Marco Castillo received his PhD in agricultural and applied economics from the University of Wisconsin, Madison. His research interests include experimental economics and development economics, especially bargaining and the behavior of groups. He has published in the American Economic Review, Games and Economic Behavior and Economic Inquiry.

"Candidate Canvassing: How Campaigns Win Votes and Influence Elections"

Abstract: We investigate the impact of candidate canvassing on voter turnout and voter choices by implementing an experiment in a recent election for county government in a Midwestern state. To our knowledge, this is the first experiment that evaluates the impact of a candidate’s direct campaigning on voters’ decisions. We randomize the content of the campaign pamphlet delivered to the household and whether the candidate delivered the pamphlet in person. The control group was left untouched. Contrary to the get-out-the-vote literature, we find no evidence that visits by a canvasser (the candidate personally) increase turnout. Importantly, we find that certain messages can depress turnout, whether delivered by the candidate or not, but more so when delivered in person. Turnout among voters that are ex-ante less likely to vote is reduced by 12 percentage points for some treatments, but the effect is negligible among more likely voters. With respect to voter support for the candidate, pamphlets alone have no robust effect. However, candidate canvassing does increase the likelihood of voting for the candidate by roughly 20 percentage points. This effect is largest among unaffiliated or persuadable voters. Consistent with signaling theories of political campaigning, the effect of campaigning on who voters support is largest among those that are ex-ante less likely to vote and less likely to support the candidate.