Leaders over parties? : the personalization of politics and voting behavior in contemporary Western democracies

Abstract

This thesis is a compilation of four empirical studies analysing the impact of the personalization of politics on voting behaviour in contemporary Western democracies. The process of personalization of politics has been theoretically described as resulting from the cumulative effects of three factors: a) an individualization of voting behaviour emerging from the erosion of cleavage-based voting and consequent process of partisan dealignment; b) transformations in the structure of mass communication, particularly with the diffusion of television as voters’ main source of political information and parties’ preferred channel of political communication; and c) a process of party change and adaption to a changing environment, reshaping their electoral profiles, modes of operating and organizational structures. The present thesis attempts at providing empirical evidence of these theoretical linkages by (i) demonstrating how personalization ultimately results from the decline of the role of partisanship in guiding vote choice, and therefore is the consequence of partisan dealignment; (ii) providing evidence of the impact of the change towards a television-based media diet in driving the personalization of vote choice; and (iii) accounting for the role of party organizational change – namely through the introduction of more open leadership selection procedures and further concentration of powers at the leadership position – in heightening leaders in contemporary politics. Furthermore, it explores a novel avenue concerning leadership effects by investigating whether leadership evaluations can also impact voters’ turnout decisions. It relies on two types of data sources: one dataset constructed using data from the Comparative Study of Electoral Systems project and the Political Party Database; and one original dataset pooling over 129 national election surveys conducted in 14 Western democracies in the period between 1961-2016. The results contribute to the ongoing scholarly debate on the topic by providing extensive evidence for the electoral dimension of the personalization of politics.