The Paradox of Voting and Candidate Competition: A General Equilibrium Analysis

Abstract

Conventional analysis of the decision of expected utility maximizing agents to vote has concluded that it is irrational to vote unless voters have a distorted view of their individual impact or place a direct value on the act of voting. On the other hand, mathematical analyses of the electoral process (see, e.g., Davis, Hinich, Ordeshook (1970)), have usually assumed that all voters vote. Each theory is incorrect in the sense that in actual elections turnout is neither zero nor 100%. In this paper we will argue that previous analyses of expected utility maximizing voters s topped too soon because of the partial equilibrium approach and that if each voter considers the simultaneous reactions of all voters in a "rational" manner, then depending on the location of the candidates' platforms, turnout will usually be positive but less than 100%. In particular we will derive a (probabilistic) vote supply function, given a distribution of voters and the choice of platforms of candidates, which has the property that, even with costs of voting, unless the candidates have identical platforms, the expected turnout is positive. The model and these results are presented in sections 1a and 1b.