Surge and Decline, Negative Voting, and the Midterm Loss Phenomenon: A Simultaneous Choice Analysis

Abstract

Campbell's surge and decline and Kernell's negative voting theories have been employed most frequently by researchers who seek to explain the invariable pattern of administration party vote loss in midterm House elections. Despite the original emphasis placed on voting as the act of selecting jointly from among the three options of abstention, Democratic choice, or Republican choice, all efforts to corroborate the theories have inappropriately treated the decisions of whether and how to vote as independent questions for investigation. Applying the more suitable technique of nested multinomial logit analysis, I find little support for negative voting's two key contentions that negative presidential evaluations hurt White House party candidates more than positive appraisals help, or that the president's backers have disproportionately low turnout; however, a rational basis for abstention emerges that differs from the one postulated by negative voting. In contrast, on-year defections to presidential party candidates by regular voters who normally favor the opposition do seem a cause of the biennial electoral cycle, as surge and decline indeed would have us to believe.