How in-person conversations shape political polarization: Quasi-experimental evidence from a nationwide initiative

Growing political polarization is often attributed to "echo chambers"among like-minded individuals and a lack of social interactions among contrary-minded individuals. We provide quasi-experimental evidence on the effects of in-person conversations on individual-level polarization outcomes, studying a large-scale intervention in Germany that matched pairs of strangers for private face-to-face meetings to discuss divisive political issues. We find asymmetric effects: conversations with like-minded individuals caused political views to become more extreme (ideological polarization); by contrast, conversations with contrary-minded individuals did not lead to a convergence of political views, but significantly reduced negative beliefs and attitudes toward ideological outgroup members (affective polarization), while also improving perceived social cohesion more generally. These effects of contrary-minded conversations seem to be driven mostly by positive experiences of interpersonal contact.