A spatial election with common values

Abstract

Does electoral competition make candidates reveal information that voters value? I study this question in a Downsian model of a repeated election consistent with six stylized facts of US Presidential Elections: (i) there are two candidates/parties, (ii) they are longlived, (iii) there is majority rule, competition is over many issues at a time (iv) some on which voters disagree, (v) others on which they do not, and (vi) prior to the election, not all information that voters value is available to them. In this election, even if candidates compete in multidimensional space and appear ex-ante identical, Nash equilibria exist.