Divide and Diverge: Political Competition and the Organization of Conflict

Abstract

Why do politicians sharpen divisions when elections reward broad appeal? I model how two identical, office-motivated parties turn voter disagreement into partisan conflict. Elections allocate political power. Each party empowers its most valuable members first, so power has diminishing returns. Platforms never converge: differentiation builds constituencies that survive adverse popularity shocks. Voter polarization pushes platforms apart and benefits both parties--even when a voter group grows more hostile to one party. Across issues, parties divide along the direction that generates more voter conflict--for elliptical electorates, a leading principal component of voter positions, consistent with ANES survey evidence.