The Democratic Value of Strategic Game Reporting and Uncivil Talk: A Computational Analysis of Facebook Conversations During U.S. Primary Debates

Abstract

This study explores discourse features on Facebook pages of news organizations during the 2020 U.S. primary debates using a state-of-the-art machine-learning model. Informing the scholarly debate about the implications of strategic game reporting in online spaces, we find that it is not necessarily linked to uncivil discourse, yet it might deter from relevant conversations. Second, addressing fears about the undesired outcomes of uncivil talk, our data suggest that incivility can coexist with rational discourse in user comments, although this relationship is not pervasive. Implications of these results are discussed in the context of the role of hybrid media for political engagement during electoral campaigns.