Friends or foes : does polarisation really undermine democracy? Understanding the two-way relationship between polarisation dynamics and the processes and institutions of democratic contestation

The relationship between polarisation and democracy has become a central concern in contemporary political science, yet empirical findings on whether affective polarisation undermines democratic attitudes remain inconclusive. In this dissertation I address this puzzle by arguing that two distinctions have been overlooked in existing research: different types of affective polarisation and different types of democratic institutional design. Having traced its origins, I define polarisation as the increasing distance between actors based on political conflict. I distinguish between idea-based and identity-based conflict, and between various forms of increasing distance. Based on that conceptual development, I investigate how polarisation relates to democracy across different contexts through four empirical studies examining both micro-level polarisation dynamics and macro-level institutional frameworks. Studies 1 and 2 employ a survey experiment (N = 2,000) in Belgium and the UK to examine the differential effects of various types of affective polarisation on antidemocratic attitudes. The findings reveal striking patterns: affective polarisation measured as dislike toward political parties shows either no relationship or is negatively associated with anti-democratic attitudes (less support for violence, more support for liberal democracy). In stark contrast, when polarisation is measured towards an individual situated within a specific societal conflict (wokeness) and operationalized through social distance and negative emotions rather than mere dislike, significant positive relationships with anti-democratic attitudes emerge. Specifically, social distance predicts higher political intolerance, while emotional distance (hate, anger, frustration) strongly predicts both support for political violence and lower support for liberal democracy. These patterns hold consistently across both countries. Studies 3 and 4 shift focus to macro-level institutional factors, analyzing crossnational data from 38 democratic countries (2000-2019) and 113 countries (1900- 2023) respectively. The findings demonstrate that consensus democratic institutions are systematically associated with lower levels of polarisation compared to majoritarian vii institutions. Countries with federalism, coalition governments, and proportional representation exhibit significantly less polarisation on average, with these findings being more pronounced for identity-based than idea-based polarisation. Furthermore, consensus institutions weaken the relationship between polarisation and political violence, with federal systems and strong constitutional rigidity serving as particularly effective buffers against polarisation's most destructive consequences. These findings contribute to both polarisation and democracy research by demonstrating that polarisation is not a monolithic concept—different types have different democratic implications. The research challenges the dominant pessimistic narrative in polarisation studies, showing that classical partisan measures may miss the mark and that not all polarisation is problematic for democracy. For democracy research, the study reveals that polarisation does not inherently undermine democratic systems. Even stronger, the findings demonstrate that democracies are not defenseless – consensus institutions can effectively buffer against polarisation’s most destructive effects. Rather than treating polarisation as inherently problematic, I suggest that democracies are fundamentally engineered to manage conflict, with institutional design playing a crucial role in determining whether polarisation becomes destructive or remains within the bounds of healthy democratic contestation.