PhD Thesis defence by Emma Hoes.
This dissertation assesses the harms of actual misinformation and compares it with the potential harms and benefits of the media's attention to as well as interventions against misinformation. The attention and interventions, what I together refer to as the coverage of misinformation, I argue not to be in proportion to the magnitude of the problem of misinformation itself. To test this argument, my dissertation disentangles which type of exposure - actual misinformation, the coverage of misinformation, partisan news, or a media literacy intervention - fosters misperceptions, lowers trust and increases skepticism about verified facts the most. The dissertation is structured in four chapters.
In Chapter 2, I assess vote intentions during the 2016 Brexit Referendum Campaign. In line with the media's minimal effects theory, I find that media exposure during the campaign period did not affect the vote intentions of the majority of people. This finding builds the groundwork for the remaining three chapters of my dissertation, as for misinformation in the media to have an effect, there should be a media effect to begin with. It raises a fundamental question: if the media in general change so little about people's opinions, then why is so much attention paid to misinformation disseminated in the media, and on what grounds are interventions against misinformation in the media implemented?
In Chapter 3 and 4, I compare the effect of exposure to actual misinformation to the coverage of misinformation and - only in Chapter 4 - to exposure to partisan news. I do so through an online survey experiment fielded in Poland (Chapter 3) and by using panel- and online trace-data in the US (Chapter 4). I find that not only misinformation itself but also the media's intention to raise awareness about the issue of misinformation may have negative downstream effects by lowering trust in various institutions, such as the media and scientists, by discouraging citizens to inform themselves, and by potentially distorting belief in factually accurate information. In Chapter 4, I additionally find that not exposure to misinformation - measured through actual visits to untrustworthy sites - but rather partisan news fosters misperceptions.
Finally, in Chapter 5, I ask how a media literacy intervention may help citizens to accurately assess the reliability of true and false news items. Leveraging a natural setting in a Dutch High School and using a Difference-in-Differences approach, I find that a media literacy intervention improved students' accuracy.
Taken together, the findings of my dissertation ask for a careful reconsideration of several interventions against misinformation, as some of these interventions may do more harm than good by further lowering public trust in a host of institutions important to a well-functioning democracy.
Emma Hoes is a postdoctoral researcher in the ERC Project Problem Definition in the Digital Democracy (PRODIGI) and at the Digital Democracy Lab of the University of Zurich. Her research focuses on several challenges that came about with the advancement of digital technologies, such as misinformation, micro-targeting, online content moderation, and - more broadly speaking - the role of social media in our daily media-diet. She is particularly interested in the extent to which the numerous interventions against these potentially harmful phenomena are effective, but especially to what extent such interventions may cause unintended spill-over effects.