Split-Ticket Voting: An Implicit Incentive Approach

This paper applies an implicit incentive approach in the principal-agent framework to explain split-ticket voting, when citizens vote for candidates of di¤erent parties in the local municipal simultaneous elections (i.e. the elections of a mayor for the city hall and of a governor for the province). The principals (voters), in each period of an in nite horizon, reward the agents (mayor and governor) with reelection based on their observed performance but through an implicit reward rule. So the voters can inuence the politiciansperformance only through the choice of an evaluation rule. We \u85rst show that, if the voters do vote split-tickets and if the politicians do care about overall representation of their party in governing bodies rather than their own reelection prospects, then the voters prefer a relative performance evaluation rule in the next period. Otherwise, they use an absolute performance evaluation rule to reappoint the politicians in the next period. Second, we show that the stationary probability that in the long-run the principals vote split-tickets is independent on the initial state, and is lower than the stationary probability that they do not vote split-tickets. JEL classi\u85cation: D72, D82