transfer 30(3) » Interpersonale Kommunikation

Political Expression in Private Group Chats: Social Capital and Perceived Discussion Safety in Instant Messaging Groups

Political communication increasingly takes place in everyday digital environments, yet research has paid comparatively little attention to political expression in instant messaging (IM) groups. Unlike social networking sites such as Facebook or Instagram, which are often studied as (semi-)public arenas shaped by context collapse, IM groups tend to be smaller, more stable, and more personally embedded. Political expression may therefore emerge incidentally in everyday conversation and be closely tied to interpersonal relationships and perceived risks of disagreement. Against this background, this study examines how relational characteristics and perceived discussion safety shape political expression within IM groups. Drawing on social capital theory, relationship strength, trust, and received social support are conceptualized as relational resources, while perceived discussion safety adapts Edmondson’s (1999) concept of psychological safety to informal political communication.

The study uses secondary survey data from a quota sample of the German population (N = 733). Respondents reported on specific IM groups, including family, friendship, hobby, and work groups. The data were analysed using multilevel models to account for individual-level characteristics and group-specific communication contexts. The analysis tests the links between relational resources, perceived discussion safety, and political expression step by step.

The findings show that political expression in IM groups is shaped not only by individual political predispositions, but also by the relational conditions of the respective group. Received social support emerged as the strongest and most consistent positive predictor, whereas relationship strength was only conditionally related to political expression and trust was negatively associated with it. Perceived discussion safety played a more limited role: although linked to some relational resources, its direct association with political expression was weak. Group type also mattered, with political expression being most common in family, friendship, and neighbourhood groups.

Overall, the multilevel design highlights the context-sensitive nature of political expression in IM groups as interpersonal environments in which relational resources and perceived safety shape whether individuals voice political views.