Abstract
Affective polarization, characterized by increased negativity and hostilitytowards opposing groups, has become a prominent feature of political discourseworldwide. Our study examines the presence of this type of polarization in aselection of European parliaments in a fully automated manner. Utilizing acomprehensive corpus of parliamentary speeches from the parliaments of sixEuropean countries, we employ natural language processing techniques toestimate parliamentarian sentiment. By comparing the levels of negativityconveyed in references to individuals from opposing groups versus one's own, wediscover patterns of affectively polarized interactions. The findingsdemonstrate the existence of consistent affective polarization across all sixEuropean parliaments. Although activity correlates with negativity, there is noobserved difference in affective polarization between less active and moreactive members of parliament. Finally, we show that reciprocity is acontributing mechanism in affective polarization between parliamentariansacross all six parliaments.