Resilient? Perceptions, Spread, and Impacts of Misinformation in the New Political Information Environment

Over the past years, Western democracies’ media and political environments have experienced important changes. The proliferation of content producers, algorithmically-driven information distribution, declining trust in the media and governments, the rise of right-wing populism, and political polarization are all conducive to a political information environment in which true and false information coexists and citizens are increasingly divided into “truth publics,” with different realities, facts, authorities, and narratives. This dissertation focuses on how the current political information environment can influence citizens’ perceptions of and vulnerability to misinformation and examines the consequences for social cohesion and democracy.Specifically, this dissertation asks: 1) How do citizens perceive misinformation, and what influences these perceptions? 2) How does the coexistence of different information environments in multilingual countries influence the spread of misinformation? 3) How is misinformation related to societal polarization? In response to the first question, the first two chapters show that citizens have a broad understanding of misinformation, perceive many different forms of misinformation as being prevalent and harmful to democracy, and continue to be critical of politicians spreading misinformation. While perceptions of misinformation form a relatively coherent belief system, citizens’ perceptions are influenced by their political information environment. Given current political discourses around misinformation, individuals with a right-wing ideology or consuming alternative right-wing media are more likely to perceive media misinformation as prevalent and less likely to perceive misinformation spread by social media users as prevalent and harmful than left-wing and centrist citizens. Right-wing citizens are also more indifferent to misinformation and less likely to support misinformation interventions, partly because they perceive current discourses around misinformation and content moderation as biased against them. I discuss how these findings can hinder the effectiveness of our response to misinformation.To answer the second question, Chapter 3 takes advantage of the high prevalence of COVID-19 misinformation in the United States and differential exposure to U.S.-based information among Canada’s English- and French-speaking populations to evaluate whether language creates a barrier to the spread of misinformation. The results suggest that Francophones insulated from the English-language information environment had somewhat lower levels of misperceptions than exposed Francophones and Anglophones, in part because of their lower exposure to U.S.-based content on social media. Exposed Francophones (i.e., bilinguals), especially heavy social media users, were slightly more likely to believe and spread misinformation online. However, compared to Anglophones, their misinformation-sharing behaviors were less dependent on their exposure to U.S. content. This chapter highlights the necessity of considering the globalized and interconnected nature of information environments when evaluating national resilience to misinformation.Finally, Chapter 4 introduces the concept of issue-based affective polarization – the distance between citizens’ positive feelings towards those who share their issue positions and negative feelings towards those who do not. It provides insights into the third question by showing that misinformation contributed to the high level of affective polarization on COVID-19 vaccines and climate change among the Canadian public by intensifying opinion divergence on these issues. Finally, it shows that affective polarization can persist even as the issue becomes less salient. The concluding chapter discusses the theoretical and practical implications of these findings

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.