Incumbency and The Presidential Vote in Senate Elections: Defining Parameters of Subpresidential Voting

Compared with the recent substantial strides in presidential election analysis, research on congressional elections has only begun. The majority of studies have been cast within the presidential-election context, with the relationship between the presidential and congressional vote the phenomenon to be explained. The present attention to presidential contests is understandable because of the inherent interest in such races in a presidential-centered political system and because reliable survey data have been limited to nationwide samples, severely restricting analysis on a state or district basis. And yet without some comparable advances at the congressional level, we cannot assess bases of electoral support nor the numerous assumptions of behavior in Congress as linked to this support, nor attempt a theory of voting behavior that does not consider voting at the subpresidential level. Put simply, the state of research is such that we have only begun to identify and measure the key variables affecting congressional voting outcomes. It is this basic task to which recent research in the field has been directed. Consider as the core phenomenon for explanation the sharp fluctuations over time in the partisan division of the vote for Senator and Representative. Since studies of voting behavior indicate the stability of party loyalties over time, evidence of sharp shifts in voting outcomes suggest factors other than party cues influencing the vote decision. Among the lines of inquiry opened, studies by Cummings, Press, and Hinckley, following the work of Key, have utilized aggregate election statistics to measure the substantial impact of the presidential vote on House election outcomes.