Political Equilibrium in Representative Democracy

Abstract

It is shown in this paper that the Median Voter Theorem lacks robustness in the sense that if voters have (even the slightest) preferences for the competing candidates, beside preferences for their current policy proposals, then no policy in the neighbourhood of the median voter's preferred policy constitutes an equilibrium (in pure strategies). This suggests that this classical theorem does not apply to representative democracy. Indeed, if voters do have candidate preferences, and these are strong enough, the policy-motivated candidates will, in general, adopt differing policy positions in equilibrium, and, under certain qualifications, the equilibrium outcome will be (close ti) a particular utilitarian optimum. More specifically, in a discretized model the policy outcome will lie between the preferred policy of the most popular candidate and this utilitarian optimum.