We develop and test a model of joint determination of economic growth and national election results in the United States. The formal model, which combines developments in the rational choice analysis of the behavior of economic agents and voters, leads to a system of equations in which the dependent variables are the growth rate and the vote shares in presidential and congressional elections. Our estimates support the theoretical claims that growth responds to unanticipated policy shifts and that voters use both on-year and midterm elections to balance the two parties. On the other hand, we find no support for “rational” retrospective voting. We do reconfirm, in a fully simultaneous framework, the “naive” retrospective voting literature's finding that the economy has a strong effect on presidential voting. We find congressional elections unaffected by the economy, except as transmitted by presidential coattails.
All Science Journal Classification (ASJC) codes
- Sociology and Political Science
- Political Science and International Relations