Political Representation in Different Electoral Settings: Measuring Issue Congruence with VAA-Generated Data

Over the last decade, the long standing findings of political representation studies have been challenged, with disagreement found over the effects of institutional design settings. The non-findings at country level suggest that more effort should be put into investigating the effect of political parties’ characteristics in relationship with congruence (Belchior, 2013). Additionally, congruence studies have systematically overlooked the effect of individual characteristics due to a lack of analytical framework (Walgrave and Lefevere, 2013). Part of the non-findings and disagreement over the effects of country and party level predictors can be attributed to the panoply of measurements of congruence. Firstly, the ambiguity of the meanings of Left and Right across different electoral contexts poses problems for comparative research. Data on party positioning are now plentiful. However, there are not many sources of data where the voter and party positioning are interconnected. VAA data offer the perfect opportunity to study congruence in a comparative perspective, as it captures the political preferences of parties and citizens on the same set of issues. This paper employs data from two pan-European voting advice tools, the 2009 EU Profiler and 2014 EUvox. Congruence is assessed as the matching between prospective voters and their most preferred party, expressed as vote intention, on the Economy, Social and EU issue areas. Secondly, recent research brought into attention the caveat of focusing on the median citizen when assessing congruence (Golder and Stramski, 2010). This paper employs two different approaches of assessing congruence, an absolute citizen and a relative measure. Using hierarchical modelling, the results show that party and individual level characteristics play an important role in explaining variation in congruence across countries, but also across issue dimensions. Furthermore, the results suggest that we need to look further than the agreement between the average voter and her party.