Do people vote for consequentialist reasons, for non-consequentialist reasons, or both? A large body of theoretical research in political economy assumes that individuals vote in order to affect the outcome of an election. In these models, individuals should be more likely to turn out when the election is close. On the other hand, the chances of an individual actually being pivotal in an election is vanishingly small. If consequentialism were the primary motivation for political participation, we should see very low levels of turnout. While researchers have examined observational data, confounds make conclusions from this data difficult.
This research uses an experimental methodology in which we test whether informing people about the closeness of the election affects turnout. According to these models, individuals should be more likely to vote when the election is close. If our findings do not support these models, then research might develop models where individuals have non-consequentialist (e.g. emotional, social norm, ...) motivations for political participation.
The broader impacts of the study include adding to existing knowledge about "what works" in increasing democratic participation, and in building intellectual bridges between political science and economics.
Why people vote is a fundamental question. In leading theory models in economics and political science, it is conjectured that individuals are motivated to vote so as to affect the outcome of the election. However, since an individual vote will only affect an election outcome in the case of a tie, the probability of casting such a decisive or pivotal vote grows vanishingly small in large elections. Despite this, evidence from psychology and behavioral economics suggests that people sometimes dramatically overestimate the probability of rare events. Finally, information is hypothesized to play a crucial role in the voting decision. When a voter learns that he or she is more likely to cast a decisive vote, theory suggests that such information will change the decision about whether and for whom to vote. The goal of this research is to examine the empirical validity of canonical pivotal voter models. To test these models, we conducted a field experiment in the 2010 gubernatorial elections with over 16,000 voters where we provided different information about the closeness of the elections, exploiting large differences across polls. Voter beliefs and voting intentions are elicited before and after the provision of information. We find that voters enormously overestimate the probability of an extremely close election, which is consistent with evidence from psychology but inconsistent with the classical pivotal voter models. Consistent with theory, voters update their beliefs substantially in response to new information. However, the experiment has no effect on turnout or vote choice. Even in a controlled setting whether voters believe that they may be pivotal and in response to an intervention that significantly changed their perceived probability of being pivotal, voting behavior is unaffected by beliefs about the probability of casting a decisive vote. Our results have important implications both for theory and for policy. For theory, we fail to find evidence to support pivotal voter models—simply put, there is nothing in the evidence to suggest that voters are primarily motivated to vote by the chance of casting the decisive vote influencing an election. The evidence is more consistent with expressive models where voters are conjectured to derive benefit from the process of casting a vote, regardless of its influence on the outcome of the election. From a policy perspective, offering voters more precise information that their vote will "make a difference" (i.e. swing the outcome of an election) is unlikely to produce increases in voter participation. Instead, increasing turnout may require appeals to civic duty or appeals to the value of expressing one's political opinion in the Democratic process.
