The Ties that Bind? Intra-party Information Exchanges of German MPs in EU Multi-level Politics

Abstract

As political authority is successively transferred from the national to the EU level, national parliaments are often considered to lose control over the domestic political agenda. Yet recent studies suggest that national parliaments cannot simply be labelled ‘losers’ of European integration. National parliaments have institutionally adapted to the EU in order to better scrutinise and control their governments in EU affairs. While existing research shows how parliaments employ their institutional opportunities to exercise scrutiny in the national arena, this paper suggests that MPs also employ informal strategies to obtain information on EU affairs to control and influence their governments. It argues that MPs primarily act through political parties, which are viewed here as multi-level organisations, and make use of their partisan ties to regional, transnational and supranational party actors to obtain information on EU issues. The article probes this argument by drawing on original data obtained through a survey of German MPs in 2009.

Acknowledgements

We would like to thank Christian Melbeck, Hans-Jörg Schmedes and Udo Zolleis for their support in the development and implementation of our online survey as well as Jessica Fortin-Rittberger, Alexander Gattig and the two reviewers for their very helpful comments on earlier versions of the manuscript. We are also grateful for the comments obtained from seminar and workshop participants at the Bremen International Graduate School of Social Sciences (BIGSSS) and Nuffield College. We also gratefully acknowledge financial support from the German Science Foundation [DFG, RI 1972/1-1].

Notes

1. Our conceptual distinction is empirically confirmed by MPs actual political activities: While MPs’ information and influence seeking exchanges with MEPs correlate quite strongly (rho: 0.62, p: 0.000, N = 88), the relationship is less strong for influence and information exchanges with MPs in other countries (rho: 0.40, p: 0.0002, N = 85) and regional ministers (rho: 0.53, p: 0.000, N = 81).

2. Moreover, one might argue that the large size of the German delegation in the European Parliament provides members of the Bundestag with a strong incentive to contact MEPs from their party to obtain information on EU policies and politics. In addition, given that party systems in the new member states are still more fluid, our findings might not hold for Central and Eastern European countries.

3. See http://true-european-voter.eu/sites/default/files/Sudulich_Nov26.pdf (accessed 21 March 2013).

4. Non-parametric Kruskall–Wallis rank test.

5. Originally, we measured our dependent variable on a five-point scale (never, seldom, occasionally, often, always). The empirical results presented here are the same for both the three- and five-point scale. We do present the trichotomous variable here, since it seems analytically and heuristically more useful to us.

6. The differences of the cut-off points of the dependent variables in our multivariate analyses are statistically significant, corroborating the appropriateness of the ordered logistic regression model.

7. Variance inflation factors indicate that the correlation does not cause problems of multicollinearity in the multivariate analyses.

8. Given the ordinal-level measurement of our dependent variable (frequency of contacts), we employ a (non-parametric) Wilcoxon–Mann–Whitney test to compare the differences in the average interaction frequencies between parliamentarians who are members of the political parties forming the governing coalitions (CDU and SPD) and those who are members of opposition parties (Grüne, FDP, Die Linke).

9. Some of our respondents did not provide information on their EU attitudes and the salience they attach to EU policies and politics. These cases had to be dropped in our multivariate analysis. If, however, we estimate without our control variables but all 94/97 cases, the estimates for our explanatory variables do not change significantly.

10. The change of probability from government to opposition status is estimated while holding all other variables constant at their mean.