Optional Anonymity and Political Balloting

Abstract

Data from a student political poll were analyzed to test the hypothesis that those voters casting a ballot for the unpopular candidate would be more likely to do so anonymously than would those voting for the popular candidate. The data tended to support the hypothesis, but were rendered less definitive by the fact that affiliates of the political party allegedly represented by the unpopular candidate tended to cast anonymous ballots regardless of their candidate preference. Implications of this and previous findings were discussed with respect to both the experimental vs survey approach to the anonymity problem and the relative wisdom of studies focusing on the anonymous response as opposed to the public response in voting contexts.