Published in Electoral Studies 22(2003): 449-468.
This article analyses turnout in the European Parliament elections. Turnout is affected partly by the same factors that affect turnout in normal national elections and partly by factors related to the EU. The results show that most of the cross-country variation in turnout can be explained by the same variables as in national elections. Compulsory voting, weekend voting and having other elections simultaneously with EP elections increase turnout. Also some EU specific factors affect turnout, but their effect is smaller. Voters in countries benefiting from the EU subsidies vote more actively than voters in the countries that pay these subsidies. Furthermore, turnout is higher in countries with strong support for the EU membership. Holding elections during weekends and having multiple constituencies in all countries could increase the turnout by approximately 10 percentage points.
The empirical data used in the analysis is available here in Stata format. And here are explanations for the variables in the data set. Finally, here is a Stata do-file, which will reproduce the results.