The Decision to Vote or to Abstain
Summary and Keywords
Why voters turn out on Election Day has eluded a straightforward explanation. Rational choice theorists have proposed a parsimonious model, but its logical implication is that hardly anyone would vote since their one vote is unlikely to determine the election outcome. Attempts to save the rational choice model incorporate factors like the expressive benefits of voting, yet these modifications seem to be at odds with core assumptions of rational choice theory. Still, some people do weigh the expected costs and benefits of voting and take account of the closeness of the election when deciding whether or not to vote. Many more, though, vote out of a sense of civic duty. In contrast to the calculus of voting model, the civic voluntarism model focuses on the role of resources, political engagement, and to a lesser extent, recruitment in encouraging people to vote. It pays particular attention to the sources of these factors and traces complex paths among them.
There are many other theories of why people vote in elections. Intergenerational transmission and education play central roles in the civic voluntarism models. Studies that link official voting records with census data provide persuasive evidence of the influence of parental turnout. Education is one of the best individual-level predictors of voter turnout, but critics charge that it is simply a proxy for pre-adult experiences within the home. Studies using equally sophisticated designs that mimic the logic of controlled experiments have reached contradictory conclusions about the association between education and turnout. Some of the most innovative work on voter turnout is exploring the role of genetic influences and personality traits, both of which have an element of heritability. This work is in its infancy, but it is likely that many genes shape the predisposition to vote and that they interact in complex ways with environmental influences. Few clear patterns have emerged in the association between personality and turnout. Finally, scholars are beginning to recognize the importance of exploring the connection between health and turnout.
Elections are at the very core of representative democracy. They are essential mechanisms for holding governments accountable, and they provide one of the main opportunities for exercising democratic citizenship. Voting in elections is one of the most basic ways in which citizens participate in politics. Elections embody the principle of political equality: every eligible citizen has the right to vote and every vote counts equally. But in practice, socio-economic disadvantage and other forms of social inequality translate into unequal participation. If non-voters have systematically different policy preferences and political values, the implications of unequal turnout for the responsiveness of democratic systems are profound. Not surprisingly, then, voter turnout has attracted a good deal of study. Indeed, a meta-analysis of individual-level empirical research on the determinants of turnout in national elections located 90 studies that were published in ten top journals between 2000 and 2010 (Smets & van Ham, 2013). These studies included 176 different independent variables.
There are at least two reasons why the literature presents so many explanatory factors. On the one hand, few studies set out to systematically evaluate competing theories; instead, the typical study focuses on assessing one particular theoretical framework (see Smets & van Ham, 2013). On the other hand, many studies take an “everything-but-the-kitchen-sink approach” to explaining why individual-level turnout varies (Fowler, Baker, & Dawes, 2008, p. 234). The net result is that “the jury is still out on what the foundations of micro-level turnout are (Smets & van Ham, 2013, p. 345).
My intent here is not to present a detailed model of voter turnout. Rather, the purpose is to identify some of the enduring themes that have structured research on voter turnout, to review some of the key debates, and to highlight some new research directions and unanswered questions. Given the staggering number of potential explanatory factors, some limits are needed at the outset. In particular, I will not discuss aggregate-level explanations of variations in voter turnout (see Geys, 2006).
Rational Choice and the Calculus of Voting
One of the most influential—and contested—theories of voter turnout is a model of parsimony. Downs (1957) assumed that people act rationally to maximize their utility. In the context of turnout this means that people will only vote if the return on voting exceeds the cost. The return depends on the voter’s party differential—that is, the difference in expected benefit, or utility, if his or her party rather than another wins—discounted by the probability of casting the decisive vote. The probability of casting the decisive vote is a function of the size of the electorate and the closeness of the election. The most important cost is time. It takes time to learn about the candidates and their stances, and to figure out which one to vote for. It also takes time to deal with the administrative demands of registering to vote, to find out where to vote, and to get to the polling place.
The problem with Downs’ calculus of voting is that, for the vast majority of voters in the vast majority of elections, the probability of casting the decisive vote is virtually nil, and so the cost will almost always exceed the expected benefit. In other words, the rational voter should abstain. Yet, we know that millions of people nonetheless vote.
Downs (1957) proposed a way out of the dilemma: rather than weighing the short-run returns against the short-run costs, some voters will focus on the long-run return. These voters see value in voting per se because they realize that democracy can only survive if people vote. It is not clear, though, that this solves the problem of rational abstention. Just as voters are likely to understand that the probability of casting the decisive vote is trivially small, so they are likely to realize that their one vote is extremely unlikely to ensure that democracy continues.
Riker and Ordeshook (1968) took up the challenge of modifying the calculus of voting to account for the fact that many people vote, even though the costs almost always outweigh the expected benefits. They did so by expanding on Downs’ argument about the value attached to the act of voting per se. Their simple expedient was to factor in the utility people derive from expressing their party preference, complying with the norm of civic duty, expressing support for the political system and so on.
Most of the explanatory power in Riker and Ordeshook’s model was attributable to the sense of duty to vote, leading Ferejohn and Fiorina (1974) to argue that incorporating this factor fundamentally changed Downs’ model. They suggested that the “paradox of not voting” could be resolved by recognizing that a rational choice explanation does not have to be premised on maximizing expected utility. They proposed instead that rational actors seek to minimize the maximum potential regret associated with not voting. This is a much less restrictive decision rule that predicts that many people will vote. People are no longer faced with the impossible task of estimating the probability that their preferred candidate might lose by a single vote if they fail to turn out: “the mere logical possibility of such an event is enough” (p. 535).1
Uhlaner (1989) took a different approach to saving a theory of rational turnout by incorporating the role of group leaders. Her key point was that people do not act as atomized individuals but as members of groups with shared interests such as labor unions. Group leaders provide collective benefits by inducing candidates to adopt policy positions that align with group interests in return for their votes.
Scholars have also sought to salvage a rational choice approach to turnout by developing a theory of expressive voting. This theory holds that people vote because they derive satisfaction from expressing their preferences: “citizens want to express who they are and what they care about” (Engelen, 2006, p. 420). This theory is still explicitly rooted in the assumption of rationality, but it rejects the notion that motivations can only qualify as rational if they are instrumental. Engelen argues that expressive motivations can be characterized as rational if “they are performed on the basis of considerations that the individual judges to be reasons worth acting on” (p. 427). This seems to be a very weak notion of rationality. Indeed, the same explanation could be derived without any reference to rational motivation.
Explanations that incorporate psychic gratifications into a rational choice model can seem ad hoc or even tautologous. Recognizing rational calculation and psychic gratification as alternative motivations for turning out offers a more promising approach. Using surveys and experiments, Blais (2000) shows that some people do decide whether or not to vote on the basis of a rational calculation of costs and benefits. If the benefits outweigh the costs, they go to the polls; if not, they stay home. However, support for the original Downsian model is very qualified. People typically over-estimate the probability that their vote will be decisive. More importantly, they do not appear to weigh the expected benefits by the probability of casting the decisive vote; instead, the probability has an independent effect on turnout. Nonetheless, Blais (2000) finds merit in the rational choice account of turnout. Some people do appear to try to maximize their utility when choosing to vote or to abstain. However, he concludes that many more decide to vote because they derive non-instrumental, expressive benefits from the act of voting. For these people, voting is a matter of civic duty and the personal satisfaction experienced in fulfilling that duty. Rational calculations only come into play when the sense of duty is weak.
What needs to be explained is why some citizens rely on rational calculations and (most) others are motivated by a sense of civic duty. One interesting possibility is that this heterogeneity has a genetic component (Loewen & Dawes, 2012): some people may have a genetic predisposition to pay more attention to social norms. Personality may also be a factor. For example, people who are more altruistic tend to have a stronger sense of duty to vote (Blais & Labbé St. Vincent, 2011). Pre-adult socialization and other environmental influences are also likely to play a role and may very well condition the effects of both genetics and personality.
The Civic Voluntarism Model
The puzzle for rational choice models is why people vote when—from a strict, rational choice point of view—it would be more rational to abstain. From the perspective of the civic voluntarism model, the puzzle is why some people do not exercise this most basic of democratic rights. The model encompasses three broad explanatory factors: resources, psychological engagement with politics, and the mobilizing power of networks (Verba, Schlozman & Brady, 1995). Put simply, people do not vote “because they can’t; because they don’t want to; or because nobody asked” (Verba et al., 1995, p. 269).2
Resources are the most important element of the model. Time, money, and civic skills are the key resources. Early experiences within the home and at school play an important role in people’s ability to acquire these resources; workplace experiences and involvement in voluntary associations and religious institutions are also critical. At first blush, it might seem that the amount of time and money required to vote is trivially small. However, time should not be viewed simply in terms of the time it takes to go to the polling place. Investing time in figuring out where the parties stand and how and where to vote—even taking off time to go to vote—may simply be too much for someone who is working at two jobs to stay afloat financially. Similarly, paying the rent and putting food on the table may leave little to spend on getting informed. Civic skills include communication and organizational skills that are acquired on the job or through involvement in non-political organizations. Being able to write a letter, chair a meeting, or give a speech are valuable skills for some forms of political activity but, unsurprisingly, they do not have a significant impact on the propensity to vote. Two indicators of skills do matter, though—education and vocabulary.
A second important factor is psychological engagement with politics. When it comes to voting (as opposed to other forms of political activity), psychological orientations toward politics are among the strongest predictors. The key orientations include political efficacy, political information, strength of partisanship, and interest in politics. Like resources, psychological engagement with politics is shaped by early experiences within the family and at school, and through non-political experiences as an adult—in the workplace, in voluntary associations, and in religious institutions.
The least important component of the model is political recruitment. Many people vote without being asked. A lack of adequate resources or psychological orientations toward politics that discourage voting are much more serious impediments. Still, people may be encouraged to vote by their family and friends, by colleagues at work or by fellow members of voluntary associations or religious institutions. Indeed, there is a positive association between the number of requests for political activity at work, in voluntary associations, and in religious institutions and turnout.
As Verba et al. (1995) argue, the civic voluntarism model is an improvement over the “standard SES model” that focuses on socioeconomic characteristics. Socioeconomic status (SES) certainly matters. Many studies have reported significant associations between both income, education, and voting (Smets & van Ham, 2013). The civic voluntarism model fleshes out the explanatory paths that account for these empirical associations. The model also highlights the role of orientations toward politics, especially political interest. Subsequent work has confirmed that political interest appears to be the single best individual-level predictor of propensity to vote (Blais & Labbé St. Vincent, 2011). Importantly, the civic voluntarism model provides insight into some of the factors that shape political interest and other psychological orientations toward politics.
Age and Political Socialization
The civic voluntarism model emphasizes the role of pre-adult experiences within the family and at school in fostering a propensity to be politically active (or not) as an adult. Verba et al. (1995) uncovered a significant association between experiencing political stimulation in the home and turnout. People who grew up in families where the parents were politically involved and politics was discussed are significantly more likely to vote. However, parental influence may not endure. Plutzer (2002) finds that parental turnout is the most important predictor of whether or not young adults will vote in their first election. However, the effect of parental turnout diminishes over time, and neither parental turnout nor parental education or parental political knowledge influences how young adult turnout evolves in subsequent elections. The only parental attribute that influences the evolution of turnout is the strength of parental partisanship.
Lower turnout is typical of young adults in a wide range of countries. This is evident when turnout rates are compared across age groups using data from module four of the Comparative Study of Electoral Systems (CSES) project. The figure presents data for all of the CSES countries that held national elections between 2011 and 2013. The difference in turnout between the under-30-year-olds and those aged 30 to 44 is over 20 points in Austria, Japan, New Zealand, Taiwan, and Ireland. Elsewhere, the turnout gaps are smaller, but even in Australia, where voting is compulsory, the under-30s are the least likely to vote.
It is typically assumed that the relationship between age and turnout is curvilinear, with turnout declining as old age begins to take its toll (Smets & van Ham, 2013). However, this pattern does not appear consistently across countries. In Mexico and Switzerland, the oldest voters actually have the highest turnout, and there is little evidence of declining turnout in this age group in Australia, Germany, Iceland, Japan, New Zealand, and Thailand.
The 20’s and early 30’s are characterized by some important life transitions—finishing school, getting a job, getting married, becoming a parent, buying a home—that might plausibly influence the propensity to vote. Highton and Wolfinger (2001) looked at the first seven years of the political life cycle and conclude that the assumption of adult roles does not consistently boost turnout; we need to look elsewhere for an explanation of turnout in young adulthood. Leaving school has a large negative effect on turnout; residential mobility also has a negative effect. Entering the workforce and leaving home have modest positive effects, and getting married or buying a home have minimal effects. Even taking account of all of these adult roles, age still has an independent effect, suggesting that turnout may increase in early adulthood as a result of political experience.
Plutzer (2002) has proposed a developmental framework to explain how turnout evolves in young adulthood and how people become habitual voters. He argues that we need to differentiate the factors that influence newly eligible voters to vote (or not) in their first election from those that motivate nonvoters to become habitual voters and those that disrupt the voting habit. The costs of voting are likely to be paramount in the first-eligible election, which is why parental attributes matter and why young adults typically turn out at lower rates than older adults. As young adults acquire resources, the costs of voting go down, and they transition from being nonvoters to being habitual voters. The more resources they accumulate, the faster the transition is likely to be. Some people, of course, will never acquire sufficient resources. These people are more likely to be the habitual nonvoters.
Looking at differences in turnout among age groups in a single election risks confounding life-cycle effects with generational differences. Differences in formative experiences can have long-lasting effects on the turnout of a generation. Franklin (2004) cites examples such as the lowering of the voting age and the introduction of compulsory voting. However, short-term factors can also leave a lasting “footprint” on the turnout of a generation. Franklin highlights the competitiveness of the first few elections in which young adults are eligible to vote. The more competitive the election, the greater the efforts of political parties to mobilize voters and the more likely young voters are to develop the habit of voting. Conversely, the less competitive their first few elections, the more likely people are to become habitual non-voters. Older voters are less influenced by the competitiveness of elections because they have already developed the habit of voting.
Understanding why younger voters are less likely to vote is essential from a policy perspective, and so is disentangling generational from life-cycle effects. Generational effects are likely to prove relatively intractable, but life-cycle effects can be countered through initiatives to reduce the start-up costs of voting.
Education and Turnout
Along with age, education is the most commonly used predictor of voter turnout (Smets & van Ham, 2013); the higher people’s level of educational attainment, the more likely they are to vote. There are a number of reasons why education should enhance turnout. First, education imparts cognitive skills that can reduce the costs of voting. Acquiring information about politics and figuring out how to get registered and where to vote are easier for better-educated people. Second, greater educational attainment likely leads to jobs that encourage the development of civic skills. Third, education fosters civic-mindedness and a sense of civic duty. As a result, highly educated are more likely to feel an obligation to vote. Fourth, education is likely to enhance people’s sense of personal political competence and their interest in politics. The more education people have, the more likely they are to believe that they can be effective political actors, and the greater the perceived benefit of voting. Finally, education fosters adherence to democratic norms.
These causal mechanisms are all typical of the “absolute” education model. In other words, they are not dependent on the level of education that other people have attained but simply on the absolute number of years of schooling completed. According to the “relative” education model, on the other hand, what matters is a person’s level of education relative to others (Nie, Junn, & Stehlik-Barry, 1996). Nie and his colleagues argue that relative education determines a person’s “social network centrality.” People who are more educated than average are more attractive as network members. As a result, they are more likely to have large and diverse social networks, which increases the odds that they will be mobilized to vote.
The relationship between education and turnout is typically considered to be one of the best-established generalizations in the political behavior literature, but much of the research on turnout has been conducted in the United States. We cannot assume that similar patterns necessarily hold elsewhere. Cross-national research reveals that the relationship between education and turnout is actually quite variable (Gallego, 2015). It is strong in some countries (such as Canada, Switzerland, and the United States), weak or non-existent in others (such as Bulgaria, Chile, and Denmark), though in most countries, the more educated do have higher turnout rates than the less educated. Gallego (2015) attributes the variation to contextual factors that influence the costs and benefits of voting. Her central argument is that the turnout of the less educated is disproportionately affected by factors that increase the cognitive costs of voting. These include complex voting and registration procedures, frequent coalition governments, and media systems that are market-oriented.
The Uncertain Causal Status Education
Plausible as the explanatory purchase of education might appear to be, its causal status has been challenged by studies claiming that education is merely serving as a proxy for pre-adult experiences rooted in the family. The difficulty in trying to establish causality is selection bias: both educational attainment and turnout in adulthood may be influenced by the very same factors. The gold standard for demonstrating causality is a randomized experiment, but obviously it is impossible to randomly assign people in childhood to attain different amounts of schooling. Instead, scholars have adopted three main strategies to mimic the logic of controlled experiments and to estimate the impact of educational attainment, over and above that of factors that could explain both educational attainment and the propensity to vote.
The first approach takes advantage of natural experiments. These have typically involved changes in compulsory schooling laws. Such changes offer a way of capturing exogenous sources of individual-level variation in educational attainment that are unrelated to turnout in adulthood. On balance, these studies suggest that education does not have a causal effect on turnout (see, for example, Pelkonen, 2012). However, these studies are only estimating the impact of additional years of secondary schooling for those who would otherwise have dropped out of school. Berinsky and Lenz (2011) take advantage of a different type of natural experiment to examine the impact of post-secondary education. The Vietnam-era draft lotteries in the United States meant that young men were drafted on a random basis, but they could defer military service by continuing their education. Berinsky and Lenz conclude that this externally induced increase in college enrolment did not boost turnout. However, their study does not consider factors such as family background, cognitive ability, and personality traits that might have influenced which individuals avoided conscription by attending college.
A second strategy for minimizing selection bias is to use matching. These methods mimic the logic of controlled experiments by creating “treatment” and “control” groups that are equivalent except for their level of educational attainment. Each individual in the “treatment” group is matched with one or more individuals who have less education but are otherwise similar. Given that the groups are similar except for their educational attainment, any difference in their turnout should, in principle, reflect the difference in levels of education.
Studies using matching methods have yielded conflicting findings, depending upon the matching method and/or the data used.3 Some authors conclude that attending college has no causal impact on political participation; others confirm the conventional wisdom that the relationship between education and turnout is causal (see the exchanges between Kam & Palmer, 2011, and their critics). Perhaps the most persuasive use of matching methods is the analysis by Persson (2014), of a British cohort study that tracked every person born during the same week in 1970. This study allows for matching on three critical characteristics—childhood cognitive ability, parental education, and cultural experiences in childhood. All three prove to be powerful predictors of both educational attainment and political activity, lending weight to the argument about the importance of pre-adult experiences.
With any matching method, though, concerns necessarily remain about the impact of sample attrition, given the difficulty of finding matches. To take an obvious example, very few high school graduates are going to have the same occupational status or household income as a college graduate. As a result, there will be many unmatched individuals who have to be omitted from the analysis. Moreover, with so many different potential predictors of voter turnout—many of which correlate with education—achieving equivalent groups is always going to be challenging.
The study that comes closest to mimicking the logic of a randomized experiment uses a series of field experiments that evaluated the effectiveness of different interventions on high school graduation rates. The children who participated were assigned randomly (or nearly randomly) into treatment and control groups. Sondheimer and Green (2010) tracked down these children and collected data on their turnout history. They conclude that, “educational attainment profoundly affects turnout” (p. 185). However, as the authors note, the interventions were primarily aimed at children from impoverished families, and so additional years of schooling may have had much more impact on their turnout than might be the case for children from less socially disadvantaged backgrounds.
Another design that might help to clarify the causal status of education is based on siblings. Demographers and public health researchers commonly use sibling-based designs to provide an approximation to controlled experiments: siblings from the same family serve, in effect, as each other’s control group. Comparing the turnout of siblings who have different levels of education but share the same family background offers a powerful way of estimating the effect of education on turnout beyond shared family background. Yet there is little systematic sibling-based research on turnout.
Studies using sophisticated designs have reached contradictory conclusions. Determining whether education has a causal impact on turnout and, if so, which level of education matters, is critically important from a normative perspective. Non-voting parents tend to have non-voting children, but education could potentially disrupt the process of intergenerational transmission.
Genes and Turnout
Evidence is emerging that at least part of the variation in turnout may be attributable to genetic factors. Using twin study data, Fowler et al. (2008) have shown that genetic differences account for a significant amount (as much as 50%) of the variation in voting. Their twin study revealed that identical twins are more likely to behave similarly by both voting or both abstaining than non-identical same-sex twins. Given that identical (monozygotic or MZ) twins share 100% of their genes, whereas non-identical (dizygotic or DZ) twins share only 50%, on average, this suggests that there is an important element of heritability in the predisposition to vote. The most common criticism of twin studies challenges the assumption that identical and fraternal twins possess the same proportion of shared environment, relative to heritability and unshared environment. However, there is a good deal of evidence that the equal environments assumption is warranted (see, for example, Fowler & Dawes, 2008). Notably, studies of twins raised apart have replicated the findings of studies based on twins reared together.
An important limitation of twin studies is that they cannot identify which genes or combination of genes might be implicated in the act of voting; they can only indicate that the propensity to vote is heritable. Candidate gene studies compare the frequency of likely genes in respondents who turn out to vote with their frequency in nonvoters. Two candidate genes have been explored to date. Fowler and Dawes (2008) have shown that the MAOA (monoamine oxidase A) and 5HTT (serotonin transporter) genes—genes that are known to influence sociability and pro-social behavior—are significantly associated with turning out to vote. However, the association between the 5HTT gene and voting only holds for people who attend a place of worship, which underlines the importance of environmental factors in conditioning genetic effects.
Fowler and Dawes’ work has come under intense critical scrutiny. Charney and English (2012) have taken issue with the data, the operationalization of turnout, insufficient allowance for the possibility that genetic effects may vary across population groups, the classification of genotypes, and the non-independence of cases and controls (many of the respondents are sibling pairs). Some of these criticisms would apply to a host of studies of voter turnout: many studies have had to rely on self-reported turnout, and it is standard practice to treat voting as a dichotomous variable (voted/did not vote) (Deppe, Stoltenberg, Smith, & Hibbing, 2013; Fowler & Dawes, 2013). Subsequent studies addressing these criticisms have been unable to replicate Fowler and Dawes’ results (Charney & English, 2012; Deppe et al., 2013; but see Fowler & Dawes, 2013). This inability to fully replicate the original findings is common in candidate gene association studies, which frequently produce inconsistent results.
As critics and proponents alike emphasize, there are likely to be many genes implicated in a complex behavior like turning out to vote (Charney & English, 2012; Fowler & Dawes, 2013). A multiplicity of genes, each of which has a small effect, is likely to result in false positives. Given that the results of any one study are necessarily only suggestive, replication studies have a critical role to play.
Identifying genetic influences is only the first challenge. The next challenge will be to trace the causal paths through which they affect the propensity to vote. One interesting possibility is that party identification plays a mediating role. Dawes and Fowler (2009) have shown that the association between the A2 allele of the D2 dopamine receptor gene and self-reported turnout is mostly indirect, via the gene’s effect on partisanship. Another promising candidate is the sense that voting is a duty. Based on their twin study, Loewen and Dawes (2012) report that heritability accounts for as much as a third of the variation in this civic norm. Finally, genetic variation in personality traits may help to explain the connection between genes and voting. Studies have found that about 50%of the variation in personality traits can be attributed to heritability (see Mondak, Hibbing, Canache, Seligson, & Anderson, 2010).
The effort to understand the genetic influences on voter turnout is only just beginning. However, it holds out the promise of deepening our understanding of the fundamental roots of differences in the propensity to vote (see Fowler & Dawes, 2008). For example, we know that church attendance is associated with higher turnout (Verba, Schlozman, & Brady, 1995), but regular church attendance is no guarantee that someone will vote. The interaction between church attendance and 5HTT, identified by Fowler and Dawes (2008), can help explain why some attendees are more likely than others to vote. Similarly, understanding the role of genes can contribute to debates about political socialization. Young adults are more likely to vote if their parents are voters, but does this simply reflect learning by example as typically assumed, or do shared genes also play a role in inter-generational transmission? Genes may also help to explain why some people are habitual voters while others are habitual nonvoters, and why some people decide to vote based on a comparison of costs and benefits while others are motivated by psychic gratifications.
Personality and Turnout
The link between genes, personality traits, and turnout has yet to be explored, but several studies have analyzed the connection between personality traits and voting. Most of these studies have focused on the “Big Five” traits the psychology literature identifies as capturing the key dimensions of variation in personality: extraversion, agreeableness, conscientiousness, emotional stability, and openness to experience. The results are decidedly mixed (see studies discussed in Gerber, Huber, Doherty, & Dowling, 2011).
Individuals high on extraversion are confident and sociable. As such, they tend to have high levels of personal efficacy and large social networks, both of which should enhance the likelihood of voting. However, studies have mostly come up with null findings, perhaps because voting is not a social activity.
It is unclear whether—and how—agreeableness might influence turnout. People high on agreeableness might be less likely to vote because they are turned off by the conflictual nature of election campaigns. On the other hand, they might be more likely to vote because altruism is an important aspect of agreeableness, and altruistic people are more likely to vote (Blais & Labbé St. Vincent, 2011). Given these opposing influences, it is not surprising that studies have failed to find a systematic connection between agreeableness and turnout.
Being conscientious is almost synonymous with being dutiful, and there is ample evidence that people vote out of a sense of duty. However, the effect of conscientiousness depends on the perceived importance of the election. The relationship is positive for those who think campaign activity is important but negative for those who consider it unimportant (Mondak et al., 2010). This pattern is attributed to another aspect of conscientiousness—a concern with instrumental benefits. If the benefits of voting are perceived to be minimal, conscientious people may prefer to devote their time and energy to more pressing duties.
The basis for expecting any relationship between emotional stability and turnout is unclear. It has been suggested that emotional stability might make it easier for people to deal with the conflictual nature of politics, but conflict avoidance appears to be unrelated to the propensity to vote (Blais & Labbé St. Vincent, 2011) and studies of the relationship between emotional stability and turnout have yielded inconsistent results.
People who are high on openness to experience tend to be more interested in politics and to possess high levels of political knowledge, as well as larger social networks (Gerber et al., 2011), all of which should enhance the willingness to vote. Yet, studies have mostly failed to detect any association between this trait and turnout (but see Mondak et al., 2010).
In light of all these mixed findings, the safest conclusion to draw is that there is no simple one-to-one connection between personality traits and turnout. Some of the inconsistencies across studies probably reflect the fact that they use different batteries to measure the Big Five traits (Gerber et al., 2011). The Big Five are multifaceted, and different batteries tap into different facets of the same broad trait. It could well be that different facets of the same trait have different effects on turnout. Another possible reason for conflicting findings is variation in the electoral context (Gerber et al., 2011). Some election outcomes may seem to be a foregone conclusion, and so a sense of duty may be particularly important; some campaigns may be very negative, and so conflict avoidance becomes a more salient consideration. Sample composition may also be a factor since the effects of traits may vary depending on people’s social background characteristics.
A further complication is the many possible causal paths that could plausibly link personality traits to turnout. A number of mediating variables, including political interest, civic duty, internal political efficacy, party identification, and external political efficacy, have been identified (see, for example, Blais & Labbé St. Vincent, 2011; Schoen & Steinbrecher, 2013). The complexity of the relationship between personality traits and turnout is compounded by the many factors that are likely to condition the connection (Schoen & Steinbrecher, 2013). Mondak et al. (2010) may well be right that we can only hope to understand how personality influences behavior if we take account of the possible interactions between traits and environmental influences. Importantly, doing so is also likely to cast light on why those environmental influences matter more for some people than for others.
The Impact of Health
Health, like personality, reflects both biological and environmental influences. Health disparities may well translate into unequal turnout. Ill health may make it difficult to acquire the resources that facilitate turnout; it may sap the mental energy required to pay attention to politics; and it may even limit the physical ability to cast a ballot. Health problems may also mean fewer opportunities to be mobilized to vote if poor health decreases involvement in social activities and voluntary associations.
Yet, until recently, political scientists have paid little attention to a possible connection between health and turnout. None of the individual-level studies of voter turnout included in the meta-analysis of Smets and van Ham (2013) address physical health, and only three look at mental health. This neglect is surprising since a health bias in voter turnout may compromise policy responsiveness. However, political scientists have recently begun to explore how health affects turnout. Pacheco and Fletcher (2014) find that people who report enjoying excellent health are significantly more likely to vote, and Mattilo et al. (2013) present data covering 30 European countries, showing that poor self-reported health depresses turnout, particularly among older people.
There is much that we still do not know about the impact of health on turnout. Studies to date have provided evidence of an empirical association, but it is unclear whether the association is causal. A major challenge for political scientists, therefore, is to provide a deeper understanding of the causal connections between health and turnout (see Pacheco & Fletcher, 2014). Plausible mediating variables include resources, motivations, and social networks. A diminished sense of political efficacy and trust in government may also play a role.
Little is known about the effects of conditioning factors. The relationship between health and turnout could well vary across countries depending on the welfare regime. The direction of the possible effects, though, is far from clear. Generous welfare provision could narrow turnout gaps between the healthy and the unhealthy by equalizing opportunities and lowering the costs of voting, but it is also possible that a lack of adequate social supports could narrow health-related turnout gaps in less generous welfare systems by mobilizing citizens with poor health.
Voting is a deceptively simple act. It turns out that there are many factors that can influence whether people decide to vote or to abstain. The search for a parsimonious model that provides a satisfactory explanation of individual-level variation in turnout has, so far, proved elusive. The calculus of voting model and the civic voluntarism model (which subsumes the SES model) are the best elaborated. The calculus of voting model relies on a very small number of explanatory factors but fails empirically unless additional factors are incorporated that arguably run counter to the model’s core assumption. The civic voluntarism model incorporates many more explanatory factors and traces complex causal paths among them, but this necessarily makes it difficult to test the complete model empirically.
Both models recognize the costs of voting. Yet, many people who have the time, money, and cognitive resources do not vote, and there are many people who are resource-poor who nonetheless do vote. If we want to understand why we need to explore the deeper roots of unequal turnout. Work on the role of genetics and personality is only just beginning, but it holds out the promise of contributing to a fuller understanding of why some vote and others stay home.
Perhaps precisely because turnout has proved difficult to explain it has spurred some innovative and methodologically sophisticated approaches. Studies of voter turnout have been at the forefront of the new interest in using laboratory experiments and field experiments to study political behavior and in using matching methods to mimic the logic of controlled experimentation and establish causality. Similarly, research on the influence of genes and personality is pushing the boundaries of political science.
Developing a fuller understanding of the factors that shape the predisposition to vote or to abstain is important. The right to vote in free and open elections is a defining characteristic of representative democracy, and yet many citizens fail to exercise this most basic democratic right. One clear conclusion from decades of research on voter turnout is that there are systematic biases in who votes and who stays home. The young, the poor, and the least educated are the least likely to vote. The social biases in turnout would not matter if voters and non-voters shared the same political preferences and issue attitudes, but this is clearly not the case. Voters are more liberal than non-voters on social and cultural issues, but they are more conservative than non-voters on economic issues (Gallego, 2015; Leighley & Nagler, 2014). The more unequal voter turnout is, the greater the differences in political opinions between voters and non-voters. In fact, where there are few systematic biases in turnout, there are few, if any, differences in political opinions (Gallego, 2015). Policy makers will pay the most attention to those who vote. If voters are less likely than non-voters to favor economic redistribution, the implication is clear—unequal turnout is likely to reinforce economic inequality.
Berinsky, A. J., & Lenz, G. S. (2011). Education and political participation: Exploring the causal link. Political Behavior, 33, 357–373.Find this resource:
Blais, A. (2000). To vote or not to vote: The merits and limits of rational choice theory. Pittsburgh: University of Pittsburgh Press.Find this resource:
Blais, A., & Labbé St. Vincent, S. (2011). Personality traits, political attitudes and the propensity to vote. European Journal of Political Research, 50, 395–417.Find this resource:
Charney, E., & English, W. (2012). Genopolitics and the science of genetics. American Political Science Review, 106, 1–34.Find this resource:
Dawes, C. T., & Fowler, J. H. (2009). Partisanship, voting, and the dopamine D2 receptor gene. The Journal of Politics, 71, 1157–1171.Find this resource:
Deppe, K. D., Stoltenberg, S. F., Smith, K. B., & Hibbing, J. R. (2013). Candidate genes and voter turnout: Further evidence on the role of 5-HTTLPR. American Political Science Review, 107, 375–381.Find this resource:
Downs, A. (1957). An economic theory of democracy. New York: Harper and Row.Find this resource:
Engelen, B. (2006). Solving the paradox: The expressive rationality of the decision to vote. Rationality and Society, 18, 419–441.Find this resource:
Ferejohn, J. A., & Fiorina, M. P. (1974). The paradox of not voting: A decision theoretic analysis. American Political Science Review, 68, 525–536.Find this resource:
Fowler, J., Baker, L., & Dawes, C. (2008). Genetic variation in political participation. American Political Science Review, 102, 233–248.Find this resource:
Fowler, J., & Dawes, C. (2008). Two genes predict voter turnout. Journal of Politics, 70, 579–594.Find this resource:
Fowler, J., & Dawes, C. T. (2013). In defense of genopolitics. American Political Science Review, 107, 362–374.Find this resource:
Franklin, M. (2004). Voter turnout and the dynamics of electoral competition in established democracies since 1945. New York: Cambridge University Press.Find this resource:
Gallego, A. (2015). Unequal political participation worldwide. New York: Cambridge University Press.Find this resource:
Gerber, A. S., Huber, G. A., Doherty, D., & Dowling, C. M. (2011). The big five personality traits in the political arena. Annual Review of Political Science, 14, 265–287.Find this resource:
Geys, B. (2006). Explaining voter turnout: A review of aggregate-level research. Electoral Studies, 25, 637–663.Find this resource:
Highton, B., & Wolfinger, R. E. (2001) The first seven years of the political life cycle. American Journal of Political Science, 45, 202–209.Find this resource:
Kam, C. D., & Palmer, C. L. (2011). Rejoinder: Reinvestigating the causal relationship between higher education and political participation. The Journal of Politics, 73, 659–663.Find this resource:
Leighley, J., & Nagler, J. (2014). Who votes now? Demographics, issues, inequality, and turnout in the United States. Princeton, NJ: Princeton University PressFind this resource:
Loewen, P. J., & Dawes, C. T. (2012). The heritability of duty and voter turnout. Political Psychology, 33, 363–373.Find this resource:
Mattilo, M, Söderlund, P., Wass, H., & Rapeli, L. (2013). Healthy voting: The effect of self-reported health on turnout in 30 countries. Electoral Studies, 32, 886–891.Find this resource:
Mondak, J. J., Hibbing, M. V., Canache, D., Seligson, M. A., & Anderson, M. R. (2010). Personality and civic engagement: An integrative framework for the study of trait effects on political behavior. American Political Science Review, 104, 85–110.Find this resource:
Nie, N., Junn, J., & Stehlik-Barry, K. (1996). Education and democratic citizenship in America. Chicago: Chicago University Press.Find this resource:
Pacheco, J., & Fletcher, J. (2014). Incorporating health into studies of political behavior: Evidence for turnout and partisanship. Political Research Quarterly, 68, 104–116.Find this resource:
Pelkonen, P. (2012). Length of compulsory education and voter turnout: Evidence from a staged reform. Public Choice, 150, 51–75.Find this resource:
Persson, M. (2014). Testing the relationship between education and political participation using the 1970 British Cohort Study. Political Behavior, 36, 877–897.Find this resource:
Plutzer, E. (2002). Becoming a habitual voter: Inertia, resources, and growth in young adulthood. American Political Science Review, 96, 41–56Find this resource:
Riker, W., & Ordeshook, P. (1968). A theory of the calculus of voting. American Political Science Review, 62, 25–42.Find this resource:
Schoen, H., & Steinbrecher, M. (2013). Beyond total effects: Exploring the interplay of personality and attitudes in affecting turnout in the 2009 German federal election. Political Psychology, 34, 533–552.Find this resource:
Schlozman, K. L, Verba, S., & Brady, H. E. (2012). The unheavenly chorus: Unequal political voice and the broken promise of American democracy. Princeton, NJ: Princeton University Press.Find this resource:
Smets, K., & van Ham, C. (2013). The embarrassment of riches? A meta-analysis of individual-level research of voter turnout. Electoral Studies, 32, 344–359.Find this resource:
Sondheimer, R. M., & Green, D. P. (2010). Using experiments to estimate the effects of education on voter turnout. American Journal of Political Science, 54, 174–189.Find this resource:
Uhlaner, C. J. (1989). The neglected role of groups. American Journal of Political Science, 33, 390–422.Find this resource:
Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in American Politics. Cambridge, MA: Harvard University Press.Find this resource:
(1.) However, the minimax regret decision rule implies other logical possibilities: “should not the minimax regretter consider the possibility that she could get killed by a car going to the polls? Is not the appropriate decision, then, not to vote?” (Blais, 2000, p. 6).
(2.) Note that this model was developed to explain political participation in general. However, Verba et al. (1995) analyze different forms of participation, including voting, and the model is frequently invoked to explain individual-level variation in turnout.
(3.) Note that the dependent variable in these studies is political participation, but self-reported turnout is one of the items that are used to create the participation measures.