The logic of voting behaviour in the European Parliament : new insights on party group membership and national affiliation as determinants of vote

The greater centrality of the European Parliament (EP) within the EU’s institutional structure has increased the importance of voting behaviour of its members (MEPs), thus making it more important to fully understand the logic they follow. A consolidated stream of literature has treated the political groups of the European Parliament (EPGs) as cohesive actors, influenced mainly by the classical dimension of left-right politics, with nationality playing a marginal role; a second, anti-/pro-EU dimension exists, but it is orthogonal to the main one and explains a fewer share of votes. In this perspective, MEPs from the same political group are likely to vote together (just like any democratic parliament), no matter which country they come from. However, other scholars suggest that the methods used to reach these findings are biased. I add new elements to this discussion, analysing: (I) the cases when EPGs are exceptionally low cohesive; (II) the different dynamics of various EP’s legislative committees; (III) the possible contrast between EPGs and national delegations; (IV) the relationship between MEPs’ exogenous policy preferences and their revealed voting behaviour. Theoretically, I originally define MEPs’ voting behaviour as a refinement of existing theories both at a macro-level (EU party system, party group cohesion) and a micro-level (MEPs’ representativeness, principal-agent theory). I firstly affirm that the current knowledge about the high cohesiveness of EPGs is a consequence, although correct, of a quantitative approach that “dilutes” the variance by looking at the aggregate scores of indexes of cohesion (such as the Agreement Index, AI); secondly, that voting behaviour of MEPs in the various committees of the European Parliament has different dynamics, following the indication of many scholars that knowledge of the internal dynamics of committees in the EP is still blurry; thirdly, that national affiliation, yet emerging only in a small minority of votes, is crucial and has some specific characteristics that can be isolated; fourthly, that MEPs’ preferences and behaviour differ more than expected along partisan (EPG) lines than national lines. Empirically I then test these hypotheses on the 6th EP legislature (2004-2009) and partially on the 7th (2009-2013) with a combination of quantitative and qualitative methodological tools: a critical use of the AI; the results of an original survey submitted to MEPs’ assistants; descriptive and inferential statistical techniques; a combination of survey preferences and behavioural data (roll-call votes). Finally, I add some qualitative insights in order to reinforce my findings. My results show that the oft-repeated claims about the resemblance of the European Parliament to a classical democratic parliament should be taken more carefully: usually political groups are highly cohesive, but when controversies arise (because the outcome of the vote may impact member states’ interests) national affiliation plays a greater role than the one usually envisaged by the mainstream literature