Candidate Faces and Election Outcomes∗

Recent research finds that naive survey participants’ rapid evaluations of the facial competence of United States Congressional candidates predict aggregate vote margins. The predictive power of facial competence has generated considerable interest because it seems to indicate a causal relationship between face and vote choice. Because there is no a priori reason to expect that candidate facial qualities are randomly distributed across electoral contests, we estimate the effect of facial competence using multiple regression analysis with controls. Using data from a survey we implemented with evaluations of more than 167,000 pairs of candidate faces, we find evidence that facial competence has a small but significant causal effect on vote choice; we also find evidence that candidates with high facial competence participate in the most competitive electoral contests. We argue that faces are an interactive part of a bigger political story that structures election results. ∗The authors thank Mike Franks, Gary Jacobson, R. Brian Law, Jeff Lewis, Elisabeth Michaels, David Sears, Alexander Todorov, Lynn Vavreck, John Zaller, and participants at the UCLA Workshop on Political Methodology. †All authors: Department of Political Science, University of California, Los Angeles; [email protected], [email protected], [email protected].