Threat Level Midnight: Examining the Relationship Between Threatening Language and Engagement with Non-Partisan, Partisan, and Hyper-Partisan Media on Facebook

This study extends the literature on shareworthiness and news values to examine threatening language as a predictor of engagement with content from the social media accounts of various media organizations, ranging in their ideological extremity. We first examine the relationship between threatening language and engagement with Facebook news content from 20 different media organizations. Next, we examined whether this relationship varied based on the ideological extremity of a media organization, ranging from nonpartisan to hyper-partisan. Finally, we examined whether, among partisan and hyper-partisan outlets, the relationship between threatening language and engagement varies based on whether the in- or out-group is mentioned. In general, our results revealed mixed findings for the relationship between frequency of threatening language and engagement levels, with the direction of this relationship varying by metric and year. Additionally, we found that this relationship varied based on the ideological extremity of the media outlet and whether group-related language was utilized in the content as well.