The Nationalization of Subpresidential Elections: Retrospective Voting & the Salience of Presidential Heuristics

Abstract

American democracy is a complex system that operates across multiple federated levels. This structure presents a challenge to the voting public: how should they judge the performance of elected officials? Vote choice in presidential elections is better understood than in less visible down-ballot races, a discrepancy this paper attempts to rectify by ascertaining how voters retrospectively evaluate candidates in senatorial and gubernatorial elections.

Scholars have traditionally subscribed to a “federalist” theory of subpresidential vote choice in which voters consider the economic conditions of their state when choosing a governor but focus on presidential performance in senatorial races. In contrast, I expect to observe that in both cases, voters rely far more heavily on the heuristic of presidential approval due to the increasing nationalization of subpresidential elections.

I utilize linear probability regression to isolate the impact of retrospective economic evaluations and presidential approval on senatorial and gubernatorial elections in 2006. My results indicate that presidential approval heavily affected vote choice in both races while retrospective economic evaluations were far less salient. The findings present a substantial contribution to the literature on subpresidential vote choice, and the absence of federalized retrospective voting has implications concerning voters’ ability to hold elected officials accountable.