Evaluating digital polarization in multi-party systems: evidence from the German Bundestag
In: WebSci '22: proceedings of the 14th ACM Web Science Conference, Barcelona, Spain, 26-29 June 2022, 296–301
New York, Association for Computing Machinery (ACM) (2022)
Social networking sites like Twitter and Facebook have become dominant sources of political activity with many politicians choosing to leverage these platforms. This rise in popularity has led many researchers to investigate the impact of these platforms on digital communication, specifically with regard to digital polarization. Most of the current polarization literature is centered on two-party systems with little attention given to multi-party parliamentary networks. As such, we leverage data from members of the German Bundestag (MdBs) to better understand digital polarization in MdB Twitter interactions. This paper expands the current literature to include a multi-level network evaluation of a parliamentary body by evaluating partisan polarization on the scale of retweets, mentions, and following-follower relationships. We employ social network analysis and text-based sentiment analysis to understand both the structure and sentiment polarity of the networks and find that polarization varies based on the level of the network that is being examined. Results indicate that the highest polarization occurs in the retweets network and the lowest occurs in the mentions network, while the sentiment analysis suggests same-party mentions are significantly more positive than cross-party mentions.