7 Conclusion

Section 5 demonstrated that the party politics of the EU are better at aggregating ‘choices of policies’ than ‘choices of leaders’. The former failure is only partial. The latter is more deep-seated. The use of national parties to structure choice on Union policies certainly encounters difficulties. Not only are the party systems somewhat less than isomorphic at the two levels but, even where there is a reasonable fit between them, choices between national parties cannot always serve as ‘good proxies’ for choice on Union matters.

Yet for all these difficulties, there would appear to be an underlying dimensionality to choices of ‘left-right’ and ‘pro-anti integration’ policies at Union level. Although it is possible to disagree on whether those dimensions are combinable, separable or fated to clash (Bartolini 2005 ), it is probably common ground that parties allow choices along one of those dimensions – left-right – to be aggregated in the EU arena.

In contrast, the EU’s political system hardly even attempts to link choices of leaders to electoral choice between parties. Even in those instances – limited to the choice of the EP itself – where office is filled by competitive election, party politics are not structured for the public control of those who are competing for office. It would be hard to argue that the ‘second-order’ characteristics of European elections have been softened to the point at which they aggregate votes around competing assessments of the performance of parties in an out-going European Parliament or around competing assessments of what parties have to offer for a coming Parliament.

Could things be better? It is not easy to answer this question when – as Section 6 demonstrated – so much remains mysterious about the party politics of the EU. More research is needed before we can tell whether incumbent parties block greater competition on EU issues or whether there is simply little voter demand for different ways of representing publics and aggregating choice at Union level.

Nor, indeed, do we even know for certain what kind of party representation has already been created in the European arena. For Simon Hix the groups in the EP are already distinct, cohesive and competitive enough to provide a basis for party-responsible government at the Union level. For Stefano Bartolini, the whole structure remains a delicate set of elite compromises between national party delegations that would not survive electoral competition linked to the performance and behaviour of the groups themselves. In between these contrasting assessments is another possibility, little noticed, hardly researched and omitted from present accounts of aggregation and representation in the Union arena. It is that EU party actors are constrained by the collective of national parties that comprise them, yet national parties are individually constrained by a wider EU party system that none can easily change on their own.