When Do Political Parties Respond to their Fringe Voters?

The idea that citizens’ policy preferences should somehow translate into preferences of their representatives constitutes a crucial feature of representative democracy (Pitkin 1967). A close congruence between parties’ policy offerings and voters demands lies at the heart of the process of representation (Schattschneider 1975; Klingemann, Gancheva, and Weßels 2017). The central question arising from this assumption is whether and when political parties indeed listen to voters’ changing policy preferences and correspondingly adjust their policy offerings. Based on the classical Downsian view on party competition, previous research has largely focused on whether political parties respond to shifts in the position of the median or mean voter (MV). By and large, these studies confirm the presence of centripetal forces across West European democracies: if the MV shifts its position on the standard left-right dimension, mainstream parties tend to follow these movements (see Adams 2012 for an overview). In this study, I advance this body of research by focusing on positional movements of those voters that are not positioned around the centre of the left-right dimension. More concretely, I ask whether and when mainstream parties respond to their non-centrist fringe voters (FV). FVs are positioned toward the extreme ends of the left-right scale and are typically undecided whether to cast their vote for a centrist mainstream party or a more radical challenger party – i.e. a radical right, radical left, or green party –. Based on Downsian ideas on third-party competition and recent advances in the field of challenger party success, I argue that the continuing rise of challengers puts increasing centrifugal pressures on mainstream parties that prompt the latter to pay attention to non-centrist voter groups. I test this claim empirically by employing a unique dataset that unifies voters’ left-right positions in combination with parties’ like-dislike scores (national election studies) and party position data (MARPOR/CMP) from 13 West European countries since 1985. Moreover, I investigate when mainstream parties respond to position shifts of their fringe voters rather than movements of the median voter. The empirical analysis shows that MV responsiveness is indeed not the only game in town. Although the analysis confirms previous findings showing that mainstream parties respond to disadvantageous MV shifts (Adams et al. 2004; Ferland 2018), it also shows that mainstream parties from the right and the left respond to shifts of non-centrist fringe voters. This holds especially for those mainstream parties that see themselves under increasing electoral pressure induced by challenger parties.