Electoral Rules and Voter Turnout
Abstract and Keywords
What is the relationship between electoral rules and voter turnout? To answer this question, this chapter begins with the more general question of why voters turn out on Election Day at all. The chapter then reviews the main theoretical arguments and empirical evidence of how electoral rules affect voters’ turnout decisions, as well as the puzzles that have emerged from that literature. It then highlights some recent theoretical, methodological, and empirical advancements using subnational data and electoral system reforms that have refined our understanding about the interaction between electoral rules, competitiveness, and turnout. Finally, the chapter explores new and promising directions for further research.
What is the relationship between electoral rules and voter turnout? To answer this question, it is helpful to begin with the more general question of why voters turn out on Election Day at all. The pioneering rational choice work by Downs (1957), Tullock (1968), and Riker and Ordeshook (1968) argued that closely contested, competitive elections will increase the chance that a single voter might become “pivotal” in determining the outcome, and thus increase the incentives for voters who care about the outcome to turn out and vote. A problem is that such pivotal voter theories, if taken seriously, would predict increasingly lower turnout rates as the size of the electorate expands—even in competitive races—since the probability of any one voter becoming pivotal would get closer and closer to zero (Palfrey and Rosenthal 1985).1 And yet citizens still turn out to vote.
To address this apparent paradox of voting, subsequent approaches shifted the focus from voters’ personal calculations of pivotality to the mobilizational incentives and efforts of elites—including politicians, their parties, and various interest groups in society aligned with those parties (e.g., Morton 1987, 1991; Uhlaner 1989; Cox and Munger 1989; Shachar and Nalebuff 1999). The basic argument is that elite actors in close races might rationally invest in mobilizing voters (i.e., through increased campaign spending and get-out-the-vote activities), and those voters might rationally respond to such mobilization efforts by turning out to vote. There are two likely targets of mobilization: (1) individuals or groups who would likely be supportive but might not otherwise turn out if not contacted and (2) individuals or groups who would be most effective at secondary mobilization (e.g., Huckfeldt and Sprague 1992; Rosenstone and Hansen 1993). The former strategy might help explain the well-documented pattern of “friends and neighbors” voting around candidates’ hometowns (e.g., Key 1949; Lewis-Beck and Rice 1983; Rice and Macht 1987; Górecki and Marsh 2012; Fiva and Smith 2017). The latter means that politicians or parties will attempt to mobilize key activists or group leaders, and those individuals, in turn, will mobilize other members of their social groups. The micro-foundations behind the elite mobilization theories may thus be related to social pressure or influence within social networks (e.g., Cox, Rosenbluth, and Thies 1998; Bond et al. 2012; Cox 2015).
Even as the debate over the exact mechanisms behind voter turnout was very much unresolved, other scholars began to consider how turnout incentives might be conditioned by the electoral system, especially the degree of proportionality generated by the seat allocation rules (e.g., Powell 1980, 1986; Jackman 1987; Blais and Carty 1990). The early arguments about competitiveness and turnout by Downs and others were based on the assumption of electoral competition in single-member district (SMD) races. However, Gosnell (1930) and Tingsten (1937) had long ago observed that turnout increased in Germany, Norway, and Switzerland following electoral system reforms in the early 1900s from a two-round runoff system with SMDs to a proportional representation (PR) system with multimember districts (MMDs). Beginning with the seminal studies of Powell (1980, 1986) and Jackman (1987), a renewed interest in the relationship between electoral rules and turnout picked up speed. In a recent meta-analysis of 185 studies on the determinants of aggregate national and subnational turnout, Cancela and Geys (2016) count 51 studies since Powell (1980) that in some way or another test the relationship between the proportionality of the electoral rules and voter turnout (and their sample of studies is not exhaustive).
In this chapter, I first review the main theoretical arguments and empirical evidence of the relationship between electoral rules and turnout, as well as the puzzles that have emerged from that literature.2 I then highlight some recent theoretical, methodological, and empirical advancements using subnational data that have refined our understanding about the interaction between electoral rules, competitiveness, and turnout—but have also opened up new and promising directions for future research.
Proportionality and Turnout
The arguments for how proportionality might increase turnout rest on four basic—and related—mechanisms. The first explanation is that voters are likely to feel more “efficacious” in electoral systems that produce more proportional outcomes (e.g., Blais and Carty 1990; Lijphart 1997; Karp, Banducci, and Bowler 2008; Fisher et al. 2008). At higher levels of district magnitude (M), the translation of votes into seats will be less distorted (meaning there will be less disproportionality between vote shares and seat shares), thereby increasing voters’ feelings that their votes actually count. Put differently, voters have less need to be concerned that their votes will be “wasted” on losing candidates or parties, and hence will be more likely to turn out to cast those votes. This explanation rests clearly on voter-driven incentives rather than elite-driven incentives as its mechanism, and is relatively underexplored in the literature.
A second explanation is that PR systems, owing to a larger M, are more permissive to the entry of a greater number of parties (Duverger 1954; Cox 1997). From the perspective of voter-based theories, this means that there may be less reason to abstain for a lack of party options matching voters’ ideological preferences (e.g., Powell 1986; Jackman 1987; Ladner and Milner 1999). From the perspective of the elite mobilization theories, this explanation implies that there will be more parties that will make efforts at mobilizing different segments of the population. For example, niche or extreme left/right parties will mobilize voters on the edge of the political spectrum who might be indifferent to the mainstream parties, hence increasing the overall level of turnout.
A third explanation, which rests firmly in the camp of elite mobilization theories, is that PR systems may feature more powerful linkages between parties and groups in society, thereby facilitating mobilization (Powell 1980). For example, many of the early postwar parties of Western Europe, where PR systems have been prevalent, conformed to the “mass party” model of party organization (Duverger 1954; Neumann 1956), in which “the fundamental units of political life are pre-defined and well-defined social groups, membership in which is bound up in all aspects of the individual’s life” (Katz and Mair 1995, 6). Given the close association with well-defined social groups—such as labor union organizations and religious movements—the mass party model facilitates mobilization, and especially secondary mobilization. Cox (2015) refers to this process as “subcontracting.” Party elites need only get the assurance of support from leaders of such groups, who then mobilize their members on behalf of the party.
The fourth and final basic explanation is that PR elections tend to be more nationally competitive across districts than SMD elections (Jackman 1987). In SMD systems, some districts will be competitive “swing” districts, but many others will be “safe” districts, where one party is overwhelmingly expected to win the seat. Given the logic of either the pivotal voter theories or the elite mobilization theories, turnout will be much lower in these safe districts.3 In PR systems, all districts are expected to be competitive, as each additional vote earned by a party can help increase the number of seats the party might win in parliament. Hence, parties have an incentive to campaign across all areas of the country, not just in the areas where such an effort might pay off. It follows from this logic that PR systems will also exhibit less variance in turnout across districts than SMD systems (Cox 1999).
More broadly, Cox (1999) postulates that elite incentives to mobilize voters will depend on three “translation” processes of a given system: (1) how effort translates into votes, (2) how votes translate into seats, and (3) how seats translate into portfolios (including executive offices, but also postelectoral legislative offices and judicial appointments).4 The first speaks to the marginal costs and benefits of different technologies or strategies of voter mobilization. The second is determined by the electoral rules. The third can vary depending on higher-order conditions such as the dynamics of coalition formation or the separation of executive powers.
Existing Empirical Evidence and Puzzles
Despite the compelling explanations for why turnout ought to be higher under proportional electoral rules, the existing empirical evidence paints a rather mixed picture. For the set of advanced industrialized democracies, there is broad and consistent evidence that mean turnout is higher under PR systems than under SMD systems (e.g., Powell 1980; Blais and Carty 1990; Jackman and Miller 1995; Franklin 1996; Blais and Dobrzynska 1998). The exceptions are Switzerland, where turnout is relatively low despite the use of PR, and New Zealand (prior to electoral reform in 1993), where turnout was high, despite the use of first-past-the-post (FPTP) plurality rule in SMDs.
However, when cross-national samples are expanded to include new and developing democracies in Latin America and Eastern Europe, the relationship between proportionality and turnout is much less consistent (e.g., Pérez-Liñán 2001; Kostadinova 2003; Fornos, Power, and Garand 2004; Blais and Aarts 2006; Endersby and Krieckhaus 2008; Gallego, Rico, and Anduiza 2012). The mixed empirical evidence from cross-national studies outside of the industrialized world leads Blais (2006, 111) to conclude that “the impact of institutional variables may be overstated.” Others, such as Selb (2009), characterize the evidence for a positive relationship between proportionality and turnout as “overwhelming,” with the main ambiguity in the literature being which of the four possible mechanisms is responsible for the effect.
In terms of voter feelings of efficacy, Banducci, Donovan, and Karp (1999) find an increase in perceptions of efficacy in the case of New Zealand after the 1993 electoral reform from FPTP/SMD to a mixed-member proportional (MMP) system. The effect they estimate is especially significant among political minorities and supporters of smaller parties. On the other hand, although aggregate turnout increased in the first election under MMP in 1996, it subsequently declined to levels that were lower than the pre-reform period (Vowles 2002, 2010), continuing a downward trend that began before the reform.
The evidence has also been mixed to poor in terms of the number of parties as a mechanism. Although many studies have found a relationship between PR and the number of parties entering competition (e.g., Cox 1997; Eggers 2015; Cox, Fiva, and Smith 2016), there is less evidence for a relationship between the number of parties and higher turnout (e.g., Brockington 2004; Blais and Aarts 2006). However, the mixed results apply mainly to increasing the number of parties within PR systems, beyond the initial difference between M = 1 and M > 1 (Grofman and Selb 2011). Within SMD systems, in contrast, there is strong evidence that increasing the number of parties or candidates does indeed increase turnout (e.g., Indridason 2008; Scheiner, Smith, and Thies 2015; Fiva and Smith 2017). The counterintuitive relationship between the number of parties and turnout may be because a greater number of party options in PR systems increases the decision costs facing voters (Downs 1957), or because more parties can lead to coalition governments and less clarity in voter choice (Jackman 1987).
In support of this latter explanation, Tillman (2015) finds evidence that pre-electoral coalitions help to mitigate the downward effect of multipartism on turnout. In terms of Cox’s (1999) three translation processes, pre-electoral coalitions might be thought of as increasing certainty about the seats-to-portfolios translation insofar as they give voters and elites more information about which parties might form a government after the election. If the pre-electoral coalitions are formed across SMDs, they might also affect the votes-to-seats translation by removing the possibility that candidates from two or more ideologically proximate parties might split the vote and lose the district to a less ideologically compatible opponent.
Another possible explanation for the downward effect of multipartism on turnout is that when there are many parties, it might not always be clear that the additional vote gains from mobilization efforts targeted at specific groups of voters will accrue only to the party making those efforts. For example, if there are two parties on the left and an imperfect mapping of leftist voters into social groups that support those two parties, then any attempt to mobilize voters by one party might increase votes for its rival due to “mobilizational spillovers” (Cox 2015), reducing the incentives for either party to attempt to mobilize voters beyond its core supporters. The more parties there are in PR systems, the more likely it is that such scenarios will arise.
Only a handful of studies have attempted to directly measure differences in elite mobilization efforts across electoral systems. Using voter survey data from the Comparative Study of Electoral Systems (CSES) dataset, both Karp, Banducci, and Bowler (2008) and Rainey (2015) find that voters actually report lower levels of direct contact by parties in PR systems relative to SMD systems. This would seem to run counter to expectations about competitiveness and incentives to mobilize across districts under PR. However, the existence of subcontracting to affiliated groups as a mobilization strategy in PR systems may help to explain these results—if parties are working through secondary mobilization channels, they are less likely to need to engage in direct contact (Cox 2015). More generally, if PR districts tend to be larger in size and population than SMDs, then direct contact may be a less efficient mobilization strategy than other elite efforts, such as running television or newspaper ads that reach a large audience at once.
Overall, of the fifty-one aggregate-level studies included in the meta-analysis by Cancela and Geys (2016)—which collectively included 239 different tests (estimates) of the relationship between proportionality and turnout—only 61 percent of tests and just 53 percent of studies overall pointed in the expected (positive) direction. This contrasts with an earlier meta-analysis of fourteen studies (Geys 2006a), in which approximately 70 percent of tests and studies reported positive findings. A key difference in the study samples is the addition of many more studies using cases from outside the set of advanced industrialized democracies.5 It would seem that the conventional wisdom regarding the effect of proportionality on turnout might be in need of rethinking.
However, one of the problems with any meta-analysis is that the aggregate pattern that emerges is only as good as the individual studies that go into the sample. One issue is measurement. About half of the studies included in the Cancela and Geys (2016) meta-analysis use a simple dummy variable to operationalize the entire electoral system as “majoritarian” or “proportional.” The rest of the studies use either a measure of proportionality such as the Gallagher index (Gallagher 1991) or simply district magnitude. Most often, these variables are also measured at the aggregate (national) level as the mean across all districts and do not take into account how the level of competition, as well as the number of parties, varies at the subnational level across districts. The dynamics of mixed-member systems that combine SMD and PR in separate tiers also tend to be oversimplified—for example, treating MMP cases like New Zealand as “proportional” despite the potential variation in district-level competition across SMDs within the system.
The fact that most of the existing literature is based on cross-national comparisons of aggregate turnout data is itself another issue. As Herrera, Morelli, and Palfrey (2014, 132) point out, empirical investigations based on cross-national comparisons are likely to suffer from a number of confounding variables and measurement challenges, including “the measurement of competitiveness, properly controlling for social/cultural factors, endogeneity of the choice of electoral system, isolating the effects of district magnitude or multimember districts and taking into account institutional variations in government formation.” This makes it harder to identify a causal effect of electoral rules on turnout.
Ironically, many of the studies that explore single-country variation are based on the well-known exceptions to the aggregate-level evidence, Switzerland and New Zealand. Nevertheless, when cross-national confounding variables are removed from the analysis, the state of the evidence appears to improve. Of the fifty-one studies surveyed by Cancela and Geys (2016), only ten use subnational variation to evaluate the relationship between electoral rules and turnout; however, 70 percent of these subnational studies (and 76 percent of individual tests) produce positive findings. In contrast, the success rate for the subsample of forty-one studies using national-level data is just 51 percent (57 percent of tests). In other words, it would appear that subnational data do a better job at capturing the expected effects of electoral rules on turnout.
Moving beyond Cross-National Comparisons of Aggregate-Level Data
The increased availability of district-level data, including newly digitized historical electoral data covering important periods of electoral system reform, has made subnational analyses of turnout possible for many democracies. For example, Selb (2009) uses district-level data from the CSES dataset to verify the hypothesis that variance in turnout across districts is lower in more proportional systems.6 There have also been several methodological advancements in how to measure competitiveness across SMD and MMD systems (e.g., Blais and Lago 2009; Grofman and Selb 2009; Folke 2014; Kayser and Lindstädt 2015). Using their “index of competition” measure, Grofman and Selb (2011) compare across districts of varying magnitude in Switzerland and Spain, two cases that feature both SMDs and MMDs. They find that once one moves beyond M = 2, subsequent increases in district magnitude do not appear to have a positive effect on either competitiveness or turnout.
A few subnational studies are also able to exploit quasi-experimental variation to provide better insight into the causal effect of electoral rules on turnout. A shortcoming is that such quasi-experimental settings are limited to the electoral systems in use, which are often not simple versions of FPTP/SMD or PR, making it harder to draw general inferences about the relationship between proportional representation and turnout. For example, Fauvelle-Aymar and Lewis-Beck (2008) compare simultaneous cantonal and regional elections in France that are held under two-round runoff (TRS) and closed-list PR (CLPR) systems, respectively. In the regional CLPR elections, turnout is estimated to be 5 percentage points higher; however, they find no effect when the sample includes only first-round elections for the TR/SMD cantonal observations.
Eggers (2015) uses a population-based regression discontinuity (RD) design applied to municipal elections, also in France. There, a population threshold of 3,500 inhabitants triggers the use of CLPR with a 50 percent majoritarian bonus to the largest party, rather than a multiple nontransferable vote system in MMDs (MNTV/MMD). In the latter system, voters get M votes to cast in an M-sized district (average M = 25). They are allowed to engage in “plumping” (not using all M votes) and “panachage” (casting votes for candidates of different parties), but not “cumulation” (casting multiple votes for the same candidate) (Cox 1997, 42–43). Eggers (2015) finds a slight (1 percentage point) increase in mean turnout under the PR system and a lower level of variance in turnout across municipalities using PR, in line with the earlier cross-national, district-level analysis by Selb (2009).
Sanz (2015) uses the same RD approach to explore the case of Spain, where municipalities above 250 inhabitants use CLPR, those with fewer than 100 inhabitants only elect a local mayor through FPTP/SMD, and those with populations in between employ a system that is similar to the MNTV/MMD system in French elections studied by Eggers (2015), except that parties may nominate up to five candidates for five available seats, and voters are only allowed to cast up to four votes for candidates, within or across party lines.7 Interestingly—in contrast to the results in Eggers (2015)—Sanz (2015) finds that turnout in the CLPR elections is 1 to 2 percentage points lower than turnout in the MNTV/MMD elections, a finding that he attributes to the fact that MNTV/MMD systems encourage competition not only between parties but also between candidates of the same party. In other words, Eggers (2015) interprets his findings from France to imply that greater proportionality under CLPR produces higher turnout than majoritarian systems; Sanz (2015) interprets the exact opposite findings to imply that open-list systems produce higher turnout than closed-list systems. A possible explanation for this apparent discrepancy is the 50 percent seat bonus given to the largest party in the French case. If elites’ incentives to mobilize are conditioned by the translations of (1) effort to votes, (2) votes to seats, and (3) seats to portfolios, then the 50 percent bonus would have a larger impact on the votes-to-seats translation, and hence encourage more eager mobilization, than the conditions under more standard methods of seat allocation (i.e., D’Hondt or Sainte-Laguë).
Building on the variance hypothesis, Cox (1999, 399) conjectures that the differences in mean (aggregate) turnout between SMD and PR systems will ultimately depend on the distribution of the cross-district variance in competitiveness in the SMDs, a conjecture that has recently been formalized into a game-theoretic model by Herrera, Morelli, and Palfrey (2014). The basic logic of the model predicts that in highly competitive districts under SMD, incentives to mobilize will actually be higher than under PR because the stakes are higher—with only one seat up for grabs and “winner take all” plurality rule. In SMDs where the result is a foregone conclusion, elites have fewer incentives to mobilize. In PR elections, incentives to mobilize, and hence turnout, will be somewhere between the two types of SMDs. Hence, moving from SMD to PR should produce a “contraction effect” in the variance of turnout (Cox, Fiva, and Smith 2016). Whether or not the aggregate level of turnout is higher under PR or SMD depends on how many competitive SMDs exist in a system.
Cox, Fiva, and Smith (2016) explore the historical case of electoral reform in Norway to provide empirical support for the “contraction effect” of PR on mobilization and turnout. In 1919, Norway switched from a TR/SMD system to CLPR. Using municipality-level data to construct stable subnational units across the two electoral system periods, they find that in the regions that were closely contested in the pre-reform period, competitiveness and turnout declined following the introduction of PR. In all other districts, competitiveness and turnout increased. Because there were more noncompetitive SMDs than competitive SMDs in the pre-reform period, the aggregate effect was an increase in turnout under PR. In the two elections immediately before and after the reform, mean turnout increased from 58 percent to 65 percent, and the standard deviation of turnout fell from 15 percentage points to 9 percentage points. The evidence from Norway thus provides compelling before-and-after evidence to corroborate the cross-sectional evidence of the variance hypothesis provided by Selb (2009) and Eggers (2015), and adds additional points to the scorecard of the cross-district competitiveness mechanism as an explanation for the proportionality–turnout nexus.8
It should be clear from the previous sections that much work remains to be done. One promising direction for future research, building on the findings of Cox, Fiva, and Smith (2016), would be to explore additional cases of electoral reform using district-level data or municipality-level data aggregated to stable geographic units. Figure 1 makes a first attempt at this approach with box-and-whisker plots of the distribution of district-level turnout before and after major electoral reforms in the lower (or only) chambers of four countries: Norway, Italy, New Zealand, and Japan. The Norwegian case is the 1919 reform from TR/SMD to CLPR explored by Cox, Fiva, and Smith (2016), though here the data are the actual district-level returns to compare it to the other three cases. Italy used open-list PR (OLPR) for the Chamber of Deputies until 1993, when it switched to a mixed-member majoritarian (MMM) system. That system was in use for three elections before a second reform in 2005 to CLPR with a majoritarian bonus. As mentioned, New Zealand switched from FPTP/SMD to MMP in 1993. Japan changed its system for the House of Representatives from the single nontransferable vote (SNTV) system with MMDs (average M = 4) to its current MMM system in 1994. For the mixed systems, the data are only for the SMD tiers.
Even with turnout measured at the district level, the contraction effect on variance is clear for the Norwegian case. For the Italian case, there is also some evidence that the reverse logic of the contraction effect—that is, an “inflation effect” on variance—might hold when a country moves from a more proportional system to a less proportional system. Although turnout was already quite high under OLPR, some SMDs under the MMM system had even higher turnout in the first two elections after the reform, even as many more experienced a drop in turnout.9 In New Zealand, mean turnout initially increased in the first two elections under MMP and then decreased. However, the variance across districts appears to have contracted under MMP. The Japanese reform coincided with a drop in mean turnout, but the cross-district effect, if any, is less obvious. Future research should further explore these and other reforms with more fine-grained data. These reform cases, along with the disparate conclusions in much of the existing literature, also suggest the need to sharpen both predictions and measurement by moving beyond simplifications of electoral systems as “majoritarian versus proportional.”
Mixed-member cases like Japan, New Zealand, and Italy provide some particularly interesting opportunities for further investigation. There are two common varieties of mixed-member systems: MMM and MMP (Shugart and Wattenberg 2001). Most are composed of two tiers—one SMD tier and one PR tier—though mixed systems can take other forms, such as in the Japanese House of Councillors (the upper chamber of parliament), which currently combines SNTV/MMD and OLPR. Under MMM, seat allocation in the two tiers occurs in parallel; under MMP, seat allocation is compensatory, meaning the PR vote determines the overall proportion of seats to which a party is entitled. Maeda (2016) argues that the patterns in competitiveness and turnout in MMM and MMP should be similar to the patterns in SMD and PR, respectively, as the SMD tier is most determinative of the final seat allocation outcome under MMM, while the PR tier matters most under MMP. Using district-level panel data from two MMM cases (Japan and South Korea) and two MMP cases (Germany and New Zealand), he finds that changes in competitiveness (measured by vote margin between the top two candidates) and changes in turnout are more strongly related in SMDs under MMM than under MMP.10
Although Maeda (2016) focuses his analysis on the change in competitiveness and turnout (i.e., the dynamic relationship), the static cross-sectional relationship is also revealing. Figure 2 plots the relationship between district-level turnout (y-axis) and the district-level vote share margin between the top two candidates (x-axis) for all elections held under the mixed-member systems in Japan and New Zealand between 1996 and 2014 (seven elections for each country). The figure shows a positive relationship between district-level competitiveness and turnout in both cases. However, it is also clear that there is more variance in competitiveness and turnout in Japan (MMM) than in New Zealand (MMP). Given that the variance in turnout was already low in New Zealand under FPTP/SMD, these differences are likely contextual. Nevertheless, such patterns are worth further exploration.
The fact that there is a positive relationship between competitiveness and turnout in SMD races in mixed-member systems suggests that elite effort is not uniform across districts despite the possibility that additional mobilization—even in less competitive SMDs—might impact a party’s seat allocation outcome in the PR tier. In other words, individual effort by candidates in SMDs may still be important for mobilizing voters. Figure 3 illustrates the point with municipality-level vote returns in the SMD and PR tiers of Japan aggregated to the level of the SMD. Japan’s MMM system combines 300 SMDs (295 in 2014) and 180 (200 in 1996) seats allocated by CLPR in eleven regional districts. All of the SMDs “map into” the larger PR districts. The top-left panel of figure 3 plots the relationship between SMD-level turnout in the SMD tier on the y-axis and competitiveness in the SMD tier (measured using the Grofman-Selb index of competition for consistency of measurement across tiers) on the x-axis. The top-right panel plots the relationship between SMD-level turnout in the SMD tier and competitiveness in the PR tier (the clumping of points is because all SMDs located in the same PR district have the same value on the index of competition). The bottom-left panel plots the relationship between SMD-level turnout in the PR tier and competitiveness in the SMD tier, and the bottom-right panel plots the relationship between SMD-level turnout in the PR tier and competitiveness in the PR tier. The figure clearly shows that SMD-level turnout is positively related to competition in the SMDs, but not competition at the PR tier level. It also clearly shows the contraction effect in the measure of competitiveness (index of competition) across districts in the PR tier relative to the SMD tier.
The impact of individual-level effort on overall mobilization across different electoral rules is a third promising avenue for future research. Much of the existing research considers a general typology of “majoritarian” versus “proportional” families of electoral systems. However, subtler differences between types of electoral systems within these two broad families are also likely to play a role. For example, how do competitiveness and turnout differ between SNTV/MMD and single transferable vote (STV/MMD) systems? Or FPTP versus TR versus alternative vote (AV) variations in SMD systems?11 Most obviously, among list PR systems, OLPR differs from CLPR in that it generates competition not only between parties but also between individual candidates on party lists. Such intraparty competition for preference votes can potentially increase mobilization and turnout—though the few existing studies to explore this hypothesis have produced mixed results (e.g., Karvonen 2004; Hix and Hagemann 2009; Robbins 2010; Nemoto and Shugart 2013; Sanz 2015). One challenge is the availability of quasi-experimental settings or subnational variation in electoral systems to get around the causal inference challenges posed by cross-national comparisons.
Much of the mobilizational effect of individual candidates, if it exists, may depend on how parties adapt nomination strategies under OLPR. Do they nominate candidates that represent different geographical regions in the district? Or different sectoral interests? Again, the case of Japan provides a useful example. Prior to 2001, the nationwide district tier of the mixed-member system for the House of Councillors used CLPR rather than OLPR.12 After the switch to OLPR, Kōmeitō—a small religious party with a stable support base of voters—optimized its mobilization of supporters through an innovative strategy: the party leadership assigned its list candidates to specific geographic regions and instructed its followers on which candidates should earn their preference vote. The candidates, for their part, focused their mobilization efforts within their assigned territories. In the prefecture-based SNTV tier of the system, the party nominated a candidate in only a small number of districts in both 1998 (under CLPR) and 2001 (under OLPR). With the switch to OLPR, the vote share for Kōmeitō increased by several thousand more votes in the prefectures where the party did not run a prefectural-tier candidate in either election, but now had a dedicated “local” candidate on the party’s nationwide list (Smith 2014).
Finally, it is important to keep in mind that the incentives for elites to mobilize voters in elections will ultimately depend on what is at stake, including the seats up for election, but also higher-order outcomes like which party or parties will be able to form a government. As Cox (1999) and Herrera, Morelli, and Nunnari (2016) argue, a complete assessment of how electoral rules affect turnout must take into account not only the votes-to-seats translation imposed by the electoral system but also the seats-to-portfolios translation in a given system, and interactions between them.
Does it make sense to believe that voters are simultaneously concerned with how their vote will be translated into the outcome of legislative seats, as well as how mundane higher-order rules will filter that choice into portfolio outcomes? Probably not. However, it may be reasonable to believe that individual elites are more likely to understand such structural incentives and target their mobilizational activity accordingly. For example, if cabinet promotion in a given system is decided on the basis of seniority rules, then an individual incumbent politician on the cusp of reaching the requisite seniority might invest a bit more effort in ensuring his or her party wins enough seats to enter government. This calculus, too, can be expected to differ under different electoral system rules. In systems such as PR or STV, where one individual’s effort can help to elect his or her copartisans due to vote pooling or vote transfers, such effort will have a higher payoff. In SMD systems, the most important thing may be simply to win; winning by a larger margin through increasing turnout does not reap the same sort of additional payoffs. Research in this area is so far nonexistent, but efforts to explore such individual-level elite mobilization incentives will no doubt yield valuable insights into the overall relationship between electoral rules and turnout.
I thank João Cancela and Benny Geys for kindly sharing detailed information about the studies included in their meta-analysis, Colleen Driscoll for assistance collecting additional literature, and Gary W. Cox, Jon H. Fiva, and Benny Geys for helpful feedback.
Aldrich, John. “Rational Choice and Turnout.” American Journal of Political Science 37, no. 1 (1993): 246–278.Find this resource:
Banducci, Susan A., Todd Donovan, and Jeffrey A. Karp. “Proportional Representation and Attitudes about Politics: Results from New Zealand.” Electoral Studies 18, no. 4 (1999): 533–555.Find this resource:
Barone, Guglielmo, and Guido de Blasio. “Electoral Rules and Voter Turnout.” International Review of Law and Economics 36 (2013): 25–35.Find this resource:
Blais, André. “What Affects Voter Turnout?” Annual Review of Political Science 9, no. 1 (2006): 111–125.Find this resource:
Blais, André, and Kees Aarts. “Electoral Systems and Turnout.” Acta Politica 41, no. 2 (2006): 180–196.Find this resource:
Blais, André, and R. K. Carty. “Does Proportional Representation Foster Voter Turnout?” European Journal of Political Research 18, no. 2 (1990): 167–181.Find this resource:
Blais, André, and Agnieszka Dobrzynska. “Turnout in Electoral Democracies.” European Journal of Political Research 33, no. 2 (1998): 239–261.Find this resource:
Blais, André, and Ignacio Lago. “A General Measure of District Competitiveness.” Electoral Studies 28 (2009): 94–100.Find this resource:
Bond, Robert M., Christopher J. Fariss, Jason J. Jones, Adam D. I. Kramer, Cameron Marlow, Jaime E. Settle, and James H. Fowler. “A 61-Million-Person Experiment in Social Influence and Political Mobilization.” Nature 489 (2012): 295–298.Find this resource:
Brockington, David. “The Paradox of Proportional Representation: The Effect of Party Systems and Coalitions on Individuals’ Electoral Participation.” Political Studies 52, no. 3 (2004): 469–490.Find this resource:
Caldeira, Gregory A., and Samuel C. Patterson. “Contextual Influences on Participation in U.S. State Legislative Elections.” Legislative Studies Quarterly 7, no. 3 (1982): 359–381.Find this resource:
Cancela, João, and Benny Geys. “Explaining Voter Turnout: A Meta-Analysis of National and Subnational Elections.” Electoral Studies 42 (2016): 264–275.Find this resource:
Cox, Gary W. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. New York: Cambridge University Press, 1997.Find this resource:
Cox, Gary W. “Electoral Rules and the Calculus of Mobilization.” Legislative Studies Quarterly 24, no. 3 (1999): 387–419.Find this resource:
Cox, Gary W. “Electoral Rules, Mobilization, and Turnout.” Annual Review of Political Science 18, no. 1 (2015): 49–68.Find this resource:
Cox, Gary W., Jon H. Fiva, and Daniel M. Smith. “The Contraction Effect: How Proportional Representation Affects Mobilization and Turnout.” Journal of Politics (2016).Find this resource:
Cox, Gary W., and Michael C. Munger. “Closeness, Expenditures, and Turnout in the 1982 U.S. House Elections.” American Political Science Review 83, no. 1 (1989): 217–231.Find this resource:
Cox, Gary W., Frances M. Rosenbluth, and Michael F. Thies. “Mobilization, Social Networks, and Turnout.” World Politics 50, no. 3 (1998): 447–474.Find this resource:
Dawson, Paul A., and James E. Zinser. “Political Finance and Participation in Congressional Elections.” Annals of the American Academy of Political and Social Science 425 (1976): 59–73.Find this resource:
De Paola, Maria, and Vincenzo Scoppa. “The Impact of Closeness on Electoral Participation Exploiting the Italian Double Ballot System.” Public Choice 160, no. 3–4 (2014): 467–479.Find this resource:
Denver, D. T., and H. T. G. Hands. “Marginality and Turnout in British General Elections.” British Journal of Political Science 4, no. 1 (1974): 17–35.Find this resource:
Denver, David, Gordon Hands, and Iain MacAllister. “Constituency Marginality and Turnout in Britain Revisited.” British Elections and Parties Review 13, no. 1 (2003): 174–194.Find this resource:
Dhillon, Amrita, and Susana Peralta. “Economic Theories of Voter Turnout.” Economic Journal 112 (2002): 332–352.Find this resource:
Downs, Anthony. An Economic Theory of Democracy. New York: Harper and Row, 1957.Find this resource:
Duverger, Maurice. Political Parties: Their Organization and Activity in the Modern State. New York: John Wiley, 1954.Find this resource:
Eggers, Andrew C. “Proportionality and Turnout: Evidence from French Municipalities.” Comparative Political Studies 48, no. 2 (2015): 135–167.Find this resource:
Endersby, James W., and Jonathan T. Krieckhaus. “Turnout around the Globe: The Influence of Electoral Institutions on National Voter Participation, 1972–2000.” Electoral Studies 27 (2008): 601–610.Find this resource:
Faravelli, Marco, and Santiago Sanchez-Pages. “(Don’t) Make My Vote Count.” Journal of Theoretical Politics 27, no. 4 (2015): 544–569.Find this resource:
Fauvelle-Aymar, Christine, and Abel François. “The Impact of Closeness on Turnout: An Empirical Relation Based on a Study of a Two-Round Ballot.” Public Choice 127, no. 3/4 (2006): 469–491.Find this resource:
Fauvelle-Aymar, Christine, and Michael S. Lewis-Beck. “TR versus PR: Effects of the French Double Ballot.” Electoral Studies 27 (2008): 400–406.Find this resource:
Fisher, Stephen D., Laurence Lessard-Phillips, Sara B. Hobolt, and John Curtice. “Disengaging Voters: Do Plurality Systems Discourage the Less Knowledgeable from Voting?” Electoral Studies 27 (2008): 89–104.Find this resource:
Fiva, Jon H., and Daniel M. Smith. “Local Candidates and Voter Mobilization: Evidence from Historical Two-Round Elections in Norway.” Electoral Studies 45 (2017): 130–140.Find this resource:
Folke, Olle. “Shades of Brown and Green: Party Effects in Proportional Election Systems.” Journal of the European Economic Association 12, no. 5 (2014): 1361–1395.Find this resource:
Fornos, Carolina A., Timothy J. Power, and James C. Garand. “Explaining Voter Turnout in Latin America, 1980 to 2000.” Comparative Political Studies 37, no. 8 (2004): 909–940.Find this resource:
Franklin, Mark. “Electoral Participation.” In Comparing Democracies: Elections and Voting in Global Perspective, edited by L. LeDuc, R. G. Niemi, and P. Norris, 216–235. Thousand Oaks, CA: Sage, 1996.Find this resource:
Gallagher, Michael. “Proportionality, Disproportionality and Electoral Systems.” Electoral Studies 10, no. 1 (1991): 33–51.Find this resource:
Gallego, Aina, Guillem Rico, and Eva Anduiza. “Disproportionality and Voter Turnout in New and Old Democracies.” Electoral Studies 31, no. 1 (2012): 159–169.Find this resource:
Garmann, Sebastian. “A Note on Electoral Competition and Turnout in Run-off Electoral Systems: Taking into Account Both Endogeneity and Attenuation Bias.” Electoral Studies 34 (2014): 261–265.Find this resource:
Geys, Benny. “Explaining Voter Turnout: A Review of Aggregate-Level Research.” Electoral Studies 25, no. 4 (2006a): 637–663.Find this resource:
Geys, Benny. “Rational Theories of Voter Turnout: A Review.” Political Studies Review 4 (2006b): 16–35.Find this resource:
Górecki, Maciej A., and Michael Marsh. “Not Just ‘Friends and Neighbours’: Canvassing, Geographic Proximity and Voter Choice.” European Journal of Political Research 51, no. 5 (2012): 563–582.Find this resource:
Gosnell, Harold. Why Europe Votes. Chicago: University of Chicago Press, 1930.Find this resource:
Grofman, Bernard, and Peter Selb. “A Fully General Index of Political Competition.” Electoral Studies 28, no. 2 (2009): 291–296.Find this resource:
Grofman, Bernard, and Peter Selb. “Turnout and the (Effective) Number of Parties at the National and District Levels: A Puzzle-Solving Approach.” Party Politics 17, no. 1 (2011): 93–117.Find this resource:
Herrera, Helios, Massimo Morelli, and Salvatore Nunnari. “Turnout across Democracies.” American Journal of Political Science 60, no. 3 (2016): 607–624.Find this resource:
Herrera, Helios, Massimo Morelli, and Thomas Palfrey. “Turnout and Power Sharing.” Economic Journal 124, no. 574 (2014): F131–F162.Find this resource:
Hix, Simon, and Sara Hagemann. “Could Changing the Electoral Rules Fix European Parliament Elections?” Politique européenne 2, no. 28 (2009): 37–52.Find this resource:
Huckfeldt, Robert, and John Sprague. “Political Parties and Electoral Mobilization: Political Structure, Social Structure, and the Party Canvass.” American Political Science Review 86 (1992): 70–86.Find this resource:
Indridason, Indridi H. “Competition & Turnout: The Majority Run-off as a Natural Experiment.” Electoral Studies 27, no. 4 (2008): 699–710.Find this resource:
Jackman, Robert W. “Political Institutions and Voter Turnout in the Industrial Democracies.” American Political Science Review 81, no. 2 (1987): 405–424.Find this resource:
Jackman, Robert W., and Ross A. Miller. “Voter Turnout in the Industrial Democracies during the 1980s.” Comparative Political Studies 27, no. 4 (1995): 467–492.Find this resource:
Karp, Jeffrey A., Susan A. Banducci, and Shawn Bowler. “Getting Out the Vote: Party Mobilization in Comparative Perspective.” British Journal of Political Science 38, no. 1 (2008): 91–112.Find this resource:
Kartal, Melis. “A Comparative Welfare Analysis of Electoral Systems with Endogenous Turnout.” Economic Journal 125, no. 587 (2015): 1369–1392.Find this resource:
Karvonen, Lauri. “Preferential Voting: Incidence and Effects.” International Political Science Review 25, no. 2 (2004): 203–226.Find this resource:
Katz, Richard S., and Peter Mair. “Changing Models of Party Organization and Party Democracy: The Emergence of the Cartel Party.” Party Politics 1, no. 1 (1995): 5–28.Find this resource:
Kayser, Mark Andreas, and René Lindstädt. “A Cross-National Measure of Electoral Competitiveness.” Political Analysis 23 (2015): 242–253.Find this resource:
Key, V. O., Jr. Southern Politics in State and Nation. New York: Alfred A. Knopf, 1949.Find this resource:
Kollman, Ken, Allen Hicken, Daniele Caramani, David Backer, and David Lublin. “Constituency-Level Elections Archive.” Produced and distributed by Ann Arbor, MI: Center for Political Studies, University of Michigan, 2016.Find this resource:
Kostadinova, Tatiana. “Voter Turnout Dynamics in Post-Communist Europe.” European Journal of Political Research 42, no. 6 (2003): 741–759.Find this resource:
Ladner, Andreas, and Henry Milner. “Do Voters Turn Out More under Proportional Than Majoritarian Systems? The Evidence from Swiss Communal Elections.” Electoral Studies 18, no. 2 (1999): 235–250.Find this resource:
Lewis-Beck, Michael S., and Tom W. Rice. “Localism in Presidential Elections: The Home State Advantage.” American Journal of Political Science 27, no. 3 (1983): 548–556.Find this resource:
Lijphart, Arend. “Unequal Participation: Democracy’s Unresolved Dilemma.” American Political Science Review 91, no. 1 (1997): 1–14.Find this resource:
Maeda, Ko. “Voter Turnout and District-Level Competitiveness in Mixed-Member Electoral Systems.” Journal of Elections, Public Opinion and Parties 26, no. 4 (2016): 452–469.Find this resource:
Morton, Rebecca B. “A Group Majority Voting Model of Public Good Provision.” Social Choice and Welfare 4, no. 2 (1987): 117–131.Find this resource:
Morton, Rebecca B. “Groups in Rational Turnout Models.” American Journal of Political Science 35, no. 3 (1991): 758–776.Find this resource:
Nemoto, Kuniaki, and Matthew S. Shugart. “Localism and Coordination under Three Different Electoral Systems: The National District of the Japanese House of Councillors.” Electoral Studies 32, no. 1 (2013): 1–12.Find this resource:
Neumann, Sigmund. “Towards a Comparative Study of Political Parties.” In Modern Political Parties, edited by Sigmund Neumann, 395–421. Chicago: Chicago University Press, 1956.Find this resource:
Palfrey, Thomas R., and Howard Rosenthal. “Voter Participation and Strategic Uncertainty.” American Political Science Review 79, no. 1 (1985): 62–78.Find this resource:
Pérez-Liñán, Aníbal. “Neoinstitutional Accounts of Voter Turnout: Moving Beyond Industrial Democracies.” Electoral Studies 20, no. 2 (2001): 281–297.Find this resource:
Powell, G. Bingham. “Voting Turnout in Thirty Democracies: Partisan, Legal, and Socio-Economic Influences.” In Electoral Participation: A Comparative Analysis, edited by Richard Rose, 396–412. Beverly Hills: Sage, 1980.Find this resource:
Powell, G. Bingham. “American Voter Turnout in Comparative Perspective.” American Political Science Review 80, no. 1 (1986): 17–43.Find this resource:
Rainey, Carlisle. “Strategic Mobilization: Why Proportional Representation Decreases Voter Mobilization.” Electoral Studies 37 (2015): 86–98.Find this resource:
Reed, Steven R., and Daniel M. Smith. “The Japanese House of Representatives Elections Data Set.” 2016. https://sites.google.com/site/danielmarkhamsmith/data.
Rice, Tom W., and Alisa A. Macht. “The Hometown Advantage: Mobilization or Conversion?” Political Behavior 9, no. 3 (1987): 257–262.Find this resource:
Rich, Timothy S. “Institutional Influences on Turnout in Mixed Member Electoral Systems: An Exploratory Analysis.” Representation 50, no. 2 (2014): 161–175.Find this resource:
Riker, William H., and Peter C. Ordeshook. “A Theory of the Calculus of Voting.” American Political Science Review 62 (1968): 25–42.Find this resource:
Robbins, Joseph W. “The Personal Vote and Voter Turnout.” Electoral Studies 29 (2010): 661–672.Find this resource:
Rosenstone, Steven J., and John Mark Hansen. Mobilization, Participation, and Democracy in America. New York: Macmillan, 1993.Find this resource:
Sanz, Carlos. “The Effect of Electoral Systems on Voter Turnout: Evidence from a Natural Experiment.” Political Science Research and Methods (2015). First View: doi:https://doi.org/10.1017/psrm.2015.54Find this resource:
Scheiner, Ethan, Daniel M. Smith, and Michael F. Thies. “The 2014 Japanese Election Results: The Opposition Cooperates, but Fails to Inspire.” In Japan Decides 2014: The Japanese General Election, edited by Robert Pekkanen, Steven R. Reed, and Ethan Scheiner, 22–38. New York: Palgrave Macmillan, 2015.Find this resource:
Selb, Peter. “A Deeper Look at the Proportionality-Turnout Nexus.” Comparative Political Studies 42, no. 4 (2009): 527–548.Find this resource:
Shachar, Ron, and Barry Nalebuff. “Follow the Leader: Theory and Evidence on Political Participation.” American Economic Review 89, no. 3 (1999): 525–547.Find this resource:
Shugart, Matthew Soberg, and Martin P. Wattenberg. “Mixed-Member Electoral Systems: A Definition and Typology.” In Mixed-Member Electoral Systems: The Best of Both Worlds?, edited by Matthew Soberg Shugart and Martin P. Wattenberg, 1–24. Oxford: Oxford University Press, 2001.Find this resource:
Simonovits, Gàbor. “Competition and Turnout Revisited: The Importance of Measuring Expected Closeness Accurately.” Electoral Studies 31, no. 2 (2012): 364–371.Find this resource:
Smets, Kaat, and Carolien van Ham. “The Embarrassment of Riches? A Meta-Analysis of Individual-level Research on Voter Turnout.” Electoral Studies 32, no. 2 (2013): 344–359.Find this resource:
Smith, Daniel M. “Party Ideals and Practical Constraints in Kōmeitō Candidate Nominations.” In Kōmeitō: Politics and Religion in Japan, edited by George Ehrhardt, Axel Klein, Levi McLaughlin, and Steven R. Reed, 139–162. Berkeley, CA: Institute of East Asian Studies at the University of California, Berkeley, 2014.Find this resource:
Tillman, Erik R. “Pre-electoral Coalitions and Voter Turnout.” Party Politics 21, no. 5 (2015): 726–737.Find this resource:
Tingsten, Herbert. Political Behavior: Studies in Election Statistics. London: P. S. King & Sons, 1937.Find this resource:
Tullock, G. Toward a Mathematics of Politics. Ann Arbor: University of Michigan Press, 1968.Find this resource:
Uhlaner, Carole J. “Rational Turnout: The Neglected Role of Groups.” American Journal of Political Science 33, no. 2 (1989): 390–422.Find this resource:
Vowles, Jack. “Offsetting the PR Effect? Party Mobilization and Turnout Decline in New Zealand, 1996-99.” Party Politics 8, no. 5 (2002): 587–605.Find this resource:
Vowles, Jack. “Electoral System Change, Generations, Competitiveness and Turnout in New Zealand, 1963–2005.” British Journal of Political Science 40, no. 4 (2010): 875–895.Find this resource:
(2) Several recent articles have reviewed the broader empirical literature on turnout (e.g., Blais 2006; Geys 2006a; Cancela and Geys 2016), as well as the more specific theoretical and empirical literature on the relationship between electoral rules and turnout (Cox 1999; Blais and Aarts 2006; Herrera, Morelli, and Palfrey 2014; Cox 2015; Herrera, Morelli, and Nunnari 2016). These articles cover some of the literature that, for space reasons, is not included here, as well as the impact of other institutional variables, such as compulsory voting, presidentialism versus parliamentarism, concurrent elections, and bicameralism, all of which have inspired additional streams of research.
(3) Multiple empirical studies on single-round SMD elections have confirmed that turnout is higher in close elections (e.g., Denver and Hands 1974; Dawson and Zinser 1976; Caldeira and Patterson 1982; Cox and Munger 1989; Denver, Hands, and MacAllister 2003). Studies on two-round SMD elections also find that close competition in the first round increases turnout in the second round (Indridason 2008; Fauvelle-Aymar and François 2006; Simonovits 2012; De Paola and Scoppa 2014; Garmann 2014; Fiva and Smith 2017).
(4) Herrera, Morelli, and Nunnari (2016) make a related argument in more formal terms. For additional formal theories of electoral systems and turnout, see Kartal (2015) and Faravelli and Sanchez-Pages (2015).
(5) In a separate meta-analysis of individual-level turnout studies (i.e., studies using voter survey data), Smets and van Ham (2013) identify only four journal articles published between 2000 and 2010; of these, only one single-test study found positive results for the relationship between proportional representation systems and voter turnout.
(7) Sanz (2015) refers to this system as “open-list, plurality-at-large.” It is different from traditional open-list PR (OLPR) systems because vote pooling does not occur at the party level, and votes can be cast across party lines.
(9) For the 2001 election, the original Italian data source (the Italian Ministry of the Interior) gives turnout in the SMD tier as 100 percent in one district—Matera in Basilicata Province—with an abnormally large number of “invalid” votes. I estimate the “correct” turnout for Matera as 76.8 percent based on the recorded turnout in the PR tier for Matera voters.
(10) An earlier study by Rich (2014) of cross-national aggregate-level turnout finds no differences between MMM and MMP but suffers from the measurement and causal inference problems already discussed.
(11) Barone and de Blasio (2013) employ an RD design to study Italian mayoral elections, which use FPTP/SMD in municipalities with a population below 15,000 persons, but a TR/SMD system in larger municipalities (among some other variations in rules across municipalities with different population sizes). They find that turnout is higher by 1 percentage point in the latter.
(12) The CLPR system was used from 1983 to 1998. Prior to 1983, SNTV was used to elect all members in the national tier. SNTV has always been used in the prefectural tier, though in some prefectures M = 1, making the system effectively FPTP/SMD.