ALGEMENE MEDEDELING

In de loop van januari 2025 wordt deze online omgeving geïntegreerd in Boomportaal (www.boomportaal.nl), waarna deze omgeving wordt opgeheven. Vanaf dat moment linkt deze URL automatisch door naar Boomportaal.

DOI: 10.5553/PLC/258999292019001003002

Politics of the Low CountriesAccess_open

Article

How to Improve Local Turnout

The Effect of Municipal Efforts to Improve Turnout in Dutch Local Elections

Trefwoorden turnout, local elections, get out the vote, campaign, the Netherlands
Auteurs
DOI
Toon PDF Toon volledige grootte
Samenvatting Auteursinformatie Statistiek Citeerwijze
Dit artikel is keer geraadpleegd.
Dit artikel is 0 keer gedownload.
Aanbevolen citeerwijze bij dit artikel
Julien van Ostaaijen, Sabine van Zuydam en Martijn Epskamp, "How to Improve Local Turnout", Politics of the Low Countries, 3, (2019):160-181

    Even though many municipalities use a variety of means to improve turnout in local elections, citizen participation in local elections is a point of concern in many Western countries, including the Netherlands. Our research question is therefore: How effective are municipal efforts to improve turnout in (Dutch) local elections? To this end, we collected data from three sources: (1) a survey sent to the municipal clerks of 389 Dutch municipalities to learn what they do to improve turnout; (2) data from Statistics Netherlands on municipalities’ socio-demographic characteristics; and (3) data on the turnout in local elections from the Dutch Electoral Council database. Using hierarchical multiple regression analysis, we found that the direct impact of local governments’ efforts to improve turnout is low. Nevertheless, some measures seem to be able to make a difference. The relative number of polling stations was especially found to impact turnout.

Dit artikel wordt geciteerd in

    • 1 Concern About Low Local Turnout and How to Tackle It

      In many Western countries, citizen participation in local elections is a point of concern (Loughlin, Hendriks, & Lidström, 2011). This is not only because the turnout in absolute numbers can be low – as is, for example, the case in the U.S. where often no more than 15% cast their vote – but also because it has declined substantially over the last decades. In the Netherlands, the 2018 local elections attracted 55% of all eligible voters to the ballot box, whereas approximately 30 years earlier, turnout was well above 70% (Hendriks, Van der Krieken, Van Zuydam, & Roelands, 2015). This decline has led to questions about the functioning of local democracy. The assumption is, after all, that low electoral turnout is bad for democracy, as it can affect the legitimacy of local governments and reduce the representativeness of the elected candidates (Franklin, 1999). This leads to questions regarding how turnout can be improved.
      The reasons why people decide to vote in (local) elections have attracted considerable scholarly attention. A particular part of that literature deals with the activities that can increase voter turnout (e.g. Green & Gerber, 2008). In the Netherlands, (local) governments seem increasingly motivated to stimulate people to cast their vote. In 2014, about three-quarters of Dutch municipalities employed extra communication efforts around the local elections, most of them with the concrete aim of increasing voter turnout (Van Ostaaijen, Epskamp, & Dols, 2016). The choice of these efforts is, however, generally not based on evidence-based campaign plans. Only 15% of all municipalities based their selection of activities on research, and only 8% of all Dutch municipalities evaluated their efforts afterwards (Van Ostaaijen et al., 2016). The selections in turn were based on the wishes of local politicians, or on ‘habit’. Consequently, the effectiveness of these municipal campaigns to increase voter turnout remains largely unknown. We therefore seek an answer to the question: How effective are municipal efforts in improving turnout in (Dutch) local elections?
      We answer this question using hierarchical multiple regression analysis on quantitative data from three sources: (1) a survey sent to the municipal clerks of 389 Dutch municipalities to learn what they do to improve turnout; (2) data from Statistics Netherlands on socio-demographic characteristics of municipalities; and (3) data on the turnout in local elections from the Dutch Electoral Council database. In doing so, we aim to contribute both to society and to the literature on voter turnout. We contribute to society in that we show which efforts are undertaken by municipalities and which efforts appear to actually be successful in attracting voters to the polling stations. There are many different things municipalities (can) do and municipalities have spent up to 600,000 Euros on such campaigns (Binnenlands Bestuur, 2014). We do not only want to show what municipalities do with that money, but also whether it has an effect on voter turnout. If these activities do not contribute to the goal of increasing turnout, one can debate whether it is worth the taxpayer’s money.
      Apart from contributing to society, the academic relevance of the paper is threefold. First, we evaluate the extent to which the efforts of municipalities to motivate citizens to cast their vote have an effect. For this, we focus on the Dutch case for several reasons. In relation to, for instance, the U.S., from which a lot of literature regarding (improving) turnout originates, there is no obligation for voters to register in order to participate. This low threshold, in theory, makes the potential effect of municipalities’ efforts on voter turnout more direct. The Netherlands is furthermore an interesting case as in the last decades turnout in local elections has decreased considerably, whereas at the same time a majority of municipalities engaged in activities to increase voter turnout (Van Ostaaijen et al., 2016). As such, this study contributes to international literature regarding the effectiveness of voter turnout campaigning. Second, while many studies take the perspective of political parties, our focus is on the more neutral actor of government. Research shows that it is difficult to increase turnout with campaigning, but when instigated by a ‘neutral’ actor this can, in some cases, increase the effect (Green & Gerber, 2015). Third, while most studies focus on national elections (Smets & Van Ham, 2013), this study focuses on local elections. Local elections often face lower turnout numbers compared to national elections (Loughlin et al., 2011), making an assessment of what can improve turnout on that level all the more important. However, what works at the local level can differ from national (or presidential) elections (Cancela & Geys, 2016), which requires research exclusively focused on this level.
      In the next section, we explore in more detail the literature underlying our research. Next, our research methods are outlined, followed by our results. After answering our stated research question, this article concludes with a discussion of the implications of our findings for literature and society.

    • 2 Improving Turnout in Municipal Elections: Lessons from Literature

      Campaigns for getting people to vote take time and money (Green & Gerber, 2008), and their effectiveness is often bounded by time and place (Cancela & Geys, 2016). Most of the related literature and research is situated in the U.S. or, at least, does not specifically deal with countries similar to the Netherlands (Loughlin et al., 2011). Two relevant sets of activities can be distinguished in this respect. A first set of activities relates to communication efforts of municipalities (Green & Gerber, 2008; Smets & Van Ham, 2013). A second set of activities is aimed at lowering the burden for citizens to go out and vote (Brady & McNulty, 2011; SCP, 2012, p. 39).
      When we look at instruments that are (somewhat) effective, confirmed in international as well as the available (though limited) Dutch research, the following stand out. Regarding communication instruments, sending letters seems somewhat effective (Green & Gerber, 2008, 2015; Van Ostaaijen et al., 2016), especially when it also expresses gratitude for a previous vote (Panagopoulos, 2011). Next to that, voting advice applications seem to have a positive effect on turnout (Ruusuvirta & Rosema, 2009), especially on young people (Gemenis & Rosema, 2014; Ladner & Pianzola, 2010). Voting advice applications are also able to contribute to the enlargement of political knowledge about, and interest in, participating political parties (Kruikemeier, 2014; Neijens & De Vreese, 2010). For other activities, mainly U.S. or non-Dutch literature is available, which shows, for instance, that organizing a festival on election day and canvassing can have positive effects (Green & Green, 2008). These festivals can draw extra attention to the election (Howard & Posler, 2012), which can be extra relevant when it concerns a local or second-order election. Reminding people to vote can also have a positive effect. In an experiment, volunteers held up signs in the street to remind people to vote, which led to a 3.48% increase in voter turnout (Panagopoulos, 2009). In other research, SMS messaging made it 3% more likely that people would vote (Dale & Strauss, 2009; Malhotra, Michelson, Rogers & Valenzuela, 2011). Leaflets, e-mail, TV and radio, on the contrary, have only limited effect. However, because of the low costs involved for some of these activities, they should not be excluded from the voter turnout repertoire completely. Green and Gerber (2008) have argued that because of the relative ease with which these instruments can be implemented, even a very low success rate can still be worth the effort (Table 1).

      Table 1 Communication instruments contributing to turnout in local elections
      Communication instruments
      Sending letterse.g. (Green & Gerber 2008; Panagopoulos, 2011)
      Voting advice applicationse.g. (Gemenis & Rosema, 2014; Ladner & Pianzola, 2010; Ruusuvirta & Rosema, 2009)
      Organizing a festival on election daye.g. (Green & Gerber, 2008; Howard & Posler, 2012)
      Canvassinge.g. (Green & Gerber, 2008)
      Reminding people to votee.g. (Dale & Strauss, 2009; Malhotra et al., 2011 ; Panagopoulos, 2009)
      Leafletse.g. (Green & Gerber, 2008)
      E-maile.g. (Green & Gerber, 2008)
      TV and radio commercialse.g. (Green & Gerber, 2008)

      Apart from communication instruments, municipalities can also take measures to ease the burden of voting. The voting facilities municipalities can use include polling stations in central places (e.g., the train station) where many people pass by anyway, a mobile polling station that comes very close to residential areas, extended opening hours or increasing the number of polling stations to make sure the distance voters have to travel to the polling station is as short as possible (Brady & McNulty, 2011; Jacobs & Van Ostaaijen, 2017). Research on the relevance of these factors for turnout is inconclusive. While some studies, for example, find a positive effect of the number of polling stations on turnout (Jacobs, Van Zuydam, Van Ostaaijen, & de Brouwer, 2018), others do not (Dyck & Gimpel, 2015). This same finding also applies to facilities like extended opening hours (Aarts, 1999 in SCP, 2002, p. 152).
      Campaign activities, however, do not take place in a vacuum. Considering voter turnout in general, campaigning seems only to have a marginal influence or is not mentioned at all when talking about the most successful factors related to (local) turnout (Cancela & Geys, 2016; Smets & Van Ham, 2013). Therefore, to understand turnout and how it can be affected, it is important to take into account the factors that affect the decision of citizens to go out and vote. This is a difficult task, as in the literature over 170 factors have been identified as potentially affecting turnout (Smets & Van Ham, 2013). Moreover, studies regarding voter turnout are differentiated, meaning that most related studies only deal with a limited number of factors – of the 170 factors, only 8% were included in more than a quarter of the 90 studies analysed (Smets & Van Ham, 2013).
      Nevertheless, some factors stand out more than others. In this respect, a distinction can be made between individual factors, contextual and institutional factors and factors related to the local political situation and the upcoming elections. Several international meta-evaluations were helpful in identifying the factors relevant to local voter turnout in the Netherlands (Cancela & Geys, 2016; Geys, 2006; Harder & Krosnick, 2008; Smets & Van Ham, 2013), as well as some Dutch studies that focused on (non-)voting (Aarts, 1999; Boogers, Ostaaijen, & Slagter, 2011; SCP, 2002, 2012; Van der Meer & Van der Kolk, 2016; Wille, 2000) and case studies that focused on a limited set of factors related to turnout (Appendix A).
      With regard to individual factors, literature shows the most positive results for age, education, political factors (knowledge of politics, trust in politics and satisfaction with politics), lifestyles, habits and practical limitations (Boogers et al., 2011; Cancela & Geys, 2016; Smets & Van Ham, 2013; Smets & Van Ham, 2013). With regard to contextual factors, municipal size, local and regional connectedness of inhabitants, choice of election day and proximity to other election days, voting obligation and electoral competition are emphasized in the literature as contributing most to influencing turnout (Cancela & Geys, 2016; CPB, 2014; Blais & Dobrzynska, 1998; Henderson & McEwen, 2010; Panagopoulos, 2011;). With regard to factors related to the local political situation and the upcoming elections, we found studies showing the positive effects of electoral competition and weather conditions (Blais & Aarts, 2006; Persson, Sundell, & Öhrvall, 2014). However, for many factors, few or no studies were found. Nevertheless, based on a limited number of studies, ethnicity, religion, happiness and health, social pressure and media attention seemed to have positive effects, and to a lesser extent this proved also the case for gender, income/employment and whether people perceive local government as relevant (Cancela & Geys, 2016; Smets & Van Ham, 2013). Last, for some factors, relevance was not clear because studies pointed in difference directions. This was especially the case for fragmentation and positive or negative campaigning (Eggers, 2015; Geys, 2006; Górecki, 2013). For many factors however, we did not find enough relevant literature to arrive at a conclusion (see Appendix A). For instance, with regard to municipal stability, media presence and the connection people feel towards their municipality, we did not find enough (convincing) research to say something about their effects on local turnout, in general, or within the Dutch local context, in particular.

    • 3 Research Approach

      Sample/data: Three sets of data are used to describe the efforts of municipalities to increase voter turnout in municipal elections and to explain what accounts for differences between municipalities in turnout: (1) data on the turnout in local elections, (2) data on the efforts of municipalities to improve turnout, and (3) data on other factors – predominantly socio-demographic factors – that impact turnout. To this end, data was collected from three sources. All data was collected in 2015, focusing on the latest local elections at the time (2014). In some parts of the analysis, we compared the turnout in 2014 to the turnout in 2010.
      For the data on turnout in local elections and socio-demographic factors impacting turnout, we relied on publicly accessible databases. Data on the turnout in the 2014 local elections in all Dutch municipalities was gathered from the database of the Dutch Electoral Council, whereas we used data on socio-demographic factors from Statistics Netherlands (CBS). As with the other data, 2014 was our reference year.
      Data on the efforts of municipalities to improve turnout (through communication or by lowering the burden to vote) were collected via a survey sent to the municipal clerks of 389 of all 393 Dutch municipalities. In the summer of 2015, the 33 largest municipalities of the Netherlands were asked to participate in the survey (more than 100,000 inhabitants). Of these 33 municipalities, 31 municipalities/municipal clerks participated. Next, in December 2015, the survey was also sent to the other 356 municipalities in the Netherlands. Of these smaller municipalities, 241 municipalities responded. Combined, this resulted in a response rate of 70% (272 municipalities). In the survey, primarily the 2014 local elections were of interest, but we also asked municipal clerks to provide information on the communication efforts of their municipalities in the 2006 and 2010 local elections.

      Measurement instruments: In this study, municipalities are the units of analysis (instead of, for example, individuals). This means that we did not study whether communication efforts and measurements to lower the burden to vote had an impact on the chance that individual citizens would vote, but we studied what might explain differences in electoral turnout between municipalities.
      The dependent variable in this study is the turnout in the 2014 local elections, whereas municipal communication efforts and means to lower the burden to vote were the independent variables of interest. Socio-demographic factors that contribute to explaining why citizens go to the ballot box were included in the analysis as control variables.
      The turnout in the 2010 and 2014 local elections was measured as the percentage of eligible voters in a municipality who cast their vote.1xIn case of municipal reorganization, we included the turnout in the local elections held between 2012 and 2015 in the analysis. After all, for these municipalities, no local elections were organized in 2014. With regard to the 2010 election turnout, we included the turnout in elections following municipal reorganization between 2008 and 2011. Inspecting the distribution of turnout scores in the data suggests that it is reasonably normally distributed. The Kolmogorov-Smirnov statistic confirms this: the statistic was not statistically significant for the data on voter turnout (D(385) = 0.041, p = 0.142), meaning that the assumption of normality is not violated. In 2014, the average turnout in local elections was 56.8% (std dev = 6.9%, N = 389).
      Communication efforts of municipalities were operationalized in 12 distinct ways in which municipalities could draw attention to the upcoming elections: (1) approaching people on the street; (2) posters, pamphlets etc. on the street; (3) commercials on radio or TV; 4) personal letters to citizens (e.g., from the mayor); (5) advertisements in newspapers; (6) Election newspaper/newsletter delivered to people’s homes; (7) voting advice application (Kieskompas/Stemwijzer/DeStemVan); (8) publicly accessible debates; (9) election fairs; (10) (municipal) website; (11) social media and (12) online videos. For each of these communication means, municipal clerks were asked to indicate whether it had been used in their municipality in the period leading up to the 2014 local elections, but also whether they were used in the 2006 and 2010 local elections. In addition, municipal clerks were given the possibility of adding three additional means of communication. In the first part of the results section we provide descriptive statistics on the prevalence of each means of communication in the 2006, 2010 and 2014 municipal elections.
      The means that municipalities can employ to lower the burden to vote were operationalized in nine facilities: (1) polling station in central places (e.g., train station); (2) special transportation to polling station (among others for specific groups); (3) mobile polling station (e.g., voting bus); (4) special opening hours polling stations (e.g., at night); (5) an election party where people could also vote; (6) signposts pointing towards polling stations; (7) clear marking of polling stations; (8) a cup of coffee or other drinks/snacks for voters at the polling station; (9) a (small) reward for voters. In addition, the number of polling stations per 1,000 inhabitants can be considered a means to facilitate voting. After all, one could argue that a higher number of polling stations means that the distance voters need to travel to the ballot box decreases, making it easier for them to cast their vote. Moreover, municipal clerks could list up to three additional voting facilities. For each of these facilities, it was asked whether their municipality had used it in the 2014 local elections.
      The control variables included in this study are the socio-demographic factors that, according to literature, contribute most to explaining turnout in elections: age, level of education, income, unemployment, ethnicity, degree of religiosity, municipality size and municipal density. A multicollinearity check showed that municipal size, municipal density and unemployment correlated highly with the other factors, in particular, with the ethnicity of voters. Therefore, we opted to include the following factors in the analysis, operationalized as follows:

      1. Age: the percentage of inhabitants aged 65 years and older

      2. Level of education: the percentage of inhabitants who completed higher vocational education or university among inhabitants aged 15-75 years

      3. Ethnicity of voters: the percentage of inhabitants with a non-Western immigration background

      4. Degree of religiosity: the percentage of inhabitants that attend a religious service at least once a month

      5. Income: the average standardized income of private households

      A final control variable was whether the municipal elections were the consequence of municipal reorganization. In municipal reorganization elections, turnout tends to be substantially lower than in ‘regular’ municipal elections (cf. database Electoral Council).2xThe database of the Dutch Electoral Council can be accessed via: https://www.verkiezingsuitslagen.nl/. It includes the results and turnout in all Dutch elections starting from 1848.

      Statistical techniques: To answer our main research question, we primarily used hierarchical multiple linear regression analysis. This type of analysis is also referred to as sequential regression analysis, and is among other techniques used to assess the unique explanatory relevance of (a set of) variables over and above variables that were found to be important in previous research to explain our dependent variable of interest. This also has consequences for the sequence in which variables are included in the regression model. Although it may seem counter-intuitive compared to standard regression analysis, it is preferred in this type of analysis to enter the control variables first (Pallant, 2013; Warner, 2013). After all, of interest is whether additional factors help to explain the dependent variable better than existing models. In this study we are interested in how effective municipal efforts are to increase voter turnout. Previous research, however, has already established a number of important (socio-demographic) factors that contribute to explaining voter turnout. Hence, a hierarchical design is warranted to assess whether the use of communication means and voting facilities can uniquely explain differences in voter turnout.
      Consequently, the standard procedure in hierarchical multiple linear regression analysis is to first add the control variables to the model (Field, 2013; Pallant, 2013; also compare other research that uses this analysis strategy: Dubois & Blank, 2018; Kemmelmeier, 2008; Sheafer, 2008). In this study, these are the socio-demographic variables and municipal reorganization. Next, the independent variables of interest are included in the regression models. That way it can be assessed whether and to what extent the variables of interest combined and individually contributed to explaining the dependent variable, after accounting for other variables that matter, as shown in previous research. In this study, two variants of the model were considered. One model included the communication efforts of municipalities, in addition to the control variables. In the other, the facilities provided by municipalities to lower the burden of voting were analysed, while controlling for the socio-demographic factors and municipal reorganization. As a final step, we compared the contribution of communication and voting facilities to explain turnout in local elections, to evaluate the extent to which municipal efforts are effective.

      Table 2 Prevalence of means of communication and voting facilities to increase turnout
      200620102014
      Overall100,000 + municipalities< 100,000 municipalities
      Digital media (websites, social media, online videos)29%58%74%87%73%
      Advertisements in newspapers45%61%65%68%65%
      Publicly accessible debates or election fairs38%57%59%68%60%
      Personal letters to citizens (e.g., from the mayor)32%52%55%71%53%
      Election newspaper/newsletter delivered to people’s homes29%46%45%61%43%
      Posters, pamphlets, etc. on the street29%40%42%68%39%
      Voting advice application (Kieskompas/Stemwijzer/DeStemVan)14%30%29%97%20%
      No communication means employed19%17%17%0%20%
      Commercials on radio or TV4%15%17%36%14%
      Other communication means5%9%17%48%13%
      Approaching people on the street6%11%11%10%12%
      No additional voting facilities48%13%52%
      Signposts pointing towards polling stations35%56%31%
      Polling station in central places (e.g., train station)19%39%13%
      Mobile polling stations (e.g., voting bus)15%31%13%
      Clear marking of polling stations12%16%11%
      Special opening hours of polling stations (e.g., at night)6%28%3%
      Other voting facilities4%6%4%
      Special transport to polling station (among others for specific groups)3%9%3%
      A cup of coffee or other drinks/snacks for voters at the polling station3%3%3%
      An election party where people could also vote2%13%0%
      A (small) reward for voters1%3%1%
      Average Std dev
      Number of polling stations per 1,000 inhabitants0.6010.144

      3.1 Results: Efforts of Dutch Municipalities to Improve Turnout

      The efforts of municipalities to improve turnout at local elections can be categorized into two distinct types of measures. On the one hand, municipalities can invest in communication towards citizens that highlights the importance of voting. On the other hand, municipalities take up activities to increase the ease with which citizens can cast their vote. While the communication efforts of municipalities are of importance predominantly in the period leading up to the elections, the activities to lower the burden of voting take place on election day.
      With regard to the communication of municipalities to its inhabitants, a pluriform practice exists in the Netherlands. Municipalities use a multitude of media and communication means to convey the message that elections will soon take place and that voting is important. In the survey, we presented no less than ten means of communication to the municipalities. The results of the survey nevertheless suggest that some types of communication are more common than others (see Table 2). Among the more popular means of communication are digital media, advertisements in newspapers, personalized letters to citizens (for example from the mayor) and publicly accessible debates or election fairs. Half to three-quarters of all municipalities in the Netherlands used these communication instruments in the 2014 municipal elections. The use of posters, pamphlets and election newspapers or letters is also not uncommon, but their use is less widespread. The use of commercials on radio and TV and approaching people on the street is the least widespread. Finally, voting advice applications (usually bought by the municipality from one of the three large Dutch suppliers at the time) are also not standard practice in Dutch local government.
      The data available for the last three municipal elections (2006, 2010 and 2014) most notably suggest that municipalities have started communicating about upcoming local elections in more diverse ways. In each of the consecutive elections, more and more municipalities professed to have engaged in various communication efforts. Hence, it is not the case that an increase in one means of communication results in a reduction of other communication instruments. In contrast, it seems that municipalities are stepping up their game and adding instruments to their communication efforts. We should, however, treat this finding with some caution. At the time of the 2006 and 2010 municipal elections, no research was done on the communication instruments municipalities employed to increase turnout. Therefore, we asked respondents in 2014 to recall which instruments they had used in the previous two elections. It could be that the memories of respondents were not fully accurate. Despite this limitation, we feel these findings do provide a useful indication of the use of communication instruments, not in the least because respondents indicate that habit was an important factor in deciding which communication instruments to employ. This suggests that in preparing for the 2014 municipal elections, municipalities built on the communication plans from previous elections.
      Finally, there are some differences between the 33 municipalities in the Netherlands, with more than 100,000 inhabitants at the time, and the other, smaller municipalities. Voting advice applications are, for example, more common in the 100,000+ municipalities than in the smaller ones, and also posters, radio and television commercials, personal letters, election newspapers/letters and digital media are more often used.
      Besides communication to citizens in the period leading up to the municipal elections, it is also possible to lower the burden of voting in order to promote higher turnout. This is nevertheless a much less common way to try to increase turnout – overall, about half of the Dutch municipalities do not use additional voting facilities. If municipalities do invest in this, banners and large signs clearly pointing towards polling stations are most common (35%), followed by opening polling stations in central places (19%). Other facilities that have been used regularly are mobile polling stations like a voting bus (15%) and clear markings towards polling stations (12%).
      Regarding the organization of the campaign, most municipalities that employ additional communication efforts have based these efforts on wishes or orders from local politicians (85%) and/or habit, the latter meaning that they employed it because they previously did as well (63%). Only 15% of municipalities (partly) based their choices on knowledge or research. About half of the municipalities did not formulate an explicit goal for the communication efforts. Only 8% of the municipalities with additional communication efforts for the 2014 municipal elections say they undertook research to see whether their actions had any effect. However, a closer look at the results shows that many of these activities can hardly be called research at all, for instance an oral evaluation of the campaign or looking at how many people made use of a voting advice application.

      3.2 Results: The Effect of Municipal Efforts and the Factors That Matter Most for Local Turnout

      We have seen what the Dutch municipalities do in terms of communication and voting facilities to improve local turnout. Now the question arises whether these measures seem to have any effect. To determine this, we combined the data from Statistics Netherlands (CBS) and from our survey with the turnout data of all Dutch municipalities. With the use of hierarchical multiple linear regression analysis, the part of the variance in local turnout attributable to local governments’ communication efforts and voting facilitation is teased out, while taking into account the explanatory value of socio-demographic factors and municipal reorganization.

      Step 1 (control variables): Adding the socio-demographic factors and the factor of municipal reorganization to the regression model as a first step shows that these factors account for a large part of the variance in local turnout. For the 2014 municipal elections, no less than 69% (R2) of the variance in local turnout between municipalities can be explained by the control variables (see Table 3; F (6, 261) = 96,619, p < 0.001). Age, education, income and attending religious services relate positively to voter turnout in local elections. In contrast, a higher percentage of inhabitants with a non-Western immigrant background seems to result in a lower local turnout, as well as having an election due to municipal reorganizations.

      Table 3 Standard multiple regression analysis relating turnout in 2014 local elections to socio-demographic variables
      BSE Bβsr2
      (Constant)27.1962.514
      % inhabitants aged 65 or older0.2940.0720.1430.014***
      % highly educated citizens (HBO or university)0.2560.0380.2630.037***
      Household income (€)0.0010.0000.1900.020***
      % non-Western immigrants in municipality-0.4440.044-0.3500.085***
      % inhabitants visiting religious services at least once a month0.3700.0190.6220.322***
      Municipal reorganization elections around 2014-8.5151.201-0.2060.042***

      * = p < 0.05; ** = p < 0.01; *** = p < 0.001

      Table 4 Multiple regression analysis relating turnout in local elections to the municipal communication effort for the elections
      BSE Bβsr2
      (Constant)29.3303.884
      % non-Western immigrants in municipality-0.4050.070-0.3200.056***
      % highly educated citizens (HBO or university)0.2760.0600.2840.034***
      % inhabitants aged 65 or older0.2370.1070.1150.008*
      Household income (€)0.0010.0000.1870.016**
      % inhabitants visiting religious services at least once a month0.3630.0280.6090.278***
      Municipal reorganization elections around 2014-7.5051.798-0.1810.029***
      Approaching people on the street0.4040.9260.0200.000
      Posters, pamphlets, etc. on the street-0.3370.621-0.0240.000
      Commercials on radio or TV-0.8520.793-0.0500.002
      Personal letters to citizens (e.g., from the mayor)-0.3160.629-0.0220.000
      Advertisements in newspapers-1.6450.788-0.0980.007*
      Election newspaper/newsletter delivered to people’s home0.3100.6070.0220.000
      Voting advice application (Kieskompas/Stemwijzer/DeStemVan)-1.0720.795-0.0740.003
      Publicly accessible debates0.9000.6700.0620.003
      Election fairs-0.6810.689-0.0460.002
      Social media-0.2280.758-0.0150.000
      Website (municipality or special election website)0.4670.8320.0260.001
      Online videos (e.g., YouTube)0.4640.7910.0250.001

      * = p < 0.05; ** = p < 0.01; *** = p < 0.001

      Table 5 Comparing municipal communication effort in 2010 versus 2014
      BSE Bβsr2
      (Constant)-5.4872.193*
      % non-Western immigrants in municipality-0.0460.038-0.0660.003
      % highly educated citizens (HBO or university)0.0560.0340.1040.006
      % inhabitants aged 65 or older0.0840.0620.0740.004
      Household income (€)0.0000.0000.0490.001
      % inhabitants visiting religious services at least once a month0.0510.0160.1570.020**
      Municipal reorganization elections around 2014-6.8821.090-0.3020.081***
      Municipal reorganization elections around 20108.4630.8830.4440.186***
      Communication instrument employed in 2014 (but not in 2010)
      Approaching people on the street0.7801.3680.0310.001
      Posters, pamphlets, etc. on the street0.7980.8280.0500.002
      Commercials on radio or TV-0.4500.949-0.0240.000
      Personal letters to citizens (e.g., from the mayor)-1.4460.776-0.0960.007
      Advertisements in newspapers-0.0111.134-0.0010.000
      Election newspaper/newsletter delivered to people’s home0.4410.9250.0260.000
      Voting advice application (Kieskompas/Stemwijzer/DeStemVan)0.0691.0220.0030.000
      Publicly accessible debates0.3800.9500.0210.000
      Election fairs-0.4561.255-0.0200.000
      Social media0.0950.4530.0110.000
      Website (municipality or special election website)0.2820.8160.0200.000
      Online videos (e.g., YouTube)0.4440.7230.0300.001

      * = p < 0.05; ** = p < 0.01; *** = p < 0.001

      The unique contribution of most of these variables to explain voter turnout in local elections is modest. The percentage of inhabitants aged 65 or older and the mean household income of a municipality accounts, for example, for respectively 1.4% (β = 0.143, p < 0.001, sr2 = 0.014) and 2.0% (β = 0.190, p < 0.001, sr2 = 0.020) of the variance in local turnout. The percentage of inhabitants of a municipality that attend a religious service at least once a month, however, has a large explanatory value with regard to local turnout. This variable uniquely explains – when the other socio-demographic variables and the variable municipal reorganization are held constant – approximately a third of all differences between municipalities in local turnout (β = 0.622, p < 0.001, sr2 = 0.322; see Table 3).
      A consequence of the fact that socio-demographic variables already account for more than two-thirds of all variance in turnout in municipal elections is that local governments’ efforts can only have a modest contribution. Nevertheless, excluding the effect that socio-demographic variables have, municipal communication activities and actions to facilitate voting could still help to increase the turnout in local elections. To this end, factors related to communication were added to the model with socio-demographic variables one by one. The results suggest that even though municipalities spend a lot of energy to increase the turnout in municipal elections, their communication actions are only of limited use in this respect. Starting with the communication efforts of municipalities and the communication means they employ, the multiple regression analysis shows that none of the communication instruments statistically contributes significantly to explaining local turnout.3xIn the analysis that follows, some voting facilities – communication specially targeted at ethnic minorities, special transportation towards polling stations, election parties where people could also vote, a cup of coffee when voting, and a (small) reward for voters – were excluded from the analysis because these facilities were employed in too few municipalities. The one exception is ‘advertisement in newspapers’ (β = –0.098 p < 0.05, sr2 = 0.007). Nevertheless, in this case, there is a negative relationship – turnout is actually lower in municipalities that use this instrument. As such, merely employing certain means of communication, for example, digital media or voting advice applications, does not help to improve local turnout (Table 4).
      It could be argued that this lack of explanatory value of various communication instruments for local turnout, or even the contrary effect of it, is due to the way in which municipalities make use of them. However, testing for the extent to which municipalities prepared their communication efforts, the aims of communication and at what point in time the communication campaign started, shows that this did not affect the explanatory value of municipal communication efforts for the local turnout in any significant way. In addition, addressing specific groups within society for whom voting is not self-evident (like the youth) does not seem to result in a higher turnout. Again, however, one could argue that this might be due to the way in which municipalities address these groups. Hence, the communication of local governments to citizens on municipal elections and the way in which that communication is organized hardly seems to contribute to the turnout in these elections.
      Naturally, based on this cross-sectional analysis there is no way to be sure that turnout would not be lower if municipalities had not engaged in these communication activities. After all, maybe municipalities are more inclined to start communication activities if they expect turnout to be low, for example compared to other municipalities. Therefore, the communication means employed by municipalities were also compared to their communication efforts in the period leading up to the 2010 municipal elections – did the introduction of new means of communication help to increase the turnout in the 2014 municipal elections compared to the elections of 2010? In line with the results from the cross-sectional analysis, employing communication instruments does not seem to contribute to increasing local turnout (see Table 5; R2-change = 0.024, Fchange (12, 328) = 0.763, p = 0.689).
      Aside from communication activities in the period leading up to the municipal elections, municipalities can also try to increase local turnout by facilitating voting on election day. In this respect, local governments’ activities relate to the number of polling stations, locating polling stations at central places (e.g., train stations), mobile polling stations, long opening hours and clear marking of where polling stations are positioned (see Table 2). The hierarchical multiple linear regression analysis suggests that activities to facilitate voting do not contribute to a higher local turnout while keeping the socio-demographic variables constant, with the exception of the number of polling stations per 1,000 inhabitants (see Table 6, R2-change = 0.069, Fchange (6, 228) = 10.913, p < 0.001). The analysis shows that the higher the relative number of polling stations, the higher the turnout (β = 0.293, p < 0.001). In terms of explanatory value for the variance in local turnout, the relative number of polling stations has a moderate effect (sr2 = 0.065).

      Table 6 Multiple regression analysis relating turnout in local elections to voting facilities
      BSE Bβsr2
      (Constant)17.3633.199
      % non-Western immigrants in municipality-0.3050.055-0.2410.032***
      % highly educated citizens (HBO or university)0.2160.0440.2220.025***
      % inhabitants aged 65 or older0.0780.0850.0380.001
      Household income (€)0.0010.0000.2900.042***
      % inhabitants visiting religious services at least once a month0.3640.0210.6120.307***
      Municipal reorganization elections around 2014-9.0401.393-0.2180.045***
      Polling station in central places (e.g., train station)-1.1640.744-0.0670.003
      Mobile polling stations (e.g., voting bus)-0.1550.712-0.0080.000
      Special opening hours of polling stations (e.g., at night)-0.1331.136-0.0040.000
      Signposts pointing towards polling stations0.1540.5190.0110.000
      Clear marking of polling stations-0.4680.757-0.0220.000
      Number of polling stations per 1,000 inhabitants14.0591.7960.2930.065***

      * = p < 0.05; ** = p < 0.01; *** = p < 0.001

    • 4 Conclusions and Discussion

      The central question in this article was to what extent Dutch municipal efforts to promote the turnout in local elections are effective. To answer this question, we sent a survey to municipal clerks and collected data on local turnout and socio-demographic characteristics of Dutch municipalities. Based on this, we analysed the effect of 16 municipal communication activities and actions to facilitate voting.4xFive activities or actions were excluded from the analysis because the facility was employed in too few municipalities: communication specially targeted at ethnic minorities, special transportation towards polling stations, election parties where people could also vote, a cup of coffee when voting and a (small) reward for voters. The results show that municipalities make use of many different communication means and voting facilities to lower the burden of voting, but that these have hardly any effect. Taking into account the effect of the socio-demographic variables, only one out of these 16 (advertisement in newspapers) showed a significant effect; this effect however was negative. The other 15 did not show a significant effect at all.
      In turn, differences in turnout in the Netherlands can be predominantly explained by socio-demographic factors and factors related to the local political situations. Socio-demographic factors account for a large part of the variance in local turnout between municipalities. More specifically, 69% of the variance can be explained by differences in the percentage of inhabitants aged 65 or older, the percentage of highly educated inhabitants, the average household income and the percentage of inhabitants with a non-Western immigrant background. The degree to which inhabitants regularly visit religious services (once a month or more often), and whether local elections are ‘regular’ or the consequence of municipal reorganizations appear to explain most of the differences in turnout between municipalities.
      As an addition to the literature, we first of all showed what municipalities actually do in terms of communication instruments and voting facilities to improve turnout. The findings regarding the effectiveness of municipal actions and activities in turn add to the literature in confirming that influencing voter turnout is difficult. However, in part, our findings contrast the findings in the predominantly U.S.-focused research in which scholars have at times found evidence that municipal communication efforts positively affect turnout (Green & Gerber, 2008).
      If we reflect on these findings, some possible explanations can be put forward, though these require further research. From a U.S. perspective, Dutch turnout in local elections is relatively high, which leaves less room for improvement. Some citizens do not vote for fundamental reasons, meaning that efforts to mobilize these voters have a smaller chance of succeeding (Krouwel, Kokx, & Pol, 2009). The higher the turnout in elections, the higher the share of this type of non-voters among the group of citizens that opts to abstain from voting. Moreover, only 15% of all municipalities based their selection of activities on research; most of them were based on ‘habit’ or wishes of local politicians. With such an arbitrary way of selection, a lack of concrete effect should also not be surprising. However, based on the cross-sectional analysis, we can equally not exclude that turnout would be even lower if municipalities had not engaged in these communication activities. And last, the lack of effectiveness can be a consequence of the fact that municipalities were our units of analysis. We have, therefore, not tested the impact of various communication instruments and voting facilities on the voting behaviour of individual citizens. It could be that on the individual level, these efforts of municipalities did have an impact, perhaps at least for some groups of voters. The effect of municipal efforts could be more conditional. For some groups of citizens, specific means of communication of voting facilities might have a positive impact, whereas for others it might not help at all. Previous research has, for example, indicated that the use of voting advice applications only impacts the turnout among particular segments of the electorate, including voters at lower education levels and younger voters (Gemenis & Rosema, 2014).
      Finally, the number of polling stations per 1,000 inhabitants appears to make a difference. The analysis showed that the higher the relative number of polling stations, the higher the turnout. Whereas literature is inconclusive on the extent to which the number of polling stations impact the turnout in elections, this study suggests that the relative number of polling stations per 1,000 inhabitants might have a positive effect. Possible reasons for this effect might be that a higher relative number of polling stations lowers the burden of voting by limiting the distance people need to travel (Dyck & Gimpel, 2005), as well as that each polling station has to serve fewer people. This could contribute to shorter lines. However, the effect found in this study of the number of polling stations on voter turnout sparks questions on the exact nature and significance of their relationship and requires further research. Is there, for example, an upper limit to the number of polling stations, above which turnout is not impacted? Or to what extent does increasing the number of polling stations between two consecutive elections contribute to the turnout?
      Perhaps more than for literature, the findings result in recommendations for municipalities. We see at least three. First, municipalities should improve the organization of their campaign. For an effective communication campaign, it is important to formulate a clear objective for the campaign and then choose resources that, based on (scientific) research, can be expected to effectively contribute to that goal. The second recommendation is that in their campaigns, municipalities should address those factors that according to research are relevant for local turnout. Studies show that political sophistication plays an important role (Van Ostaaijen et al., 2016) which can be better addressed by municipalities. Next to that, the importance of socio-demographic factors derived from this research leads to possibilities to improve turnout. While significantly altering their socio-demographic profile for most municipalities might still be only theory, more specific targeting of groups might prove to be effective in the short run. Many Dutch municipalities currently invest in getting young people to vote (with limited success) but refrain from targeting groups with lower income and lower education, or ethnic immigrants, groups of inhabitants among whom turnout, in general, is low as well. The last recommendation is that municipalities should pay more attention to local differences in voter mobilization. In certain municipalities, some turnout factors may play a larger role than in others. For instance, in one municipality more voters may stay at home due to practical problems (such as a lack of time) compared to another where more people stay at home because they think the municipality is of no importance. Both factors require a different approach from the local government. More local research is therefore needed to find out in what way general turnout factors play a role in individual municipalities. For the same reason, municipal elections should evaluate with research among citizens whether their efforts had the desired effect.

    • References
    • Aarts, K. (1999). Opkomst bij verkiezingen. Enschede: Faculteit bestuurskunde van de Universiteit van Twente (rapport voor het Ministerie van BZK).

    • Baekgaard, M., Jensen, C., Mortensen, P. B., & Serritzlew, S. (2014). Local news media and voter turnout. Local Government Studies, 40(4), 518-532.

    • Binnenlands Bestuur (2014). Bewustwording centraal bij verkiezingscampagnes. Online publication, 26 February.

    • Blais, A., & Aarts, K. (2006). Electoral systems and turnout. Acta Politica, 41(2), 180-196.

    • Blais, A., & Dobrzynska, A. (1998). Turnout in electoral democracies. European Journal of Political Research, 33(2), 239-261.

    • Blais, A., Gidengil, E., & Nevitte, N. (2004). Where does turnout decline come from? European Journal of Political Research, 43(2), 221-236.

    • BMC (2014). Gemeenteraadsverkiezingen 2014: Wie stemt, wat en waarom? Amersfoort: BMC onderzoek.

    • Boogers, M., Van Ostaaijen, J., & Slagter, L. (2011). Lokale kiezers, lokale keuzes? Onderzoek naar de achtergronden en de betekenis van het stemgedrag bij de gemeenteraadsverkiezingen van 2010. Bestuurswetenschappen, 65(6), 35-54.

    • Brady, H., & McNulty, J. (2011). Turning out to vote: The costs of finding and getting to the polling place. The American Political Science Review, 105(1), 115-134.

    • Cancela, J., & Geys, B. (2016). Explaining voter turnout: A meta-analysis of national and subnational elections. Electoral Studies, 42, 264-275.

    • COS: Rotterdam Research Agency (2010). Analyse Gemeenteraadsverkiezingen 2010. Rotterdam: Centrum voor Onderzoek en Statistiek (COS).

    • CPB (2014). Beantwoording vragen opkomstpercentage en herindelingen. Den Haag: Ministerie van Binnenlandse Zaken en Koninkrijkrelaties.

    • Dale, A., & Strauss, A. (2009). Don’t forget to vote: Text message reminders as a mobilization tool. American Journal of Political Science, 53(4), 787-804.

    • Dubois, E., & Blank, G. (2018). The echo chamber is overstated: The moderating effect of political interest and diverse media. Information, Communication & Society, 21(5), 729-745.

    • Dyck, J. J., & Gimpel, J. G. (2005). Distance, turnout, and the convenience of voting. Social Science Quarterly, 86(3), 531-548.

    • EenVandaag (2014a). ‘Als je niet stemt word je ook niet bedrogen’. Retrieved June 18, 2015, from www.EenVandaag.nl.

    • EenVandaag (2014b). PVV-stemmers vooral naar lokale partijen. Retrieved June 18, 2015, from www.EenVandaag.nl.

    • Eggers, A. C. (2015). Proportionality and turnout evidence from French municipalities. Comparative Political Studies, 48(2), 135-167.

    • Eisinga, R., Te Grotenhuis, M., & Pelzer, B. (2012). Weather conditions and voter turnout in Dutch national parliament elections, 1971-2010. International Journal of Biometeorology, 56(4), 783-786.

    • Field, A. (2013). Discovering statistics using SPSS. Thousand Oaks: Sage Publications.

    • Franklin, M. (1996). Electoral participation. In L. LeDuc, R. G., Niemi, &P. Norris (Eds.), Comparing democracies: Elections and voting in global perspectives (pp. 214-233). Thousand Oaks: Sage.

    • Franklin, M. (1999). Electoral engineering and cross-national turnout differences: What role for compulsory voting. British Journal of Political Science, 29(1), 205–216.

    • Franklin, M. (2002). The dynamics of electoral participation. In L. Leduc, R. G. Niemi, & P. Norris (Eds.), Comparing Democracies 2 (pp. 148-168). Thousand Oaks: Sage.

    • Franklin, M. (2004). Voter turnout and the dynamics of electoral competition in established democracies since 1945. Cambridge: Cambridge University Press.

    • Gallego, A. (2009). Understanding unequal turnout: Evidence and voting in comparative perspective. Electoral Studies, 29(2), 239-248.

    • Gemenis, K., & Rosema, M. (2014). Voting advice applications and electoral turnout. Electoral Studies, 36: 281-289.

    • Geys, B. (2006). Explaining voter turnout: A review of aggregate-level research. Electoral Studies, 25(4), 637-663.

    • Gomez, B. T., Hansford, T. G., & Krause, G. A. (2007). The Republicans should pray for rain: Weather, turnout, and voting in US presidential elections. Journal of Politics, 69(3), 649-663.

    • Górecki, M. A. (2013). Electoral context, habit-formation and voter turnout: A new analysis. Electoral Studies, 32(1), 140-152.

    • Green, D., & Gerber, A. (2008). Get out the vote: How to increase voter turnout. Washington, D.C.: Brookings Institution Press.

    • Harder, J., & Krosnick, J. A. (2008). Why do people vote? A psychological analysis of the causes of voter turnout. Journal of Social Issues, 64(3), 525-549.

    • Henderson, A., & McEwen, N. (2010). A comparative analysis of voter turnout in regional elections. Electoral Studies, 29(3), 405-416.

    • Hendriks, F., Van der Krieken, K., Van Zuydam, S., & Roelands, M. (2015). Bewegende beelden van democratie: Legitimiteitsmonitor Democratisch Bestuur 2015. Den Haag: Ministerie van Binnenlandse Zaken en Koninkrijksrelaties

    • Howard, L., & Posler, B. (2012). Reframing political messages: Using a festival to reach young voters. Journal of Political Science Education, 8(4), 389-407.

    • Jacobs, D., & Van Ostaaijen, J. (2017). De stembus voorbij: Een onderzoek naar het effect van een gemeentelijke campagne op de lokale opkomst bij de herindelingsverkiezing in Meierijstad. Tilburg University.

    • Jacobs, D., Van Zuydam, S., Van Ostaaijen, J., & de Brouwer, L. (2018). De wet van de grotere getallen: Het effect van meer stembureaus op de opkomst bij verkiezingen. Den Bosch/Tilburg: Provincie Noord-Brabant / Tilburg University.

    • Kemmelmeier, M. (2008). Is there a relationship between political orientation and cognitive ability? A test of three hypotheses in two studies. Personality and Individual Differences, 45, 767-772.

    • Korsten, A. F. A., & Janssen, J. (1999). De provinciale statenverkiezingen van maart 1999 en de voorstellen voor opkomstbevordering. Openbaar bestuur 4, 2-9.

    • Kranendonk, M., Michon, L., Schwarz, H., & Vermeulen, F. (2014). Opkomst en stemgedrag van Amsterdammers met een migratie-achtergrond tijdens de gemeenteraadsverkiezingen van 19 maart 2014. Amsterdam: Instituut voor Migratie en Etnische Studies.

    • Krouwel, A., Kokx, W., & Pol, M. (2009). Onderzoek opkomstbevordering Almere Eindrapport. Amsterdam: Centrum voor Toegepaste Politieke Wetenschappen.

    • Ladner, A., & Pianzola, J. (2010). Do voting advice applications have an effect on electoral participation and voter turnout? Evidence from the 2007 Swiss Federal Elections. In: Tambouris, E., Macintosh, A., Glassey, O. (eds), Electronic Participation: ePart 2010 Lecture Notes in Computer Science, vol 6229 . Berlin: Springer.

    • Landman, L., Kik, Q., Hermans, L., & Hietbrink, N. (2015). Nieuwsvoorziening in de regio 2014. “Gelukkig zijn hier geen journalisten”. Den Haag: Het Stimuleringsfonds voor de Journalistiek.

    • 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.

    • Loughlin, J., Hendriks, F., & Lidström, A. (2011). The Oxford handbook of local and regional democracy in Europe. Oxford: Oxford University Press.

    • Malhotra, N., Michelson, M., Rogers, T., & Valenzuela, A. (2011). Text messages as mobilization tools: The conditional effect of habitual voting and election salience. American Politics Research, 39(4), 664-681.

    • Neijens, P., & De Vreese, C. (2010). Hulp voor kiezers in referendums: Is de Informatie en Keuze Enquête een steun voor niet-geïnformeerde of juist voor geïnformeerde kiezers? Res Publica, 52(1), 130-132.

    • Pallant, J. (2013). SPSS survival manual. Maidenhead: Open University Press.

    • Panagopoulos, C. (2009). Street fight: The impact of a street sign campaign on voter turnout. Electoral Studies, 28(2), 309-313.

    • Panagopoulos, C. (2011). Social pressure, surveillance and community size: Evidence from field experiments on voter turnout. Electoral Studies, 30(2), 353-357.

    • Persson, M., Sundell, A., & Öhrvall, R. (2014). Does election day weather affect voter turnout? Evidence from Swedish elections. Electoral Studies, 33, 335-342.

    • Ruusuvirta, O., & Rosema, M. (2009). Do online vote selectors influence electoral participation and the direction of the vote? ECPR General Conference, Potsdam, Germany, September 10-12.

    • Schmitt-Beck, R., & Mackenrodt, C. (2010). Social networks and mass media as mobilizers and demobilizers: A study of turnout at a German local election. Electoral Studies, 29(3), 392-404.

    • SCP (2002). Niet stemmers: een onderzoek naar achtergronden en motieven in enquêtes, interviews en focusgroepen. Den Haag: Sociaal en Cultureel Planbureau.

    • SCP (2012). Burgerperspectieven 2012|4: kwartaalbericht van het Continu Onderzoek Burgerperspectieven. Den Haag: Sociaal en Cultureel Planbureau.

    • Sheafer, T. (2008). Charismatic communication skill, media legitimacy, and electoral success. Journal of Political Marketing, 7(1), 1-24.

    • Smets, K., & Van Ham, C. (2013). The embarrassment of riches? A meta-analysis of individual-level research on voter turnout. Electoral Studies, 32(2), 344-359.

    • TNS NIPO (2014). Gemeenteraadsverkiezingen gaan eindelijk echt lokaal. Retrieved February 27, 2014, from www.TNS Nipo.com.

    • Van den Bent, E. (2010). Proeftuin Rotterdam: Bestuurlijke maakbaarheid tussen 1975 en 2005. Rotterdam: Erasmus School of History, Culture and Communication (speech by the achievement of the thesis).

    • Van der Meer, T., & Van der Kolk, H. (2016). Rapport Lokaal KiezersOnderzoek 2016. Stichting Kiezersonderzoek Nederland.

    • Van Ostaaijen, J. (2014). Bevlogen en begrensd: Een analyse van de gemeenteraadsverkiezingen 2014. Tilburg: Tilburgse School voor Politiek en Bestuur, Tilburg University.

    • Van Ostaaijen, J., Epskamp, M., & Dols, M. (2016). Verbetering op komst: Een verkenning naar een effectieve gemeentelijke inzet van communicatiemiddelen voor de opkomst bij lokale verkiezingen. Tilburg: Tilburg University.

    • Vergeer, M. (2006). Lokale medialandschappen in Nederland 2005. Nijmegen: Radboud Universiteit (research assigned by the business fund for the press).

    • Vetter, A. (2015). Just a matter of timing? Local electoral turnout in Germany in the context of national and European parliamentary elections. German Politics, 24(1), 67-84.

    • Warner, R. (2013). Applied statistics: From bivariate through multivariate techniques. Thousand Oaks: Sage Publications.

    • Wille, A. C. (2000). De verdwenen kiezer terug: Opkomstbevorderende strategieën bij gemeenteraadsverkiezingen. Leiden: Instituut voor Publiek en Politiek.

    • Appendix A: Literature study results regarding factors relevant for turnout in Dutch local elections (Van Ostaaijen et al., 2016)

      FactorRelevant factor for Dutch local turnout?
      Individual factors
      GenderNo (Aarts, 1999; Harder & Krosnick, 2008; Krouwel et al., 2009; Smets & Van Ham, 2013)
      AgeYes; middle-aged group highest turnout, in the younger and older groups turnout decreases (Blais, Gidengil, & Nevitte, 2004; Harder & Krosnick, 2008; SCP, 2002, p. 26; SCP, 2012, p. 39; Smets & Van Ham, 2013)
      EducationYes; lower educated vote less (Boogers et al., 2011; SCP, 2002, 2012, p. 39; Smets & Van Ham, 2013)
      Income/EmploymentUndetermined, even though having a job seems more relevant than income (Cancela & Geys, 2016; SCP, 2012, p. 39; Smets & Van Ham, 2013)
      EthnicityYes; internationally less so than in the Netherlands (Cancela & Geys, 2016; COS, 2010; Kranendonk, Michon, Schwarz, & Vermeulen, 2014; Van den Bent, 2010)*
      ReligionYes; people that frequently attend religious services vote more often (SCP, 2002; Smets & Van Ham, 2013)*
      Interest in politicsYes (BMC, 2014; Boogers et al., 2011; EenVandaag, 2014a; Smets & Van Ham, 2013; SCP, 2002, 2012; TNS Nipo, 2014)
      Knowledge of politicsYes (SCP, 2002, 2012; Smets & Van Ham, 2013)
      Assumed relevance of local politicsUndetermined (BMC, 2014, p. 17; Lefevere & van Aelst, 2014, pp. 163-164)*
      Trust/confidence in politicsYes (Aarts, 1999; BMC, 2014; Boogers et al., 2011; Harder & Krosnick, 2008)
      Satisfaction with politicsYes (BMC, 2014; Boogers et al., 2011)
      Happiness and healthYes; especially happiness (SCP, 2002; SCP, 2012)*
      LifestyleYes; voluntary work improves turnout, people with no contacts have a bigger chance of not voting (SCP, 2002, 2012; Smets & Van Ham, 2013)
      Social obligationUndetermined
      Social pressure / environmentYes (Harder & Krosnick, 2008; Panagopoulos, 2011; Schmitt-Beck & Mackenrodt, 2010)*
      HabitYes; people who voted before probably will again (Cancela & Geys, 2016; Harder & Krosnick, 2008; Smets & Van Ham, 2013)
      PracticalitiesYes (SCP, 2002, 2012)
      Context- and institutional factors
      Municipal size and densityYes; municipal size matters (Blais & Dobrzynska, 1998; Cancela & Geys, 2016; CPB, 2014; Panagopoulos, 2011)
      Municipal stabilityUndetermined*
      Local and regional connectednessUndetermined*
      Economic circumstancesUndetermined*
      Electoral systemNo (Aarts, 1999; Blais & Aarts, 2006; Blais & Dobrzynska, 1998; Eggers, 2015; Geys, 2006)
      Day of the election/concurrent electionsYes; best day is a weekend day (Franklin, 1996, 2002, 2004; Henderson & McEwen, 2010) and together with or close to first order elections is beneficial (Cancela & Geys, 2016; Korsten & Janssen, 1999; Vetter, 2015)
      Compulsory vote and registrationYes; compulsory vote has a positive influence (Cancela & Geys, 2016; Gallego, 2009; Smets & Van Ham, 2013)
      Media presenceUndetermined (SCP, 2002; Vergeer, 2006)*
      Political situation and local elections
      Number of polling boothsUndetermined (Dyck & Gimpel, 2005; Krouwel et al., 2009)*
      Fragmentation and new political partiesUndetermined (BMC, 2014; EenVandaag, 2014b; Geys, 2006; Van Ostaaijen, 2014)
      Electoral competition / closenessYes (Cancela & Geys, 2016; Franklin, 2004; Geys, 2006; Górecki, 2013)
      Campaign expendituresUndetermined*
      Positive or negative campaigningUndetermined (Kirchgassner & Schulz, 2004 in Geys, 2006, p. 649)
      Importance of individual politiciansUndetermined*
      Media attentionYes (Baekgaard, Jensen, Mortensen, & Serritzlew, 2014; Landman, Kik, Hermans, & Hietbrink, 2015)*
      Weather conditionsYes (Eisinga, Te Grotenhuis, & Pelzer, 2012; Gomez, Hansford, & Krause, 2007; Persson et al., 2014)

      * Difficult to find substantial amount of research

    Noten

    • 1 In case of municipal reorganization, we included the turnout in the local elections held between 2012 and 2015 in the analysis. After all, for these municipalities, no local elections were organized in 2014. With regard to the 2010 election turnout, we included the turnout in elections following municipal reorganization between 2008 and 2011.

    • 2 The database of the Dutch Electoral Council can be accessed via: https://www.verkiezingsuitslagen.nl/. It includes the results and turnout in all Dutch elections starting from 1848.

    • 3 In the analysis that follows, some voting facilities – communication specially targeted at ethnic minorities, special transportation towards polling stations, election parties where people could also vote, a cup of coffee when voting, and a (small) reward for voters – were excluded from the analysis because these facilities were employed in too few municipalities.

    • 4 Five activities or actions were excluded from the analysis because the facility was employed in too few municipalities: communication specially targeted at ethnic minorities, special transportation towards polling stations, election parties where people could also vote, a cup of coffee when voting and a (small) reward for voters.


Print dit artikel