Political Competition in Two-Stage Elections

We study a two-stage election with two parties, each nominating a candidate in the first stage primary election to compete with the other party’s nominee in the second stage general election. This resembles the U.S. presidential election and is used in many countries in the world, but has largely escaped attention in the literature. We extend Coleman (1971) and Roemer (1997) to model such elections in the presence of uncertainty about the median voter’s position in the general election. For a given nominee by the other party, each voter in the primary chooses an optimal policy which takes into account both her ideological preference and each candidate’s winning probability in the general election. We show that under certain conditions, the two parties’ equilibrium policies are their respective median voter’s optimal policy against each other. Such a stable outcome is possible only if we allow candidates to freely enter or withdraw from the primary elections. The degree of policy divergence is positively correlated with the uncertainty in the general election, but is bounded by the distance between the two party median voters’ ideal points. JEL Classification: D72