Does Costly Signaling Matter? Preliminary Evidence from a Field Experiment

Abstract

This paper presents a preliminary experiment designed to determine whether costly signaling plays a role in political interactions. Drawing on the expansive signaling literature in international relations and elsewhere, we propose that the quality of solicitation materials matters because voters respond to costly signals from candidates as a shortcut for determining both a candidate’s investment in their own campaign and the degree of commitment from other voters to that candidate’s cause. Thus, it is not face-to-face interaction per se that drives existing the findings demonstrating thatprovescanvassing is the superior method a of marshalling voters to the polls on election day, but rather the relatively higher cost of this action. We test these hypotheses in a campaign where the candidate allowed us to treat his campaign materials, enabling us to control for the difference between face-to-face and mail interaction while varying the costliness of the signal in ways that would have not otherwise been possible. Our findings indicate that costly signals do have an independent effect on both turnout and voter choice. This conclusion holds lessons for campaigns and the study of voter behavior, but it also has implications for the microfoundational validity of signaling arguments more generally.