Research on voting behaviour in elections to the European Parliament (EP) has demonstrated that voters tend to systematically favour parties which are "more extreme" than their own ideological or programmatic preferences. While the median voter theorem by Hotelling, Downs, and Black assumes posits that political competition converges towards the position of the median voter, more recent accounts have highlighted centrifugal incentives affecting both voters and parties: given the constraints elected officials face to deliver their advertised policy goals, voters may factor in checks, balances, political compromise and an abundance of veto players and ``overshoot the mark'' to get their desired policies enacted (i.e. ``discounting'' or ``compensational voting models'' by Grofman or Kedar). Ultimately, voters may cease to care about preferred positions on a policy
continuum, but instead aim to push for change into their desired direction (i.e. ``directional voting'' by Rabinowitz and Macdonald).
Clearly, party strategies need to react to these centrifugal patterns of political competition. In this paper, we apply the model by Adams, Grofman, and Merrill (2005) to the analysis of party competition in EP elections. Focussing on multiparty competition in the two-dimensional European political space, we demonstrate that the presence and magnitude of centrifugal competition depends on the (1) issue salience, (2) the impact of non-policy motivations, (3) the dispersion and shape of ideal points in the respective electoral constituencies, and (4) the extremity of the positions taken by the respective partisans. Our empirical findings illustrate that some member states (for instance Austria or the Netherlands) are more open and available for "issue entrepreneurs" than others (for instance Germany or the United Kingdom).
