Exploration Of Ticket Splitting

"Why would a single voter cast a ballot for two public officials from different parties who represent different political philosophies? I explore the aggregate level variables that may influence net split-ticket voting and whether or not a state observes a presidential election years from 1988-2008. This paper examines the expected effects of socioeconomic and candidate characteristic variables on ticket splitting. In particular, this study focuses on whether ticket splitting dynamics in Electoral College battleground states are distinct from those in noncompetitive states. I posit that the substantial amount of attention from presidential candidates in battleground states polarizes voters and contributes to straight ticket voting. The analysis conststs of two sections using linear and logistic regression analysis, respectively, of the following dependent variables: (1) the difference in vote share between the Democratic presidential candidate and the Democrat Senate candidate as a measure of net ticket splitting; (2) the split/not split election outcome observed by a state. Each section consists of two parts: (1) independent variables excluding candidate characteristics; (2) independent variables including Senate candidate characteristics of incumbents. Introducing the Setnate candidate characteristic variables does not have a sizeable expected effect on the vote share difference or ticket outcome. In my multivariate analysis, I find that socioeconomic variables of a state best explain the degree of voter partisanship and, ultimately, the results of the election at the aggregate level."--Abstract.