Do party primaries select leaders unwelcome by party elite? A comparative analysis of twelve leadership races in five European countries

Since the beginning of the 1990s many European parties have progressively adopted new forms of selection that allow their members or even voters to directly choose the candidates for the different public offices or the party leaders (Kittilson and Scarrow 2003; Massari 2004; Hazan 2006). But do party primaries really represent a new chance for an effective citizens’ participation to political life or they are just a weapon in the hands of the parties in order to re-attract disappointed voters or members? In fact someone could suggest the democratization simply provides the grassroots with the illusion to intervene in the choice, but actually they restrict themselves to support decisions already taken by party oligarchy. Anyway many scholars (Massari 2004; Valbruzzi 2005; Hazan 2006; Pasquino 2006) claim that party primaries and membership ballots tend to select leaders and candidates who do not represent the interests and preferences of the party core, also because sometimes these mechanisms just change in a sort of protest vote against the elite. Accordingly I hypothesise that the more the process of leadership selection (LS, intended as both the selection of party chairmen and chief executive candidates) is inclusive the more it is likely to select a leader who doesn’t coincide with the one who would be selected by the party elite. In order to assess this hypothesis I will recur to a qualitative comparative analysis of the five biggest European Western countries (UK, France, Italy, Spain and Germany), focusing on twelve contested (i.e. at least two candidates) leadership races (LRs) promoted by the main centre-left party during the last twenty years. On the one side the operationalization of the independent variable (LS inclusiveness) will require to focus on two dimensions (selectorate and candidacy) that must be explored according to a continuum related to the level of inclusiveness/exclusiveness (Le Duc 1999; Hazan 2006). First of all it means to analyse the parties’ formal organization (Statutes and other internal documents disciplining LS rules), in order to assess who are the persons entitled to vote (selectorate) and to run as candidate (candidacy). Anyway the empirical reality quite often differs from the rules, and in this sense I also have to focus on the number of persons who actually go to vote (actual selectorate) but mainly on the number of “real” candidates (i.e. candidates with actual possibilities to obtain the nomination) considering the whole of political factors that could affect the candidates’ offer, maybe hindering some other potential challengers to run (actual candidacy). On the other side the dependent variable (i.e. the coincidence between the actual winner of the LR and the elite’s favourite candidate: CWEF) is assessed on the basis of the formal and informal endorsements for the different candidates by party elite’s members at the national and local level during the electoral campaign. To this end I will recur to press analysis (focusing on LR’s coverage of the two main quality newspapers for each country) and questionnaires submitted to party experts for each country.