CONCLUSION

MEP-group alignments constitute an important contextual aspect of EU legislative lobbying and party competition. In particular, we demonstrated that the alignment of parties and organized interests reflects a left-right cleavage. Organization type, namely the distinction between business and public interests, coincides to some extent with party cleavages. Overall, business is more closely aligned with EPP, ECR and ALDE. In contrast, NGOs are strongly aligned with the S&D, GEU/NGL and the Greens. Finally, a considerable number of MEPs take a more moderate or centrist position, which is particularly the case for the EPP. This structure of alignments is reflected in the institutional context surrounding legislative proposals. For instance, the type of DG – being business-or NGO-friendly – is a significant predictor of MEP-group alignments. Proposals stemming from business-friendly DGs arouse closer alignments with interest groups for ALDE, ECR, but also the Greens, while they result in lower alignments for the EPP. The evidence also shows that, compared to the DG, the political background of the Commissioners is much less crucial, which confirms the key importance of bureaucratic politics in the EC. It is obvious that EU interest group mobilization corresponds to a consider-

able extent with party competition, whereby on average NGOs (business) are closer to left (right) parties and distanced from right (left) parties. Nonetheless, the absence of a perfect match NGO-left parties (or business-right parties) also urges us to remain careful and refrain from making highly deterministic claims. For some dyads, we observed more articulated positions and larger political distances, while in other instances parties and groups adopt more moderate positions. It is important to emphasize that we expected that some actors are on average more aligned. This qualification ‘on average’ is rather relevant, given the absence of a perfect match for all observations. How can we explain this? Although the EC (in 2008-10) was more to the right than the left, for many

of its policies support from centre-left parties, especially the S&D, was needed in order to get a majority endorsement from the EP. Generally, European legislative politics is characterized by a consensual political culture and a strong propensity to establish large majorities. This is one of the reasons why MEPs and, possibly, some interest groups refrain from adopting radical positions, but rather present moderate views, as this puts them in a position to become

only one ordinal category; and 3 ¼ group and party take highly divergent positions in favour or against the legislation. The models test whether group type, the Commissioner and the DG predict

the ordinal distance. The DG variable, which reflects the institutional venue where the proposal was drafted, distinguishes between DGs that are, because of their task orientation, (1) more business-minded, (2) more NGO-minded or (3) others (see Bernhagen et al. 2015). One of our questions is the extent to which similar organization types flock together around specific DGs. Interestingly, compared with business-MEP dyads there is a significantly higher chance that NGO-MEP dyads are tied to the NGO-minded DGs (Chi2 ¼ 84.95; df ¼ 4; p , 0.0001). Of the NGO dyads, 38 per cent are connected to these DGs, while 27 per cent of the business-MEP dyads concern proposals that originate from these DGs. In contrast, there are relatively more business-MEP dyads showing an interest for business-minded DGs; 60 per cent of the business dyads are connected to these DGs while only 34 per cent of the NGO-MEP dyads focus on business-minded DGs. Generally, we expect that if the proposal originates from a business-minded DG, the distance betweenMEPs and groups will be larger for leftist parties, while the distance will be larger for rightist parties in case of a NGO-minded DG. Another key explanatory variable is the party group of the responsible Commissioner: ALDE; EPP; S&D; or a combination of ALDE and EPP.5 The expectation is that MEPs stemming from the same party group as the Commissioner will align less well with the mobilized interest group community. Finally, we control for the type of legislative procedure, which affects the EP’s legislative power; a dummy variable distinguishes the codecision procedure from other procedures.6