Antecedents and consequences of fake news exposure: a two-panel study on how news use and different indicators of fake news exposure affect media trust Sangwon Lee 1, *, Homero Gil de Zu ´ ~ niga 2,3 , Kevin Munger 4 1 Department of Communication Studies, New Mexico State University, Las Cruces, USA 2 Department of Political Science, University of Salamanca, Salamanca, Spain 3 Department of Film Production and Media Studies, Pennsylvania State University, University Park, USA 4 Department of Political Science, Pennsylvania State University, Philadelphia, USA *Corresponding author: Sangwon Lee. Email: swlee@nmsu.edu Abstract Despite abundant studies on “fake news,” the long-term consequences have been less explored. In this context, this study examines the dynamic relationship between traditional and social news media use, fake news exposure—measured as perceived fake news exposure and exposure to actual fake news stories, and mainstream media trust. We found interesting patterns across two U.S. panel survey studies. First, we found that exposure to fake news—regardless of how we measured it—decreased people’s trust in the mainstream media. Yet, we also found that while both social media and traditional news use were positively associated with exposure to actual fake news stories, only social media news use was positively associated with perceived fake news exposure. This finding implies that while many people believe that social media is the culprit of fake news exposure, traditional news use may also contribute to people’s exposure to popular fake news stories. Keywords: fake news exposure, perceived fake news exposure, social media, traditional media, news media trust, panel data According to a recent Gallup survey, Americans’ trust in the media to report the news fully, accurately, and fairly is much lower since Gallup first started this survey in 1972 (Brenan, 2021). Only 9% of respondents said they trust the mass me- dia “a great deal” and 60% said they have little to “no trust at all.” Similarly, according to the Gallup and Knight Foundation (2018), more Americans have a negative view of the news media (43%) than a positive one (33%), and 66% of Americans think most news media fails to separate facts from opinions. Low trust in news media is not simply a matter of concern for journalists; rather, it is directly related to the quality of democracy. In a democratic society, news media plays the essential role of generating and distributing informa- tion, allowing citizens to make important political decisions based on information acquired from the news media (Brundidge et al., 2014). If citizens become skeptical toward even the simplest of facts they encounter through the media while downgrading factual information to mere opinion, then the fundamentals of democracy will be threatened. Against this background, scholars have been interested in finding factors contributing to declining trust in the main- stream media. Previous studies show that (dis)trust in the mainstream media is related to multiple factors such as news exposure, education, partisanship, news coverage style, and so on (see Valenzuela et al., 2022). In addition to these fac- tors, scholars are now paying attention to the role of fake news, defined as “the deliberate creation of pseudo- journalistic disinformation” (Egelhofer & Lecheler, 2019, p. 97), as it often contains information that denigrates main- stream news coverage (Ognyanova et al., 2020). Although some studies are beginning to explore the relationship between fake news exposure and mainstream media trust (Guess et al., 2020; Ognyanova et al., 2020), there are still gaps to fill. First, previous studies have largely relied on a single mea- surement of fake news exposure. For instance, some studies measured fake news exposure by measuring people’s self- perceived exposure to fake news (e.g., “How much have you seen fake news stories?”; Jones-Jang et al., 2021; Wasserman & Madrid-Morales, 2019), which has limitations in capturing the actual exposure to fake news stories, especially given that many people tend to categorize news that they do not like as fake news (Tong et al., 2020). Some studies took a fairly differ- ent approach by using computational methods, where they measured fake news exposure by tracing whether people had access to untrustworthy websites (e.g., Guess et al., 2020; Ognyanova et al., 2020). This approach also has limitations in that what they measured is one’s exposure to untrustworthy websites, while most people do not encounter fake news stories from untrustworthy websites (Tsfati et al., 2020). To comple- mentarily build on these studies, we attempt to use two differ- ent indicators of fake news exposure. That is, in addition to perceived fake news exposure measure, we employ an addi- tional type of fake news exposure measure from Valenzuela et al. (2019) and Diehl and Lee (2022). This is operationalized by showing people actual examples of popular fake stories cir- culated online during the presidential election. Subjects then reported whether or not they had heard about each statement. Like all scientific measurements, neither of these measures is a perfect reflection of the reality of fake news exposure, but our triangulation across measures represents an advancement over previous research in this area. Received: 25 April 2022. Revised: 7 March 2023. Accepted: 9 March 2023 V C The Author(s) 2023. Published by Oxford University Press on behalf of International Communication Association. All rights reserved. For permissions, please email: journals.permissions@oup.com Human Communication Research, 2023, 00, 1–12 https://doi.org/10.1093/hcr/hqad019 Original Research Downloaded from https://academic.oup.com/hcr/advance-article/doi/10.1093/hcr/hqad019/7111256 by guest on 10 April 2023