Deviating from the Norm? Changing Preference Structures as a Hindrance to Party Responsiveness

Normally distributed voter preferences represent one core assumption of spatial models of party competition. As already stated in the Downsian framework, political parties will cater to the demands of centrist voters if citizens’ preferences approximate a normal distribution. Recent trends in established democracies, however, suggest that voter distributions have changed fundamentally: Increasing polarization, multi-dimensional preference structures, and the fragmentation of party systems indicate that normally distributed voter preferences are no longer the rule but an exception if competition is reduced to a single dimension. Interestingly, there are more or less no studies focusing on the consequences of this non-fulfilment of a core assumption of spatial models of competition in current empirical applications of the spatial framework. Our paper, first, sets out to investigate these changes regarding voter distributions in more than 20 established democracies since the early 1990s. Focusing on three crucial features of preference distributions (skewness, kurtosis, the number of local maxima) and making use of comparative survey data (Eurobarometer, CSES), we show how the preferences structures of citizens have changed. In a second step, we investigate the consequences of these developments. Linking survey data to party position data, we show how changes in voter distributions negatively affect party responsiveness toward the mean or median voter. Being confronted with less clear messages concerning voters’ preference structures, political parties become increasingly uncertain about which programmatic shifts could help them to maximize their electoral support. Hence, the more strongly voter distributions deviate from a normal distribution, the less responsive we expect political parties to be (able to be) toward centrist voters. This finding has important implications for our understanding of a crucial feature of representative democracy, namely, how political parties respond to shifts in public opinion. Moreover, it might have additional consequences for the application of ‘standard’ spatial models and could explain heterogenous findings when comparing different contexts.