Uncivil and personal? Comparing patterns of incivility in comments on the Facebook pages of news outlets

Social media and its embedded user commentary are playing increasingly influential roles in the news process. However, researchers’ understanding of the social media commenting environment remains limited, despite rising concerns over uncivil comments. Accordingly, this study used a supervised machine learning–based method of content analysis to examine the extent and patterns of incivility in the comment sections of 42 US news outlets’ Facebook pages over an 18-month period in 2015–2016. These outlets were selected as being broadly representative of national, local, conservative, and liberal-news media. The findings provide the first empirical evidence that both the level and the targets of incivility in the comments posted on news outlets’ Facebook pages vary greatly according to such entities’ general type and ideological stance.