A Multi-issue and Stochastic Model of Candidate Competition

Stefan Krasa, University of Illinois
Mattias K. Polborn, University of Illinois, Urbana-Champaign

Abstract

We introduce a general framework in which there are several policy dimensions (``issues'') and finitely many possible positions on each issue. Two competing candidates are exogenously committed to particular positions on a subset of these issues (e.g. candidate gender, race, previously committed positions), while they can choose any policy for the remaining issues to maximize their winning probability. Voters have general preferences over policy vectors and vote sincerely. Candidates are uncertain about the distribution of voter preferences.

We characterize a condition on voter preferences (satisfied in most existing models) under which candidates' equilibrium policies generically converge in equilibrium. In contrast, if there is complementarity between a candidate's fixed and flexible position, strong policy divergence arises in equilibrium. As a normative criterion, we define competition-efficiency and provide conditions under which the equilibrium is or is not competition-efficient.