The President's Party in Midterm Elections: Going from Bad to Worse

Abstract

Models of aggregate vote change in midterm congressional elections have largely ignored the role of short-term party evaluations. This paper argues in favor of including such evaluations in the analysis of midterm voting change. We find that short-term party evaluations link both presidential popularity and economic conditions with election results and help to account for the gap between the outcomes of first and subsequent midterm elections: since 1945 the president's party has done considerably worse in subsequent midterm elections than in first midterm elections after taking control of the White House. Analysis of survey data from the 1974, 1978, and 1982 midterm elections corroborates the influential role of short-term party evaluations.