Media Use and Democratically Corrosive Attitudes

• In Event: News Media Diets and Their Consequences

Abstract

Democratic backsliding is a major concern in the US, as in many other countries. One locus of particular concern is what can be called democratically corrosive attitudes. These attitudes involve support by publics for authoritarian candidates, acceptance of force or violence to obtain specific political objectives, rejection of the legitimacy of political disagreement, false or conspiratorial beliefs about institutions and democratic process, and so on.

The literature shows that democratically corrosive attitudes are predicted by a set of underlying psychological characteristics and orientations. These include Right-wing Authoritarianism, Social Dominance Orientation, ethnocentrism, perceived victimhood, and others. We are interested in how these become translated into specific, corrosive attitudes about public affairs. Perhaps the most well-known influence in the literature is the availability of political elites who visibly violate democratic norms, and who offer the public populist messages, especially in the context of perceived threats to economic security and identity.

The role of media use in the translation of psychological characteristics into democratically corrosive attitudes is not well understood. There are a host of reasons to expect that media matter, from studies of populist leaders’ communication strategies to research on social media and disinformation. There are also many reasons to expect that different forms of media play different roles. In this study, we aim to bring media squarely into the picture, by asking: How does exposure to political communication through media shape the relationship between psychological characteristics and democratically corrosive attitudes?

We approach this question from the perspective of media “diets” in the US. Our approach draws on two distinctions that have been explored by scholars in other contexts. The first is the difference between mainstream news media and the right-wing media ecosystem. We expect differences not only in content but also in the extent of selective exposure exhibited by audiences. The second distinction is among types of social media, especially between those that facilitate thick-tie as opposed to thin-tie networks. We expect more heterogeneity in content among thin-tie apps such as Twitter and Youtube, compared with thick-tie apps such as Facebook.

Media diets should matter through both mediation and moderation. For instance, where audiences are more selective and content narrower, we expect a positive mediating effect of media use, as people with psychological predispositions select media content that promotes anti-democratic norms with respect to specific problems and events in the news. Where audiences are less selective and content is broader and falls within democratic norms, media use may exhibit a moderating effect in the negative direction. Our paper will present several hypotheses about these expectations.

Our analysis will draw on original survey data that we collected in the US in 2021 (N=2000), which we are analyzing now. We measured a variety of psychological characteristics and orientations using standard measures (e.g. SDO, RWA, ethnocentrism) as our independent variables. Our dependent measures include standard scales for populism and conspiracy orientation along with a number of our own custom measures about beliefs in specific conspiracies, beliefs about election fraud, trust in democratic process.

We also employed an extensive battery of media questions of our own design that asked about attention to specific news media brands (e.g. Fox News, New York Times, CNN) and about use of specific social media apps (e.g. Facebook, YouTube, “free speech” apps like Truth Social). These serve as mediators and moderators in our model.

Our primary analysis will involve a set of OLS models with moderation and mediation by media use between the independent and dependent measures.

Part of our analytic effort will involve an attempt to identify a factor structure among our various dependent measures of democratically corrosive attitudes. There has been a lot of debate about predictors of populism, and we see a lot of room for improvement in how scholars conceptualize democratically corrosive attitudes. The literature is ripe for an improved understanding of how various measures of corrosive attitudes such as populism and mistrust in elections are structured.

Another part of the analysis will involve an examination of the structure of media diets, focusing on overlap between various mass media and social media brands.

Authors

• Bruce Bimber, UC, Santa Barbara

• Daniel M Gomez, New Mexico State University

• Julien Labarre, California State University, Dominguez Hills

• Ilia Nikiforov, University of California, Santa Barbara

• Karolina Koc Michalska, Audencia Business School