Can political discussions reduce affective polarization? Evidence from online chat discussions in the Netherlands

Can political discussions reduce affective polarization? Most studies on this question to date have taken place in a lab-based, non-political and non-divisive setting, usually priming and paying participants before they discussed. In contrast, this study uses data collected in the run-up to the Dutch 2021 general election from a platform that ran voluntary political discussions. The data provides a unique opportunity to study contact between different political groups in a real-world environment. Using expectations derived from contact theory, it finds that discussions can reduce affective polarization, that discussions perceived as negative increase it and that discussions between political ingroups have no effect. In addition, using open-ended response items that asked participants to evaluate their experience, it shows that mechanisms predicted by contact theory seem to hold well in a setting of political ingroups and outgroups. Taken together, this study shows the real-world potential of political discussions to reduce affective polarization, extends mechanisms proposed by contact theory to political ingroups and outgroups and has important implications for broader deliberative processes.