Activating Antagonism : Populism and Ego-Network Structure among German MPs on Twitter

Sammanfattning

The thesis explores how populist rhetoric by German parliamentary members shapes the structure of engagement communities on Twitter. From a minimal definition of populism, a systems-theoretic, social-psychology-informed account of populism’s meso-level dynamics is developed. This extends prior work centred on engagement metrics toward the structural question of whether politicians usage of populism on Twitter induces denser local engagement networks around the politician. The analysis is basedon a dataset of German MPs’ Twitter posts from 7.–14. February 2022 combined with the reply threads beneath the tweets. A Qwen3-235B LLM rates tweets on core dimensions of populism through few-shot classification with an expert-validated systemprompt. The rated tweets are combined into a user-level reply network. Politicians’ ego networks are extracted and filtered to regress a politician’s use of populism on the connectedness of the respective engagement community. Populist German MPs occupy more strongly connected engagement communities than non-populist MPs. The impact of alters on tie creation is additionally explored. Alter interaction patterns show slight populism-homophily, with populist alters’ replies bridging alters of differing activity volumes. Populist community formation is consistent with established self-selection accounts of populist mobilisation.