Procedural Effects and Party Pressure in European Parliament Roll Call Votes

The literature on voting in the European Parliament (EP) highlights the role of legislative parties. This is often done by showing that the uncovered ideal point estimates cluster more heavily around party groups than extant measures of ideol- ogy. The standard approach is to study all roll call votes within a legislative term, effectively treating each vote as equally important. Yet, we know that most roll call votes are non-consequential in terms of their effect on EU policies, as they are taken on non-legislative resolutions. This raises two concerns. First, failure to account for vote specific party inducements may make the estimated ideal points appear more clustered around parties than what would be the case if such inducements were ac- counted for. Second, most of the roll call votes in the EP are on resolutions, to the extent that voting behaviour in non-legislative resolutions is unrepresentative for voting behaviour on codecision legislation, the estimated ideal points will generate a distorted picture of voting behaviour on the votes most researchers care about.
Building on the framework of Clinton, Jackman and Rivers (2004), this paper (1) estimates to what extent MEPs change their position when voting on codecision legislation compared to resolutions, and (2) evaluates to what extent party groups are able to influence the voting behaviour of the MEPs on specific votes. The re- sults show that some MEPs behave differently on codecision and resolution votes. Furthermore, there occur substantive vote-specific party inducements on codeci- sion legislation. As the standard approach ignores such effects, it risks painting a distorted picture of politics in the European Parliament.