Disentabgling components of PTV measures, with an application to the EES 2009

Conventional analyses of turnout and party support treat the question of why people vote separately from the question why they support a certain political party. Yet it is most unlikely that voters themselves distinguish the two processes. People vote for a party, they do not vote at random. In this paper we suggest a framework for addressing the relationship between the two questions. To what extent do features of individual-level party support contribute to our understanding of turnout? Is it meaningful to attempt the development of a unified model that accounts for both turnout and party choice simultaneously? Our general intuition is that having a distinct preference for one party over all others, and pronouncing oneself likely to vote for that party, will not only largely determine which party is voted for but also help to determine whether someone votes at all. Our approach seeks to analyze propensity to vote questions in the context of a multi-level model that permits us to isolate factors that influence party support at one level of analysis and factors that influence turnout at a higher level of analysis.