Throughout the 2016 U.S. presidential election, public debate and media coverage was shaped by so called "fake news" – news articles which were intentionally false, and designed to influence opinion and policy. Although fake news itself is not a new concept, the way in which it was covered, and the was it was spread on social media platforms, was. Given this, scholarly literature examining fake news, and specifically the content or stylistic characteristics of fake news, is minimal. My research seeks to address that gap through examining different content characteristics of fake news articles spread on social media in 2016, and testing the impact of those characteristics on Facebook engagement (the number of likes or shares an article received).

I find political fake news circulated during the 2016 U.S. election is relatively homogeneous in content: it avoids policy discussion, is highly partisan, and negative in tone. Furthermore, personal content, policy discussion, partisan lean, and article tone have no detectable effect on the engagement received on Facebook. My research serves to provide avenues for future research, and increase our understanding of how fake news is spread. More importantly, given the negative influence fake news has on public discussion and democratic legitimacy, my research also increases our understanding of how to best combat the influence of fake news, and how to limit its spread.