Second-Order Election Effects in the European Multilevel Party System

A significant growth in number and importance of regional and European elections has induced a rich literature on non-national elections. The dominant perspective to analyze non-national elections is the second-order election model which assumes that regional and European electoral outcomes are driven by electoral dynamics in the first-order, national electoral arena. The literatures studying regional and European elections hardly speak to each other which leaves important questions unanswered. Are European and regional elections to a similar degree second-order? Do similar factors impact second-order election effects in both regional and European elections? And in how far are there spill-over effects between second-order electoral arenas? In this paper I set out to explore when and how national politics conditions election outcomes in subnational and supranational electoral arenas. A unique research design compares European and regional elections to the same and previously held national election for elections held in 180 regions in eight EU member states for 1979 until 2014. This set-up allows for a direct comparison between second-order election effects in European and regional elections and one may observe in how far the same second-order election (SOE) model can account for outcomes in both European and regional elections. Furthermore, the unique research design allows me to develop hypotheses on spill-over effects between second-order electoral arenas. I will test a regional identity hypothesis which expects SOE effects to decline when a SOE is held in a region with strong identities and a regional authority hypothesis which expects that SOE effects decline in regions which exercise more authority. The results indicate strongly that SOE effects are smaller for regional than for European elections and they provide strong evidence for both hypotheses.