Primaries, General Elections, and Voter Turnout

Abstract

This article uses a multinomial logit model in testing whether or not the same process is at work in determining both primary and general election turnout (the study also tests a number of hypotheses concerning why people vote). The advantage of this methodology is that it allows the incorporation of primary and general election voting into the same model. It also allows direct comparisons between groups of voters. The results of the individual-level analysis indicate that a variety of short-term (campaign interest and concern about the election outcome) and long-term factors (efficacy, civic duty, and perceived level of information) influence the decision to vote. It appears that the same process is at work in determining both primary and general election turnout. Although a simple vote/no vote dichotomy may seem reasonable for most turnout studies, the current model demonstrates that there are subtle differences between voters who participate at different levels.