Seminario di ricerca

In common value elections voting signals a voter’s level of information, since, relatively uninformed voters find it optimal to abstain (Feddersen and Pesendorfer, 1996). If voters have reputational concerns, i.e., if they care to be perceived as “informed” by the rest of the voters, and they are selfish, then turnout should be higher, but the competence of the average voter should be lower compared to the case without reputational concerns, inducing efficiency losses. We test the empirical relevance of the above argument through a laboratory experiment and find strong evidence in its support. This suggests that reputational concerns might be a plausible explanation for the rather frequent victories of sub-optimal alternatives in elections