Electoral Strategy and Economic Policy

This paper develops an approach to political equilibrium in a twoparty setting. The approach characterizes political resource allocation as trading o¤ the utility of core party supporters and swing voters, i.e. those who are not attached to either party. The model has three stages. First parties determine their electoral strategies. Second, an election is held. Third, policy is chosen. To illustrate the model at work, I apply it to two policy settings –the standard spatial model and the distributive politics problem of targeting transfers to groups of voters. The latter application is particularly interesting since the game of electoral strategy that emerges is supermodular. The paper then brie‡y discusses strategies for making political equilibria credible, how some of the key magnitudes suggested by the theory can be measured and the additional considerations that arise in dynamic models of political competition. This paper is based on the Walras-Bowley lecture presented at the World Congress of the Econometric Society in August 2005. Thanks to Paul Milgrom for some useful discussion following the lecture and to Jim Snyder and Xavier Vives for useful advice along the way. The paper draws on joint work with Anne Case, Steve Coate, Torsten Persson, Ian Preston and Daniel Sturm to whom I am grateful for their collaboration. I also thank Konrad Burchardi, Masa Kudamatsu and Juan-Pablo Rud for able research assistance.