Territorial inequalities as drivers of affective polarization in Europe: disentangling within and between countries dynamics

In recent years, a number of political events - such as Brexit and the increasing share of votes for extremist parties throughout Europe - have made evident how territorial inequalities are driving political polarization, thus emphasizing the political consequences of regional socio-economic conditions. At the same time, this rise of populism and radicalism has induced scholarly attention on a previously overlooked form of political polarization, namely affective polarization. Broadly defined as the degree of resentment towards opposing parties and their supporters, affective polarization has proven to be particularly dangerous from different perspectives, e.g. it contributes to emotional reactivity, causing negative effects both inside and outside of the political sphere. However, recent studies have mainly adopted either a cross-country, party or an individual level approach in the study of the affective polarization phenomenon. In this paper, we propose that there is scope for investigating affective polarization at regional level, since country-level analysis may hide significant within country heterogeneity, both in degrees and trends of affective polarization and of its drivers. In so doing, we make use of the Comparative Study of Electoral Systems (CSES) dataset to compute an affective polarization score for 130 regions, nested in 30 countries, across a period ranging from 2013 to 2019. First, we demonstrate the theoretical and empirical interest of a disaggregated approach, since the range of affective polarization scores is larger across regions than between countries, with over half of the variation in scores being attributable to within-country heterogeneity. Then, we make use of econometric techniques to analyse (i) the impact of a set of explanatory variables on regional levels of affective polarization; (ii) if such direct effects are mediated by between- and/or within-country dynamics. We put special emphasis on variables pertaining to economic performance and quality of institutions, hypothesizing regions that perform worse – especially compared to other regions in the same country – are affectively more polarized. We will also account for the commonly used political variables (ideological polarization, partisanship), cultural aspects and sociodemographic structure in our analysis. Our findings will contribute to a better understanding about the structural foundations of affective polarization and exemplify the relevance of regional inequalities.