01402382.2014.887881.

• Topf, R. (1995). Beyond Electoral Participation. In H.-D. Klingemann & D. Fuchs (Eds.), Citizens and the State (pp. 27-51). Oxford University Press.

• van der Meer, T. & van Deth, J. (2010). Opkomstplicht: stimulans of frustratie. Res Publica, 52(1), 73-93. doi: 10.5553/RP/048647002010052001005.

• van Ostaaijen, J. J. C., Epskamp, M. & Dols, M. (2016). Verbetering op komst (113p.). Tilburg University.

• van Ostaaijen, J. J. C., van Zuydam, S. & Epskamp, M. (2019). How to Improve Local Turnout. The Effect of Municipal Efforts to Improve Turnout in Dutch Local Elections. Politics of the Low Countries, 1(3), 160-181. doi: 10.5553/PLC/258999292019001003002.

• Verba, S. & Nie, N. (1972). Participation in America: Political Democracy and Social Equality (452p.). Harper & Row.

• Verba, S., Schlozman, K. & Brady, H. (1995). Voice and Equality: Civic Voluntarism in American Politics (664p.). Harvard University Press.

• Verlet, D., Carton, A. & Callens, M. (2010). Opkomstplicht in Vlaanderen: een gespreide slagorde? Onderzoek naar de gelaagdheid van houdingen ten aanzien van de opkomstplicht. Res Publica, 52(1), 45-71. doi: 10.5553/RP/048647002010052001004.

• Wauters, B. & Devroe, R. (2017). Forced to Vote, But Not for Women. The Effect of Compulsory Voting on Voting for Women. Acta Politica, 53, 469-487. doi: 10.1057/s41269-017-0065-x.

Appendix
Table 1 Overview variables explaining voter turnout – Belgium

	Ackaert and De Winter (1993)	Hooghe and Pelleriaux (1998)	Billiet (2002)	Verlet et al. (2010)	Reuchamps et al. (2015)
Resource model
Income	ns			ns	+
Occupational status	ns			ns
Occupational type	+	+
Age	ns	ns		ns (+ for regional and federal levels)	ns (+ only for municipal level)
Male	ns	+	ns	+	ns
Education	ns	+	+	+	ns (+ only for the federal level)
Marital status	ns
Mobilisation model
Attendance of religious services	+
Identification with religion		ns
Associational membership			ns	+
Engagement volunteering work				+
Social isolation	ns
Political campaigns and news exposure	+
Rational choice model
Utilitarian individualistic attitude				−
Perceived policy impact					+
Socialisation model
Regular social contacts
Political conversations
Psychological model
Political interest	+		+		+
Political knowledge	+		+		ns
External political efficacy	+		+	+	+
Internal political efficacy					ns
Party identification					+
Political cynicism	−
Satisfaction with democracy				ns	ns
Trust				ns	ns
Health and well-being

+= significant positive correlation; − = significant negative correlation; ns = non-significant result; [empty] = not investigated.

Table 2 Overview variables explaining voter turnout – the Netherlands

	Irwin (1974)	Aarts (2001)	Boelhouwer (2002)	Howe (2006)	Lefevere and Van Aelst (2014)	den Ridder and Dekker (2016)	Steenvoorden and Van der Waal (2016)	Jansen and Denters (2018)
	Resource model
Income	+	+	+		+		ns	+
Occupational status			+
Occupational type
Age	Curvilinear	+	Curvilinear	+	ns	+	+	Curvilinear
Male	+	+	ns		+	+	ns	ns
Education	+	+	+		+	+	+	+
Marital status			+
	Mobilisation model
Attendance of religious services			+
Identification with religion	+			+
Associational membership			+
Engagement volunteering work			+
Social isolation			−
Political campaigns and news exposure	+				+		+
	Rational choice model
Utilitarian individualistic attitude
Perceived policy impact					+			+
	Socialisation model
Regular social contacts			+
Political conversations			+
	Psychological model
Political interest	+		+		+	+	+
Political knowledge		+		+			+
External political efficacy	+					+	+
Internal political efficacy						+	ns
Party identification	+	+	+	+
Political cynicism		−
Satisfaction with democracy		+	+			+
Trust		+				+	+
Health and well-being			+

+ = significant positive correlation; − = significant negative correlation; ns = non-significant result; [empty] = not investigated.

Noten

• 1 The original aim of the article was to include data and literature on Belgium, the Netherlands and the Grand Duchy of Luxembourg. Our search on the official website of the elections in the Grand Duchy of Luxembourg (elections.public.lu) revealed, however, that election results (including absenteeism rates) are not available over time, either for parliamentary or for municipal elections. In addition, our search for English-language literature on compulsory voting and turnout in the Grand Duchy of Luxembourg yielded hardly any relevant contributions. We therefore decided not to include the Grand Duchy of Luxembourg in our analysis.

• 2 www.scp.nl.

• 3 www.dpes.nl.

• 4 VOORSTEL VAN DECREET – van de heer André-Emiel Bogaert – houdende wijziging van de wet van 19 oktober 1921 tot regeling van de provincieraadsverkiezingen, met het oog op de afschaffing van de stemplicht.

• 5 ONTWERP VAN DECREET tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie.

• 6 790 (2020-2021) – Nr. 7 23 juni 2021 (2020-2021) Verslag namens de Commissie voor Binnenlands Bestuur, Gelijke Kansen en Inburgering uitgebracht door Tom Ongena en Yves Buysse over het ontwerp van decreet tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie.

• 7 790 (2020-2021) – Nr. 6 ingediend op 16 juni 2021 (2020-2021) Verslag van de hoorzitting namens de Commissie voor Binnenlands Bestuur, Gelijke Kansen en Inburgering uitgebracht door Tom Ongena en Yves Buysse over het ontwerp van decreet tot wijziging van diverse decreten, wat betreft versterking van de lokale democratie.

• 8 For Belgium, only data from 1976 onwards provides a valid picture of the evolution of absenteeism. Until then, the members of the college of mayor and aldermen reviewed the list of voters in an election year between 1 July and 20 September. Because of the large time frame between the closing of the voter register and the election date, it regularly happened that voters who passed away during this time span, were considered absent in the elections and included in the official abstention numbers. Following the introduction of the elections law of 5 July 1976, the registers of voters are permanently kept up to date by the college of mayor aldermen. As such, the time span between the establishment of the voters’ list and the election date has been shortened (De Winter, 1978). Furthermore, until 1970 it regularly happened that no elections were organised in a significant number of municipalities owing to a lack of competition. In 1970 it concerned 12.6% of municipalities, in 1964 even 15.5% (Dewachter, 1973). This also hinders the mapping of reliable data regarding absenteeism at Belgian local elections before 1976. Therefore, only the figures from 1976 onwards to evaluate electoral absenteeism in Belgium are presented.

• 9 The political institutional model is not considered in the remainder of the article. The political institutional model considers the decision to vote to be related by the political and institutional context. In the third paragraph we already elaborated on the most striking institutional difference between Belgium and the Netherlands, namely the (non-) compulsory voting system and its impact on voter turnout.

• 10 For the sake of parsimony and clarity, only quantitative studies that contained at least three variables explaining voter turnout were included in the tables.

• 11 During the most recent federal elections the total number of blank and invalid votes was 6%. For the regional elections, 5% of the Flemish voters voted blank or invalid, 8.3% in Wallonia and 6.7% in Brussels. In 2018, 3.8% of the voters voted blank or invalid during the Flemish municipal elections.

• 12 www.verkiezingsuitslagen.nl.