The results of the European Parliament elections in the Czech Republic were surprising for party participants, as well as for observers and students of political science, both from the viewpoint of election turnover, and success achieved by certain political parties. Even though the pre-election polls predicted election turnover of forty percent, no one expected participation at 28 %. High election gains of the opposition parties, ODS (Civic Democrats) and KSČM (Communist Party), were expected.
Yet, substantial failure of both government parties (ČSSD and US-DEU) came as a surprise. Whereas in the elections to the Chamber of Deputies in 2002 ČSSD gained more than 30 %, in the EP elections they gained less than 8 %, with a loss of 22 % in the course of two years. Freedom Union (US-DEU) didn’t get even a single mandate in the EP, because they gained less than 2 % and didn’t reach the required threshold. On the other hand, two parties that had never gained a mandate in the lower chamber of the Parliament – Association of independent candidates-European Democrats (Sdružení nezávislých kandidátů-Evropští demokraté) and Independent Movement (hnutí Nezávislí) – which are relatively successful in some regions and which designate themselves as a non-party entity- gained two chairs in the EP.
These facts suggest that the basic frame for discussing EP elections, which was presented almost a quarter of a century ago by Karlheinz Reif and Hermann Schmitt in the model of second-order national election [Reif, Schmitt 1980], could contribute to the understanding of election behaviour of Czech voters. In this text I will use the model to explain the different participation in the 2002 elections to the Chamber of Deputies and in the 2004 EP elections in the Czech Republic. In the first part of my paper I will present some theoretical reasons for different election participation in EP elections and in the elections to the Chamber of Deputies. With the help of the model I will propose a hypothesis about election behaviour on the individual level. In the second part I will put the election turnover in the 2004 EP elections in the context of other elections in the Czech Republic and I will explore basic differences in the election turnover between the 2002 elections to the Chamber of Deputies and the 2004 EP elections. I will also present the results of the empirical analysis of the proposed hypothesis and I will complete them by final notes and implications for further research of election behaviour.