Presidential Debates in the Age of Partisan Media: A Field Experiment

Abstract

To investigate media effects in political campaigns, we administered a field experiment around the first general election presidential debate of 2016. In this three-wave study, subjects were randomly incentivized to watch the debate and post-debate television coverage on Fox News or MSNBC. We find that post-debate coverage has strong effects on performance evaluations, with subjects' perceptions moving in the direction of the partisan slant of the channel they were assigned to watch on. Though these effects evolve over time, with subjects coming to express beliefs more in line with the media consensus, they do not disappear altogether. Moderate partisans were still affected by the post-debate coverage a week later. The effects we observe are limited to debate evaluations; at no point do we observe partisan media affecting vote choice. Our results offer evidence that media effects can persist over time, but are confined to candidate evaluations.

Keywords: presidential election, presidential debates, media effects, Trump, Clinton, public opinion, partisan media, Fox News, MSNBC