Spatial voting with endogenous timing

We consider a model of (spatial) voting with endogenous timing. In line with actual political campaigns, candidates can decide endogenously when and where to locate. More specifically, we analyze endogenous timing in a two-period n-candidate spatial-voting game. We show that this game possesses a pure-strategy equilibrium* (OSBORNE [1993]) but no-or only very complex-subgame-perfect equilibria. We demonstrate the latter point by analyzing the subgame-perfect equilibria in a three-candidate game. Our results show that allowing for endogenous timing can eliminate some of the more unappealing equilibrium characteristics of the standard model.