Analyzing Interactions between Political Demand and Supply in Europe: The Case of Protest Voting

The poor health of citizens’ relations with their democracies is a long debated topic in the field of social and political sciences (e.g. di Palma 1970; Montero and Torcal 2006). However, to say it in Hirschman's terms (1970), people’s discontent towards their authorities does not necessarily lead to an exit from the political life, but also to a possible voice strategy, i.e. the adoption of new and specific patterns of political action. Casting a protest vote would be among these possibilities. So far the effectiveness of this kind of electoral behavior has been often inferred by its alleged consequences at the macro level, such as electoral earthquakes and shocks in the morphology of the existing party systems. Nevertheless, it is not sure whether these contextual dynamics completely coincide to protest voting, i.e. casting a vote with the only aim to frighten and to punish the political system and/or èlite (Van der Eijk et al.1996; Van der Brug et al. 2000). Several scholars have tried to deal with this topic in the past, but a unified analytical framework still lacks in this field (Van der Brug and Fennema 2003). In particular, the issue about how do voters come to perceive specific parties as credible conveyers of an electoral protest has been so far partially disregarded. The aim of this paper is to make some first improvements in this direction, offering a renewed framework and EU-wide comparative analysis of how individual level attitudes and dynamics of party competition may produce, by interacting, protest voting under varying conditions of the political macro-contexts. My main hypothesis in this regard is that the more a party appears as different from its establishment competitors, the higher its chances to mobilize protest votes, i.e. negative attitudes towards the political system and mainstream party performance (e.g. institutional distrust, political inefficacy and government evaluation) as fundamental considerations preceding vote choice. I also suppose this difference to develop through different party dimensions over time: an office-seeking dimension, a policy-seeking dimension and an organizational dimension. Such expections will be tested through several multi-level regression models on data from the European Elections Studies (EES) over a considerable time span (1989-2014). Further tests will be realized also by rearranging the data in the so-called ‘stacked’ form .