Revisiting the protest voting hypothesis : an empirical analysis on six EU countries

Often, in public debate, the effectiveness of an electoral protest is inferred by specific dynamics at the macro level, such as the occurrence electoral earthquakes and/or shocks in the morphology of the existing party systems. Yet sudden increases in the electoral outcomes of specific parties are not necessarily due to pure protest. Indeed, one could well contribute to their success because (s)he likes their policy platform or because (s)he feels ideologically and/or psychologically closed to them. Reworking the existing literature, in this paper I assume that protest motivations consist of two main components. The first one is political discontent, i.e. people\u2019s dissatisfaction about day-to-day actions of political leaders and in particular the operation of governmental parties. The second one is political disaffection, which is a more general feeling of estrangement from politics, due to a perceived lack of transparency and fairness of institutional processes and outputs. Within this framework, I will test the standard assumption that protest motivations at the individual level are indeed electorally effective, i.e. that they really contribute to those electoral shocks we instinctively recognize as \u201cprotest\u201d election results at the aggregate level. In particular, I expect that new and/or anti-political-establishment parties are those who benefit most from such electoral dynamics. These hypotheses will be tested through several multivariate models on data from the 2014 European Elections Study (EES) with a focus on six Western European countries (Austria, Greece, France, Spain, Netherlands, UK and Italy). Technically, the data matrix will be reshaped in the so-called stacked form (e.g. Tillie 1995a; Van der Eijk and Franklin 1996) while the electoral consequences of protest voting will be assessed by several counterfactual analyses, in which voters\u2019 expected preferences (PTVs) for each party will be estimated under various scenarios in which everything is the same except for the relevance of political discontent and disaffection in voters\u2019 decision-making