Résumé : Affective polarization – the extent to which voters feel positively about their own party and its voters and negatively about other parties and their voters – has become a central concept in the study of political behaviour in contemporary democracies. A key finding is that voters are more affectively polarized around elections, which has raised concern about the potential destabilizing role of electoral competition. This dissertation argues that assessing the normative implications of election-induced affective polarization requires understanding both (1) how elections drive affective polarization and (2) how affective polarization influences political behaviour during elections. Drawing on five empirical chapters that use original and secondary survey data from European multiparty contexts, the findings show that elections generate small increases in vertical (toward parties) but not in horizontal (toward voters) affective polarization. Crucially, these changes are not driven by rising hostility toward political opponents, but by differentiated increases in sympathy toward all parties and electorates. In addition, policy congruence with other parties – an inherent feature of multiparty systems – reduces hostility toward their voters, highlighting how multiparty systems contain institutional safeguards that constrain extreme affective polarization. Regarding political behaviour, affectively polarized citizens are more likely to vote across a range of electoral contexts, and this effect is more robust than that of ideological polarization. Finally, in the aftermath of elections, affective polarization widens winner-loser dynamics, but this is primarily driven by electoral winners, who may reduce support for institutional constraints. Taken together, these findings nuance the impact of elections on affective polarization and show that election-induced increases in affective polarization do not uniformly undermine democratic functioning. They suggest that democratic stability does not require the absence of affective polarization, but rather its containment within manageable bounds that structure political conflict.