Party Identification in Germany: A Dynamic Analysis of Panel Data, 1984-2006.

The analysis of party identification outside the U.S. has featured disagreements about such basic questions as whether citizens of other democracies form meaningful and stable partisan identities. We use the German Socio-Economic Panel Study (GSOEP), a longitudinal study including twenty-three waves of annual data since 1984, to analyze the dynamics of partisanship in Germany. The panel spans the period of German unification, allowing one to examine the development of partisan identities in the newly democratized east --alongside the evolution of partisanship in the west, where unification was widely seen to have caused a swift breakdown of partisan ties. A handful of scholars have used this dataset to study partisanship (e.g. Zuckerman and Kroh 2006), but these studies have not addressed the problem of measurement error and have not fully exploited the dynamic nature of the data. In this paper, we present a series of analyses of partisan stability using dynamic discrete choice models which allow for state dependence and unobserved heterogeneity. The richness of the data allows us to examine which groups are most likely to change their partisanship over time and to explore the conditions fostering greater partisan stability. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]