Strategic party placement with a dynamic electorate

Abstract

This article expands upon formal research on elections by considering competition in a dynamic environment of multiple elections. The key assumptions are that the ideology of the electorate is changing in a known way, parties cannot change their position from one election to the next, and one party has a non-ideological advantage in the first election. A deterministic version of this game shows that with a large valence advantage for one party, both parties will converge to the median for a future election. With a small valence advantage, there is no pure-strategy equilibrium, so a stochastic version of the game is considered. With probabilistic voting, parties place themselves closer to the present median, but move towards the future median the more highly they value winning in the future and the less uncertain they are about election outcomes.