Explaining Seat Changes in the U.S. House of Representatives, 1950–98

Abstract: Recent U.S. House elections have challenged existing models of congressional elections, raising the question of whether or not processes thought to govern previous elections are still at work. Taking Marra and Ostrom's (1989) model of congressional elections as representative of extant theoretical perspectives and testing it against recent elections, we find that the model fails. We augment Marra and Ostrom's model with new insights, constructing a model that explains elections from 1950 to 1998. We find that, although presidential approval ratings and major political events continue to drive congressional elections, the distribution of open seats must also be taken into account.