Competitive and Coherent? Profiling the Europarties in the 2009 European Parliament Elections

Abstract

On the bases of the new EU Profiler data for the 2009 European Parliament elections, this work looks at two basic criteria to assess the representative potential of the EU party system: its competitiveness and the policy coherence of its parties. It is here argued that, if the national parties are successfully able to aggregate their programmes and agendas at the EU level, proposing different options to the European voters, the EU ‘democratic deficit' might not be as severe as it is often lamented. It is found that the Europarties, despite the enlargements towards Central and Eastern Europe, are sufficiently coherent and different to seek to fulfil an expressive, or representative, function. By selectively placing its focus on the ‘supply-side' of politics, this work shows that European voters could indeed make meaningful choices, which the Europarties might turn into concrete policies through their parliamentary activity.

Key words:

• EU Profiler
• Europarties
• representation
• EP elections

Acknowledgements

A previous version of this article was presented at the Workshop “Where now for Europarties? Reflections post Lisbon”, University of Maastricht, 20-21 June 2011. A sincere thanks to the organizers: Simon Lightfoot and Natalia Timus. I am very grateful to Alexander Trechsel for introducing me to the EU Profiler data and for his generous advice. Thanks also to Niccolò Conti, Eugenio Pizzimenti and the two anonymous reviewers. My deepest gratitude goes to the late Peter Mair for his exceptional patience and wit.

Notes

1. Here, the label 'Europarties' refers to both the extra-parliamentary and parliamentary parties of the EU, when a distinction between the two does not appear necessary.

2. Non-legislative report on “Application of Regulation 2004/2003 on the regulations governing political parties at European level and the rules regarding their funding”, 2010/2201(INI).

3. The concepts of 'cohesion' and 'coherence' are often used interchangeably (Janda, 1980).

4. The policy statements are based on dimensions employed successfully in both manifesto research and expert studies. See also Trechsel and Mair 2011, p. 5-6.

5. See the full hierarchy of the sources in Trechsel and Mair 2011, p. 12.

6. Scales ranges from 0 to 100 – the maximum value indicates full agreeement by the party.

7. Note that the EU Profiler dimensions are explicitly positional.

8. In our measure, party positions are weighted by the share of seats in the EP, rather than by vote share.

9. The Profiler spiders have been developed on the bases of similar voting-advice applications, such as the Dutch 'Kieskompass' and the Swiss 'Smartvote'.

10. The spider display has an element of arbitrariness in the ordering of the policy dimensions and the assumption of equi-angular distances among them. However, as authoritatively argued: “it is, perhaps, tempting to dismiss these displays as simplistic novelties; however, doing so would ignore their strengths and their potential power in an exploratory data analysis strategy” (Jacoby 1998, p. 26).

11. The principal factor analysis retains two factos with eigenvalue bigger than 1. A first factor, with eigenvalue 3.6, accounts for the 68.7 percent of the variance among the dimensions. A second factor, with eigenvalue 1.3, acccounts for the 24 percent of the shared variance.

12. The correlation between the first extracted factor and the Profiler left-right dimension is .96, significant at .000

13. Of the original 28 policy questions in the EU Profiler survey, 8 have been a priori regarded as left and 10 as right (the others were disregarded for the left-right scale). As for EU integration, 7 policy questions have been used. Arithmetic means have then been computed from the party placements on each policy category. For details, see EU Profiler 2010, p. 14-15.

14. The left-right and the integration dimension are independent (r=-0.02). Left-right squared accounts for the 31 percent of the variance in the integration dimension.

15. Omega squared, rather than other measures of effect size (i.e. eta squared) is used as it provides an unbiased measure in small samples.

16. The correlation between the EU Profiler and the Euromanifesto left-right scale is 0.86 (p<.05), between the two EU integration dimensions 0.89 (p<.01)

17. Regulation no. 1524/2007, art. 8, states: “The expenditure of political parties at European level may also include financing campaigns conducted by the political parties at European level in the context of the elections to the European Parliament”.

18. Art. 17.7 of the Lisbon Treaty reads: “Taking into account the elections to the European Parliament and after having held the appropriate consultations, the European Council, acting by a qualified majority, shall propose to the European Parliament a candidate for President of the Commission. This candidate shall be elected by the European Parliament by a majority of its component members” (emphasis added).

19. Importantly for the external validity of the results here presented, several recent research papers reach very similar conclusions on the feasibility of a system of political representation at the EU level, on the bases of survey (Schmitt and Pütz 2009), expert (McElroy and Benoit 2010) or Euromanifesto data (Bressanelli 2012).