The Efficacy of Voting
Prepared for the Western Political Science Conference, April 2014
M. Emilee Booth Chapman
Popular voting is central to the current practice of large-scale democracy, and most citizens of established democracies regard voting as the primary practice of citizenship. Voting is the main, and in some cases the only way that most citizens currently participate in the collective exercise of authority over institutions of government. Because of this central focus on voting as the main form of citizen participation, evaluation of contemporary democracies requires an assessment of the meaningfulness of electoral participation as an exercise of citizens’ political authority.
The question of what makes electoral participation meaningful does not yield an obvious answer. The four most commonly invoked standards of electoral efficacy – individual decisiveness, partisan victory (or collective causal efficacy), mandate production, and preference aggregation – all face serious empirical and conceptual challenges and lead many to doubt whether voting can ever be regarded as a meaningful exercise of citizens’ political authority.
In this paper, I argue that voting is efficacious when it allows citizens to contribute to the shared project of democratic self-rule according to their community’s shared plan. Democracy is a collective activity that requires participants to share a plan for how their individual contributions fit together to accomplish their shared goal of collective self-rule. Evaluating the efficacy of electoral participation therefore requires understanding the particular role that voting plays in a contemporary plan for democracy. We should not expect voting to be a microcosm or instantiation of all the values of democracy; it cannot bear the entire normative burden of democracy on its own. Instead, I argue, voting plays a particular role within a broader plan for collective self-rule as a unique form of mass participation that concretely and formally establishes the equal political authority of all citizens over at least a few important decisions.
With this understanding of the role of voting in contemporary democracies in mind, I explain why the most common standards of electoral efficacy fail to evaluate voting in the right way. I then develop a set of conditions for electoral efficacy that are sensitive to the particular role that voting plays in contemporary plans for democracy. In developing these standards, I make use of the common distinction that scholars of voter attitudes make between internal and external political efficacy. I argue that these two attitude constructs reflect two kinds of evaluation of the citizens’ authoritative role in the political system and when considered in an institutionally specific way, both can offer insight into the standards for assessing when voting effectively fulfills its role in a plan for democracy.
These standards are not fully determinate of all electoral institutions. Though voting as a form of participation seems to play a similar role in nearly all contemporary democratic communities, the practice of voting is embedded within different kinds of plans for democracy. The particular way that a democratic community meets the standards for meaningful electoral participation will therefore depend on the particular plans that members of that community have worked out together. The conditions for efficacy I offer here will not settle all debates about the design of electoral institutions, but it can still offer critical purchase for intervening in those debates and in other questions of electoral administration. In the final section of this paper, I briefly discuss how my account of electoral efficacy can illuminate the debates over compulsory voting and limited ballot access.
Section I: The Role of Voting in the Plan for Democracy
Evaluations of voting as a democratic practice tend to think of democracy as an abstract ideal, a collection of evaluative standards that might be applied to any system or institution to assess how closely that institution adheres to those standards.1 But democracy is not just an abstract ideal. It is a collective activity. At its most basic, democracy means that the people rule.2
Understanding democracy not just as a set of ideals, but as a collective activity should affect how we think about evaluating democratic institutions. This is because once we see democracy as a collective activity, we recognize the necessary particularity of democracy’s instantiation.3 Democracy is a thing that the members of a particular political community do together. Annie Stilz and Eric Beerbohm have argued that this understanding of democracy as a collective activity can generate loyalty to and obligations to participate in particular democratic communities.4 Likewise, the philosopher Christopher Kutz has argued that civic and political obligations, like the obligation to vote, arise from individuals’ commitment to a collective project of self-rule.5 Individuals involved in a collective project form “participatory intentions:” they intend to play their part in producing the collective goal.6 Kutz argues that voting enables a citizen to act on her democratic participatory intention: “individual voting is rational, indeed instrumentally rational, because voting is a constitutive element of a collective project that voters can conceive as their project.”7
But seeing voting as a constitutive element of the project of democracy depends on the community’s shared understanding of how they will rule together. As participatory and deliberative democrats have observed, voting is not the only way to make collective decisions. Even assuming a representative system, officials might be selected by lot instead of election. Democracy, as rule by the people, is not fully institutionally determinate. Since there may be lots of ways of doing democracy, the members of a political community need a shared understanding for how they will undertake the collective project of self-rule.
Philosophers like Michael Bratman have pointed to the importance of shared plans for enabling collective action. Bratman argues that one of the conditions for a group of people to share an intention to engage in a particular activity is that each of the members intends to participate according to a set of “meshing sub-plans.”8 If the members of a group are to meaningfully contribute to a collective project, they need to share an understanding of how each of their actions will fit together to produce the shared aim. In his account of the “planning theory of law,” Scott Shapiro emphasizes the importance of shared plans in group life. According to Shapiro, shared plans “guide and coordinate behavior by resolving doubts and disagreements about how to act.”9 If a group of individuals is to accomplish a collective goal, they need a shared understanding of how they will accomplish that goal together.
Like any collective activity, democracy depends on citizens’ sharing a set of plans for how they will accomplish their shared goal of collective self-rule. Citizens need to share an understanding of how their individual acts of participation will fit together to produce democracy. A plan transforms mere expression and meaningless gestures like pulling a lever into political participation and the collective exercise of authority.10
Recognizing the importance of shared plans to democracy allows us to formulate how we should understand individual efficacy in elections. Voting should be understood as efficacious if by voting a citizen is able to contribute to the collective project of self-rule according to her community’s plan for democracy. As Christopher Kutz explains, when individuals act on their “participatory intentions,” their actions derive their meaningfulness from the way that they combine with others’ actions to produce a shared goal. The meaningfulness of an individual’s vote thus depends on the existence of an effective shared plan for democracy that identifies how individual votes constitutively contribute to the collective project of self-rule.
Evaluating the efficacy of voting, then, requires attention to the role that voting plays in particular plans that guide democratic participation in actual communities. This cannot be achieved simply by reasoning from abstract principles of democracy. It is not enough to know what would be the best plan for democracy. As Scott Shapiro argues, the authoritativeness of a plan cannot depend on its being the best plan, since the point of plans is just to settle uncertainty about how to act, in the face of disagreement about which is the best way to proceed.11 Even when citizens doubt whether mass voting (or particular electoral rules) should play an important role in the best plan for democracy, they can still recognize that the collective project of democracy requires a generally shared plan. So long as the existing plan for democracy is more or less effective, that plan still offers citizens the surest guide for how to participate, confident that their contributions will fit together with those of their fellow citizens to produce democracy.12
In this paper, I aim to evaluate the practice of voting on its own terms, looking at the particular role that voting plays in contemporary plans for democracy, and investigating what conditions are necessary for voting to effectively play that role. This is an exercise of constructive interpretation. In Law’s Empire, Ronald Dworkin defines constructive interpretation as “imposing purpose on an object or practice in order to make of it the best possible example of the form or genre to which it is taken to belong."13 Thus, the constructive interpretation I offer here examines the beliefs, principles, and practices that characterize voting in contemporary democracies and identify a purpose that makes the most sense of them as part of a plan for democracy. A constructive interpretation is thus constrained from two ends: by empirical reality – the interpretation needs to “fit” the actual practices I interpret, and by the conceptual requirements of a plan for democracy – the interpretation of the role of voting I offer should be part of a plausibly effective way for the people to achieve collective self-rule.
The project of constructive interpretation then has a further normative component in which existing practices are critically evaluated in light of the best purpose that can be ascribed to them. My principle aim in this paper is to offer a set of conditions that need to be met if voting is to effectively fulfill its role in contemporary plans for democracy. These standards can then be used to evaluate existing practices and provide some guidance for the reform of electoral institutions and practices.
A constructive interpretation of the role of voting in contemporary democracies
Though it is embedded in very different plans for democracy, voting, as a form of participation, plays a similar and central role in most contemporary established democracies.14 Survey data on citizens’ attitudes toward voting and habits of participation, as well as public discourse and common practices reveal that voting plays a unique role as a form of mass participation that concretely and formally instantiates the equal political authority of all citizens in a democracy. Elections reserve certain important decisions for the final political authority: the citizenry as a whole, and it is expected that these important decisions depend on the actual inputs of all citizens. Voting thus identifies a minimum level and form of participation that citizens must engage in if they are to contribute to the project of collective self-rule according to their community’s shared plan for democracy. Because they are such mass moments of participation in which individual influence is minimal, elections draw attention to the distinctively collective character of the democratic project. By voting, citizens do not just contribute to government, but they clearly participate in the democratic project of collective self-rule
Global public discourse and activism reflects the centrality of voting to our understanding of participation in modern democracies. The watchdog group Freedom House keeps a running tally of electoral democracies,15 while the Universal Declaration of Human Rights declares that the authority of governments should be derived from the “will of the people…expressed in periodic and genuine elections” 16 This way of defining democracy does not reflect the triumph of the minimalist claim that democracy just is a regime with elections. The Freedom House and UNDHR definitions of democracy attach a normative value to the institution of elections. They see popular voting as a method for rule by the people.
If we want to assess the efficacy of electoral democracy, though, we need a better understanding of the role that voting plays in a shared plan for collective self-rule. After all, popular voting isn’t conceptually necessary for democracy (perhaps we could have a lottery system to select representative rulers). Nor is voting the only way that citizens actually participate in decision-making today’s democracies. Citizens might also contribute to public deliberation, lobby or petition their representatives, engage in protests or public shaming, donate time or money to a campaign, or even stand for office themselves. As numerous scholars have pointed out, compared to the myriad of other forms of participation, voting is a very “blunt instrument” for influencing policy.17 Popular voting alone surely cannot bear all of the normative weight of the idea of rule by the people. So, just how does voting contribute to rule by the people?
Since a plan for democracy is essentially a shared understanding among citizens about how their participation fits together with that of their fellow citizens to generate democracy, the best place to begin looking for evidence of the role that voting plays in the plan for democracy is the actual beliefs and attitudes that citizens hold toward the practice of democracy.
In surveys, citizens of democracies all over the world overwhelmingly endorse the idea that there is a duty to vote.18 In one of the broadest based surveys of political attitudes, the 2004 ISSP citizenship module, respondents were asked to rate on a seven point scale how important it is for a good citizen to “always vote in election.” In 37 of 40 countries included in the survey, over fifty percent of respondents rated this a 6 or a 7.19 In twelve of these countries, more than 75% of respondents rated always voting in the top two categories. By contrast, the percentage of respondents who rated “keeping watch over the government” in the top two categories of importance exceeded 50% in only twenty-two of the 40 countries studied, and it exceeded 75% in only three. And only in one country (The Phillipines) did more than 50% of respondents rate “being active in social and political organizations” in the top two categories of importance. These results suggest that the widespread belief in the duty to vote is not just one example of a more general duty of political participation. Rather, voting is regarded as a distinctively important kind of participation.20
In depth studies of citizen participation in the United States consistently show support for the distinctiveness of the duty to vote. Russell Dalton finds that support for the duty to vote in the US is more akin to belief in the duty to obey the law than in the importance of other forms of participation.21 Other studies show similar results, with voting ranked very highly on lists of citizen duties – and often freely reported - more in the league with unconditional duties like obeying the law than with other forms of political engagement or political interest.22 Scholars of participation support the distinctiveness of voting by observing that the reasons and background factors that lead people to vote exhibit a completely different pattern from those that lead people to participate in other ways.23
Voting’s distinctiveness seems to arise for two related reasons: 1) the meaningfulness of voting seems tied to the fact that so many citizens actually participate in it, and 2) voting need not be policy or outcome oriented. American survey respondents consistently affirm that it is important to vote even when many other people also vote,24 and In a survey of Canadians, Andre Blais finds overwhelming agreement with the statement "In order to preserve democracy, it is essential that the great majority of citizens vote"25 Widespread participation in voting is not incidental; citizens see mass participation as actually constitutive of the electoral process.
It is probably because voting is meant to be a mass form of participation that it is seen to be less outcome-oriented than other forms of participation. If voting were primarily outcome oriented, then a citizen should only vote when she actually has a preferred outcome to promote, and her voting is likely to contribute to producing this outcome. But the popular understanding of the duty to vote holds that a good citizen should “always vote in elections.”26 If all citizens are supposed to participate in voting, then voting can’t be meaningful only when it is primarily outcome-oriented. Actual voting behavior supports the understanding of voting as distinctly non-policy-oriented. In his broad study of voter participation in the United States, David Campbell finds high voter turnout in very politically homogeneous where electoral outcomes are a foregone conclusion, but the sense of the duty to vote is very strong.27 Chris Achen and Andre Blais likewise find that in predicting voter turnout, belief in the duty to vote has a negative interaction with the effect of caring about the outcome on the decision to vote. How much a person cares about the outcome of an election is generally a strong predictor of the likelihood that she will vote, but its effectiveness as a predictor decreases among survey respondents who believe more strongly in a duty to vote. Individuals who believe there is a duty to vote are not just more likely to actually vote; they actually think about the decision to vote differently.28
The belief in the distinctive importance of voting that these surveys reveal reflects a publicly shared norm. Numerous countries (thirty-one by one count29) have some form of compulsory voting.30 Even where voting is not compulsory, official language often endorses the idea of the duty to vote. The US Citizenship and Naturalization Information document describes voting as “a duty as well as a privilege,”31 and voting is actually listed among the duties of a citizen in the Italian Constitution.32 Our political practices also reflect a belief that widespread participation in voting is distinctively important. Governments go to great pain and expense to ensure that voting is as accessible as possible.33 The special attention that we give to making voting accessible – where other forms of participation notoriously are not – suggests the expectation that everyone participate in voting. Indeed, we wring our hands over voter turnout rates of 60%. If voting were just another kind of participation, this would be astonishing given the fact that other forms of political engagement enjoy much lower rates of participation.34
But voting is not just another form of participation; instead, it plays an important and distinctive role in contemporary democracies as a special form of mass participation. In large-scale contemporary democracies, most decision-making takes place at some remove from “the people.” Though citizens can theoretically influence political decision-making through a variety of channels, in reality, accessing these different channels of influence often requires civic skills and resources that ordinary citizens lack. Moreover, modern political life is tremendously complicated, and citizens might not even know where to begin. In this modern reality of distant decisions making, elections focus citizens’ attention on a set of decisions that they are supposed to make, and offer unique occasions on which the equal political authority of all citizens is made formal and concrete.
The plan for democracy allows for the people to formally and informally delegate a great deal of their decision-making authority, but the plan also reserves a set of important questions to be decided by the final political authority: the citizenry as a whole. The plan offers citizens a clear guide for how their individual actions can contribute to collective self-rule by focusing their attention on a particular set of questions that belong to them, that they are meant to decide. Voting thus acts as a minimum required form of participation. Citizens can (and should) seek to influence political life in many other ways, but if they are to be part of “the people” that rules in a democracy, then they must at least contribute to deciding the questions that have been reserved for the exercise of the people’s collective authority.35
Voting is the clearest way an individual can act, confident that her actions will fit together with her fellow citizens’ to achieve their shared aim of collective self-rule. Because it is such a massive undertaking, voting also makes plain the truly collective nature of the democratic project. When citizens cast their ballot in a polling place or watch electoral returns on the news, it is very difficult to ignore the way that their political influence depends on their fellow citizens. By voting a citizen cannot see herself as just an individual aiming to elect a candidate or influence a policy, but rather, as Christopher Kutz observes, by voting she “orients [herself] around the agency of others.”36 She affirms her role in a collective project of self-rule, a project in which all citizens are equal co-agents.
Scholars often emphasize the symbolic value of voting as a civic ritual, an expression of democratic solidarity.37 But voting is not a mere symbol of democracy. When citizens vote, they actually participate in collective decision-making. When citizens vote, they carry out their part in a plan for democracy. When I vote, I actually contribute to the collective exercise of political authority on equal terms with my fellow citizens. Though my contribution seems only marginal, it is precisely this minimal individual influence (no greater or smaller than anyone else’s) that identifies my action as part of a distinctly democratic project. When I vote, I do not just take part in political rule, but rather, I take part in a particular shared project of ruling together with my fellow citizens.
Voting is efficacious when it allows citizens to participate according to the shared plan for collective self-rule. The efficacy of voting therefore depends on voting’s actually fulfilling its role in the plan for democracy; voting should really be a unique form of mass participation in which a few important decisions depend on the equal political authority of all citizens. In the second half of this paper, I develop some conditions that political systems need to satisfy if voting is to effectively fulfill its role in the plan for democracy. But first, I briefly discuss four commonly used standards of electoral efficacy. I argue that these standards are inadequate because they do not appropriately account for the distinctive role of voting contemporary plans for democracy.
Section II: What efficacy isn’t
Scholars have often been invoked one of four common standards as a way of evaluating when voting is meaningful or worthwhile: individual decisiveness, partisan victory, electoral mandate, and preference communication and aggregation. Each of these standards offers a different way of thinking about the meaningfulness of individual participation in elections, and all have been sharply criticized by political science and positive political theory in the 20th century. I argue that the failure of any of these standards of electoral efficacy to stand up to reality should not be taken as a failure of voting itself, because these standards do not account for the particular role that voting plays in the plan for democracy.
The standard of individual decisiveness as electoral efficacy is the electoral version of rational instrumental reasoning. Because voting entails some cost (at least an opportunity cost), a person should only vote if by voting she can bring about an outcome whose benefit outweighs the cost of voting (this applies both to act-consequentialist and to self-interested rationality). If this is the standard we use to evaluate the efficacy of voting in elections, then we would have to say that voting is almost never efficacious. Elections almost never come down to a single vote. For all except a small handful of local elections, each voter can see that the electoral outcome would be the same regardless of whether she voted or not, so by voting, she effectively wasted the opportunity to act in a way that would make some difference to her well-being or to the common good.
Rational choice theorists’ efforts to rescue the rationality of voting have succeeded by pointing out that an individual can decisively contribute to her own well-being by voting if she enjoys the act of voting itself.38 But this way of recasting the rational efficacy of voting eliminates the critical purchase of the concept of efficacy for evaluating electoral democracies. Efficacy comes to be a function of individual psychology, rather than of the political context.39
One way of rescuing the critical purchase of this individual decisiveness view of efficacy is to suggest that the electoral system should maximize the probability that an individual voter will be decisive. Though officials can certainly look at partisan affiliations and voting trends to draw electoral districts that will be more or less competitive, evaluating the effectiveness of an election based on the probability that an individual will be decisive ignores the importance of the individual choices that determine the probability of decisiveness. It would be strange, I think, to conclude that voting is not an effective form of democratic participation just because a sizeable majority of voters prefers one candidate over the other. Maximizing the probability of individual decisiveness in elections would require constantly redefining electoral districts in a way that would undermine the point of electoral choice. It would also be basically impossible to manipulate this probability for the kinds of offices and ballot questions for which the electorate is predefined.
Though some individual elections might happen to meet the standard of efficacy as individual decisiveness, elections in general can’t be expected to meet this standard. But once we understand the role that voting plays in a democracy, it is clear that individual decisiveness isn’t the appropriate understanding of efficacy against which we assess the meaningfulness of voting. The individual decisiveness approach just misses the way that for citizens committed to democracy, policy pursuit is imbedded within an intention to rule together with others. Voting is not just one way that I can pursue my policy goals. It is the way that my fellow citizens and I have agreed we will make decisions together. In fact, voting is special in that it is meant to be a mass form of participation in which all citizens contribute to deciding a few important questions that are reserved for the citizenry as a whole. Voting makes clear to participants the collective nature of democracy, and the radical implications of engaging in a collective endeavor with millions of their fellow citizens. Demanding individual decisiveness in elections simply denies the role of voting in a democracy as formally establishing the equal political authority of all citizens at once.
Recognizing that individual decisiveness in elections is an impossible standard, some consequentialists have suggested that voting should be understood as efficacious when an individual’s preferred outcome prevails. Contrary to the version of rationality which holds that individuals should avoid paying the opportunity cost of voting unless they can be decisive over outcomes, causal consequentialists argue that individuals should want to be part of a group that is jointly efficacious in bringing about a good outcome.40 A similar logic underlies many competitive understandings of democracy: individual citizens relate to politics through parties or associations of which they are apart. Individuals who contribute to their team’s victory can be said to have efficaciously voted.
This view of efficacy renders the losers’ participation meaningless in hindsight, though, and doesn’t fully capture the way we think about the point of voting. As our exploration of attitudes toward voting revealed, most citizens believe that a person should vote even when the preferred outcome is unlikely to “win” and many even believe that voting does not require having a preferred outcome at all. Partisan causal efficacy clearly can’t be the standard of meaningfulness if voting is important even when an individual does not have a preferred partisan outcome.
Voting can’t be motivated solely by a desire to contribute to good outcomes or to ‘win’ in political competition. As critics of the consequentialist argument for the duty to vote have pointed out, there are lots of ways that an individual can contribute to good outcomes,41 and many more effective ways to gain political influence. What directs people to pursue their political aims by voting is a plan that identifies a set of decisions that citizens are supposed to participate in answering. The plan for contemporary democracy defines voting as a form of mass participation in which decisions are made by all citizens, but this plan clearly requires the participation of the winners as well as the losers.42 The individual citizen who votes for a losing candidate has still participated according to a shared plan for democracy. He has contributed to the collective project of self-rule to exactly the same extent as his fellow citizens.
The partisan victory understanding of the efficacy of voting does not give enough weight to the way that partisan competition and the pursuit of policy goals are embedded within a shared intention to rule democratically with one’s fellow citizens. Even when my preferred outcome does not prevail, I may still have effectively contributed to the shared project of self-rule by participating in the process of collective decision-making according to the plan that I share with my fellow citizens for how we will work out the process of governing together.
The electoral mandate
An account of individual efficacy in elections needs to be attentive to the role of voting in the plan for democracy as a form of mass contribution to a decision that is meant to be truly collective. The standard for individual efficacy can’t be based on “getting one’s way,” but rather should be based on how one’s vote contributes to the collective exercise of authority according to the plan for democracy. One of the more commonly accepted standards for evaluating electoral efficacy, the “electoral mandate” view takes seriously that voting really is a contribution to a collective decision. The electoral mandate account of efficacy evaluates the effectiveness of elections as an instrument of the citizenry as a whole, but this view still neglects the significance of voting’s special role in a plan for democracy.
The idea of electoral efficacy that lies behind the electoral mandate is the idea that when citizens decide which candidate should hold a certain office, they also settle policy questions which, though not themselves on the ballot, are asserted to be implicitly at stake in candidate elections. Upon election, public officials often claim a “popular mandate” to enact the policy program on which they campaigned.
Political theorists and political scientists have raised serious doubts about the idea of an electoral mandate, by observing that citizens vote for candidates on the basis of many different reasons, and that majority support for a candidate may not reflect majority support for any particular component of a policy platform.43 An electoral majority can be constructed from a coalition of minorities who feel very strongly about one or two issues. This led Robert Dahl to characterize pluralist democracy as “minorities rule”44
Despite the dubiousness of interpreting elections as mandates for entire policy programs, the mandate conception of electoral efficacy remains implicit in the assumption that many political theorists and political scientists make that citizens ought to decide who to vote for on the basis of policy issues. But as so many critics have pointed out, voting for public officials is not an effective way of communicating policy preferences.45 Voters are unable to list their reasons on the ballot.
The problem with the electoral mandate view of efficacy is that it tries to make elections do too much. Rather than seeing voting as a particular kind of participation that plays a particular role in the plan for democracy, the electoral mandate view takes the centrality and universality of electoral participation to mean that it must encapsulate the full normative value of democracy. If the people rule, then they must do so entirely by voting. This means that rather than understanding elections as identifying a few important questions that are reserved for the citizens as a whole, elections identify a few questions to serve as proxies for all political decision-making. Because it expects to use a very simple mechanism to enable complex communication on a massive scale, the mandate conception of electoral efficacy is doomed to fail.
It is the simplicity of the electoral decision that makes meaningful mass participation in voting possible. As many scholars have repeatedly observed, voting is an information poor kind of participation. The people can only translate their choice in a small set of narrowly defined questions. Nevertheless, Voting is uniquely structured to enable mass contribution to these few decisions. Elections play a central role in the plan for democracy, because they represent distinctive moments of mass participation in which a few important decisions have been reserved for the citizenry as a whole. By voting, individual citizens participate in making these decisions together with all of their fellow citizens.
The mandate conception of electoral efficacy misunderstands the distinctive role that voting plays in elections, but it also neglects the way that voting is embedded within different communities’ broader plans for democracy. The electoral mandate view assumes a direct translation between the election of public officials and policy decision-making, but different communities’ plans for democracy may imagine different roles for elected officials, and define the relationship between mass electoral decisions policy outcomes in different ways. They also include different roles for non-electoral forms of citizen participation, and more or less formal involvement of intermediate groups like parties. All of these features of a plan for democracy may affect the ways that electoral outcomes actually affect the character of a community’s public life.
Preference Communication and Aggregation
Like the electoral mandate conception, the preference aggregation understanding of electoral efficacy rests on too narrow a conception of how voting might function within the plan for democracy. Unlike the electoral mandate conception, the preference aggregation assumption does not expect electoral decisions to act as proxies for other decisions, but preference aggregation does expect that the question on the ballot will be decided in a particular way. This view holds that voting is efficacious when citizens can successfully communicate their preferences over the question (or choice set) at hand and aggregate them in such a way that an election yields a “social choice” determined solely by voters’ preferences.
The Social Choice Theory of the 1960s and 70s troubled this conception of efficacy by revealing both the impossibility of reliably identifying a unique social preference,46 and the significance of agenda-setting for constraining or even determining outcomes.47 Meanwhile, surveys of voter attitudes have repeatedly revealed that many citizens cannot be said to have political preferences at all.48
These results present a devastating challenge to the preference communication and aggregation account of electoral efficacy. In Liberalism Against Populism, William Riker leverages the results of social choice theory to argue that reality could not support the “populist” interpretation of voting as an expression of the people’s will. Instead, Riker suggests, voting could at best support a “liberal” interpretation – elections allowed for the possibility that people might occasionally be able to kick a tyrant out of office. Meanwhile, scholars who wish to retain the preference aggregation view of democracy labor to demonstrate the limited applicability of certain social choice results,49 and to broaden the way that we think about political preferences.50 On the other hand, deliberative democrats often use these empirical results as part of their argument for shifting the attention and normative weight of democracy away from voting.51 The core of democracy, deliberative democrats argue, should not be about the aggregation of individual preferences, but about that way citizens form their political preferences in a process of shared reasoning with others.
The preference aggregation understanding of electoral efficacy, though, again misses the point of the role that voting plays in a democracy. Democracy is not rule by preferences; it is rule by the people.52 By voting, citizens exercise their agency in making a choice. To vote is an act of will. More importantly, it is an act of will within a particular collective context. When citizens vote, they do not just make a decision, rather, they participate in making a collective decision – within a broader project of collective self-rule.
The preference aggregation view of electoral efficacy reflects an impoverished view of collective life; it takes group decisions to be some kind of predictable mathematical combination of individual decisions. But this neglects how the collective context itself shapes the way that individuals reason and decide. Political participation is participation in a collective project that depends on massively shared agency. Taking part in the process of political decision-making is just a fundamentally different kind of activity than making a decision in circumstances where the individual is sovereign. The shared work of collective self-rule involves a messy mixture of conflict and compromise as citizens develop a sense of the concerns, priorities, and primary disagreements that characterize their shared public life as a community. The actual interactions of its citizens distinguish a community from a random collection of individuals, and the public life that results from these interactions can’t be predicted just by knowing the preferences of the community members.53
Deliberative democrats have long criticized “aggregative” democracy for ignoring the way that individuals’ preferences are shaped in public discussion, and in their renewed focus on deliberative preference formation, they tend to relegate voting to the periphery of public life. But this is a mistake. The collective work of democracy involves not just making collective decisions, but also identifying a plan for how citizens will make these decisions together, and even identifying the questions that they will decide. Communities settle on decision procedures that will allow them to answer the kind of questions that need to be settled in their political life, but at the same time, these procedures influence the kinds of questions or concerns that a community will address publicly.
In contemporary democracies, the political process of developing political agendas, priorities, and preferences takes place in a context in which a few important decisions will be settled by the mass participation of the entire citizenry. These mass moments of participation provide a frame for public discussion and structure the possibilities of political life.54 Even though elections do not reveal some underlying social preference or popular will, they nevertheless play a central and distinctive role in citizens’ collective project of self-rule.