In today's fragmented online media ecosystem, does partisan media have a measurable effect on political attitudes and behavior? Here we show results from a pre-registered, randomized field experiment embedded in a nationally representative online panel survey (N = 1,339) in which participants were incentivized to change their browser starting page, boosting the likelihood of encountering news with either a left-leaning (HuffPost) or right-leaning (Fox News) slant during the 2018 U.S. midterm election campaign. Data on respondents' web visits indicate that changes in media consumption habits persisted up to six weeks later. While we generally find negligible persuasive and agenda-setting effects, we uncover a large, durable decrease in overall media trust among those exposed to Fox News (ATT: -0.62 SD; t=-2.04, p<0.05). In addition, HuffPost strongly liberalized respondents' views on immigration policy (ATT: -1.02 SD; t=-3.02, p<0.005). Consistent with this finding, we demonstrate with topic models that immigration was an especially salient topic during this period.
Andrew Guess is an assistant professor of politics and public affairs at Princeton University.
His research sits at the intersection of political communication, public opinion, and political behavior. He uses a combination of experimental methods, large datasets, machine learning, and innovative measurement to study how people choose, process, spread, and respond to information about politics.
Current or recent projects investigate online selective exposure, how to accurately measure media exposure on the internet, the dynamics of interest group mobilization over Twitter, and the persuasive effect of new information on individuals’ attitudes and beliefs.