The Campaign Disinformation Divide: Belief and Sharing in the 2019 UK Election

• In Event: Misinformation: How and Why Misinformation Is Spread

Abstract

We explain why citizens believed and shared false political information during the December 2019 UK general election campaign. We find that the media sources people use to get campaign news are key to explaining why they believe disinformation and share it online. A new campaign disinformation divide—between those who get their information from professional news organizations and those who get it from social media—shapes citizens’ ability to distinguish between true and false information, which, in turn, predicts the likelihood that citizens will share information online.
Using an experimental design, we conducted two surveys of samples representative of the British population (total N=4018). We showed respondents 24 different news statements and asked if they had seen them before, whether they believed them, and how likely they would be to share them on social media. Statements focused, in similar proportions, on the main political parties and political divisions, including Brexit. Some of the statements derived from a database we compiled of all published fact checks from the UK’s four main fact-checking outlets during the six-week election campaign (N=176). These summarized disinformation, often originating from political elites, that was circulating on social media or in news coverage and which had been debunked. Other statements were placebos, which did not feature in the campaign but were carefully constructed to read as if they were actual false statements spreading during the campaign. Finally, we included some true statements that were being reported during the campaign. Including placebos and true statements allowed us to control for the extent to which respondents were likely to believe and share any kind of message, and to estimate their abilities to distinguish between true and false information. We measured exposure to campaign news via twelve different sources. A confirmatory factor analysis combined these into two groups: professional news media (television, radio, broadsheet, tabloid and local newspapers, and their online websites) and social media (Facebook, Twitter, Instagram, YouTube, blogs, and instant messaging apps).
Overall, respondents were more likely to believe true than false statements. However, a mixed-effects multivariate regression, with relevant controls and with random intercepts for respondents, revealed that the more people got their campaign information from professional news organizations, the more they were able to distinguish between true and false statements. In other words, people who reported more frequent exposure to professionally produced news were significantly more likely to believe true than false statements. In contrast, the more respondents got their campaign information from social media, the less discerning they were: respondents who reported more frequent exposure to news via social media were significantly more likely to believe false than true statements. Respondents were also more likely to believe a statement if it focused on a policy issue they considered salient and if it was consistent with their prior political preferences, though these effects were more pronounced for true statements.
Perplexingly, respondents were slightly more inclined to share false than true statements. Also, the more respondents believed in a statement, the more likely they were to share that statement if it was false rather than true. Thus, although false information was less likely to be believed than true information overall, once a threshold is breached and false information is believed, that false information becomes more shareable than true information. In addition, statements that were true were less likely to be shared if they conflicted with respondents’ political views. The use of social media for campaign news and the fact that a statement focused on issues salient to a respondent also increased the likelihood that a respondent would share any statement, both true and false.
Our study makes a unique contribution to knowledge about disinformation in contemporary democracies. We reveal a stark division caused by the extent to which sources of campaign information enable citizens to distinguish false from true content. Exposure to campaign coverage produced by news organizations substantially increases citizens’ ability to differentiate true from false information, even if such false information originates with political elites and has circulated widely. By contrast, exposure to campaign news on social media reduces citizens’ discernment, rendering them more vulnerable to believing false information. And because false information that is believed is also more likely to be shared than true information, the relationship between exposure to news and belief in disinformation has a behavioral outcome, in the form of online sharing. These findings highlight the centrality of professional news media in enabling citizens to protect themselves and others from disinformation.

Authors

• Cristian Vaccari, Loughborough University

• Andrew Chadwick, Loughborough University

• Johannes Kaiser, Loughborough University