We explore the 2nd-order elections framework in the context of the 2014 EP elections through a twofold strategy. We first attempt to falsify the core predictions of the second-order model through tests on aggregate electoral results, in line with extant applications on previous EP elections. We then investigate specific explanatory mechanisms for the 2014 EP elections, by relating party performance with party stances on major issues, as well as with country-level electorate configurations on the same issues. Both party-level and voter-level data are drawn from the euandi 2014 Voting Advice Application (VAA) project (the successor of the EU Profiler 2009), which includes expert survey data (for parties) and VAA users’ responses (for voters) to a battery of policy issue statements.
