Affective and Ideological Polarisation in Swedish Party System (1985-2015): Voter and Elite Level Analysis

Research on political polarisation has traditionally concentrated on ideological differences and extremity among voters and political parties. Recently, several US scholars have started to address another manifestation of polarisation, namely affective polarisation, which is defined as (increased) partisan animosity between the Democrats and the Republicans. While a certain degree of ideological polarisation is usually perceived as necessary to make party competition function properly, affective polarisation is unequivocally perceived as an undesirable phenomenon that bears negative consequences such as hindering co-operation between parties and decreasing satisfaction with democracy among the supporters of the non-incumbent party/parties. Despite the gravity of the topic in the United States, the debate on affective polarisation in Europe has been nearly non-existent. We aim to address this gap by studying affective polarisation in Sweden over the course of a 30-year period, utilizing vast datasets covering both voter and party elite (MP) level. We have identified three important and puzzling phenomena that merit further attention. First, there has been a substantial in-bloc convergence in partisan affect within the left/centre-left parties on the one side and the right/centre-right on the other, meaning that since the 1980’s both the MP-s and party supporters have started to exhibit much more positive feelings towards the other parties in their own ideological bloc. Secondly, the MPs and voters of the two biggest parties - the Conservatives and Social Democrats - have experienced a sharp decline in affective polarisation during these years (contrary to the Democrat-Republican trend in the USA). Interestingly, this has happened without any similar convergence in the parties` (and voters`) ideological positions on the central left-right dimension. Finally, emergence of the right-populist Sweden Democrats has drastically altered the dynamics of affective polarisation in Sweden. The MPs and voters from all other parties have consistently disliked the party at the highest level, while the Sweden Democrats` feelings towards their opponents have been more moderate, indicating asymmetric polarisation. This extremely negative sentiment towards Sweden Democrats has increased the overall level of affective polarisation in the party system by 2014-2015 to a similar level as in the 80’s, despite the convergence of the two biggest parties. This paper aims to determine the causes behind these phenomena. We consider (perceived changes in) ideological distances between parties/partisans as the main potential independent variable. However, it is clear that left-right positions do not explain all of the variance in affective polarisation, so we measure ideological polarisation also on the cultural dimension and test whether some specific issues (e.g. immigration) invoke stronger affective reactions. In addition, we will study whether party elite behaviour and attitudes could drive the affect among partisans (or vice versa) independently or interactively with ideological distances. As such, our paper contributes into the nascent literature on the determinants of affective polarisation and the hitherto unclear linkage between affective and ideological polarisation. Moreover, as our data permits measuring affective polarisation also on the party elite level, we can scrutinize the elite-voter link in a more direct way than previous research.