The Impact of Candidate Selection Processes on the Representation of the Pool of Candidates: The Case of the Elections to the European Parliament

One of the main motivation to study candidate selection processes is their alleged impact on the type of candidates they produce. After all, understanding the way the gatekeeping function is performed matters mostly for what it entails in term of (mis)representation in the pool of candidates and, as a consequence, of elected representatives. In this view, the absence of systematic investigation of the link between the intra-party candidate selection methods and the profile of those that are retained on the lists appears as a glaring omission, if not a mystery. This paper is a first attempt to remedy this lacuna by asking what type(s) of selection procedure lead(s) to a better representation of women and ethnic minorities on candidates’ lists. Although there exists contradictory evidences, recent literature has most often come to suggest that more closed, centralised and complex selection procedures leads to a better representation of ‘minorities’. This seems to stem from the more complex coordination needed to attain certain equilibriums when more people are involved in selecting party candidates, as well as from the fact that party leaderships generally include higher-educated individuals with more liberal attitudes towards gender equality and minority rights. In contrast, having members or even party sympathisers choose the candidates would inhibit competition since they tend to reselect incumbents (who are also disproportionately men). In this research, the EP electoral setting is used given the ideal comparative arrangement it provides: a quasi-simultaneous electoral contest in all the member states (allowing for a genuine cross-country comparison), a very large pool of candidates, as well as electoral rules of a similar nature (more notably the use of PR systems). The study draws on the one hand on a new and original dataset, collected at the party-level, which records the formal selection processes used in the 198 national parties having gained representation in the 8th EP legislature (2014-2019). On the other hand, data on the number and ranking of female candidates and of candidates of ethnic origin is collected based on the official lists of candidates, through name and facial recognition. Thanks to the perspective employed, an unprecedented comparative perspective (given the number of parties and countries concerned) is offered in the paper.