Analysing roll calls of the European Parliament - A Bayesian application

This article explores the performance of a Bayesian application of spatial voting models to the roll calls of the Fifth European Parliament (EP). Focusing on two distinct voting behaviours of members of the EP (MEPs) - high absenteeism and the defection from majority alternatives caused by the influence of national parties - it shows that the Bayesian method is complementary to the standard NOMINATE method. In general, the two methods produce very similar estimates and work as robustness checks for the results from each other. However, the Bayesian method enables us to measure the uncertainty of estimates resulting from the estimation with a large number of missing data and some random-appearing roll calls. In this way, it helps us draw more confident inferences about MEPs' voting behaviour.