Local equilibrium equivalence in probabilistic voting models

Electoral equilibria depend upon candidates’ motivations. Maximization of expected vote share may not lead to the same behavior as maximization of the probability of winning the election. Accordingly, it is desirable to understand when electoral equilibria are insensitive to the choice of candidate motivations. This paper examines sufficient conditions for local equilibrium equivalence between expected vote share maximization and maximization of probability of victory in the spatial model of elections with probabilistic voters. Journal of Economic Literature Classification Numbers: C72, D72. ∗This paper has benefited enormously from discussions with Jeff Banks, Kim Border, Mark Fey, Richard McKelvey, Tom Palfrey, and especially John Duggan. In addition, the helpful comments of an anonymous referee and financial support of the Alfred P. Sloan Foundation are both gratefully acknowledged.