Fact-checking in a post-truth world

How do you write about facts in the era of Trump, Brexit, and Ford? Four years of collectively witnessing repeated, concussive attacks on the norms of public debate, science, and democracy make the challenges of defending truth seem insurmountable. Donald Trump ran for re-election on a strategy built around lying about everything from health care to the postal service. Daniel Dale’s three and a half-minute fact-check of Trump’s speech to the Republican convention in August is a triumph of reporting but a chilling reminder of life in a ‘post-truth’ world.ⁱ Worse, the stakes extend well beyond partisan political fights. COVID-19 conspiracies—ranging from simple anti-science libertarianism to outright racism—undermine efforts to curb the pandemic.

This is part of the difficulty about writing about facts and post-truth: how panicked should we feel? While the United States is increasingly and horrifyingly showing us the logical conclusions of post-truth politics, Ontario has not yet descended to that level. Nonetheless, racism in Ontario is real. We have COVID-19 conspiracy theorists and anti-maskers here too. The Ford Conservatives based their 2018 election strategy on populist rhetoric and a scant platform. During the last round of bargaining, OSSTF/FEESO and our allies churned out fact-check after fact-check to correct the Minister of Education’s frequent mischaracterizations of the bargaining process, policy disagreements, and funding claims. At the same time, Canadians appear to have trust in science and its spokespeople.

If we are falling further into a world of post-truth politics, it is worth asking, what is the impact of fact-checking? This article looks at several studies and analyses related to the spread of false claims, fact-checking, political support, and policy preferences to see what the research tells us.

Here are some spoilers for people who don’t want to read all the way through—and also for people who are going to read all the way through, but who may want to take some deep breaths and prepare themselves first.

• Fact-checking appears to bring people closer to the truth. That is, it appears to correct or at least moderate false beliefs.
  
• At the same time, fact-checking does not appear to significantly change political behaviour. Even when politicians are shown to be making false statements, their supporters continue to support them.
  
• Perceptions of credibility and authoritativeness outweigh the actual truth-content of individual statements in forming political opinion.

These findings suggest that fact-checking is a necessary part of political communications, but it is not enough on its own. Instead, successful political communication depends on building up credibility over a longer period. It is this credibility that gives weight to political communications rather than the actual truth-content of any given claim or counter-claim.

Let’s dive into some research.
First, there’s an important study by Vosoughi et al. investigating the spread of true and false news online. In their 2018 study, the authors examined what they called ‘rumour cascades’ on Twitter from 2006 to 2017. Rumour cascades are “instances of a rumour-spreading pattern that exhibit an unbroken retweet chain with a common, singular origin.” ⁱⁱ Cascades begin with a user making an assertion about a topic in a tweet (including text, photos, and/or hyperlinks), which is then propagated through re-tweets. The study found approximately 126,000 cascades, which were spread by 3 million people 4.5 million times.

The authors found that false rumours spread “significantly faster, deeper and more broadly than the truth in all categories of information.” ⁱⁱⁱ Importantly and somewhat discouragingly, the authors were able to control for the extent to which bots—automated Twitter accounts—were involved in spreading both true and false news. They found that humans were more likely than bots to spread false news. This type of study is not able to identify the reason why falsehoods tend to spread faster, but the authors suggest there is good reason to believe it is because false claims tend to be more novel and express greater surprise. By contrast, true claims tended to inspire expressions of sadness and trust. It appears that novelty and outrage outpace empathy, at least on Twitter. This dynamic may put fact-checkers at a built-in disadvantage. At least in terms of novelty, insofar as fact-checkers are responding to an existing claim, then their message is already less novel than the falsehood it challenges.

The dynamic Vosoughi et al. identify is concerning, but it actually gets worse. Two studies look specifically at the cognitive and behavioural impacts of fact-checking. These studies ask: when people receive fact-checks, do they update their beliefs from a false belief to a true one? As importantly, are people less likely to support a politician when they’ve been shown to make false statements? The results are surprising.

In 2019, Nyhan et al. used two studies to look specifically at the impact of Trump’s falsehoods on his supporters and supporters of Hilary Clinton. The studies broke participants into four groups and gave each group an article including one of the following: a false claim, a fact-check, a fact-check with a rebuttal, or a fact-check with a rebuttal and an attack on the official source cited by the fact-check. On the positive side, exposure to fact-checks did tend to reduce misinformation among both Clinton and Trump supporters. Unfortunately, prior beliefs and political commitments then start to creep in. When they received an article that included a fact-check, Clinton supporters were more likely to see the article as fair and balanced, while Trump supporters were more likely to see it as less accurate and less fair. In fact, it did not appear to matter whether the article included an attack on the fact-check’s source: the mere presence of a fact-check made Trump supporters more likely to perceive the article as biased. Worse, even though Trump supporters were more likely to believe the truth after seeing the fact-check, they still managed to see the source of truth as biased and re-confirm their support for Trump. “In other words, factual corrections can achieve the limited objective of creating a more informed citizenry but struggle to change citizens’ minds about whom to support.” ^iv

In 2020, Barrera et al. found similar results in their study of extreme right-wing candidate in the 2017 French presidential elections, Marine Le Pen.^v In a study designed similarly to Nyhan et al.’s, Barrera et al. exposed participants to a neutral article, an article with a fact about immigration, an article with a false claim about immigration (made by Marine Le Pen), or an article with a false claim plus a fact-check. Here again they found that participants updated their factual knowledge after seeing a fact-check (that is, they switched from a false belief about immigration to a true belief about immigration). Interestingly, participants who knew the truth before seeing a false claim were not persuaded by the claim. Participants who did not know the truth before seeing the false claim moved further away from the truth. Put another way, false claims appeared to have little influence on people who are already know the truth about an issue, but they make ill-informed people even more ill-informed.

Further, like Nyhan et al., they found that awareness of false claims did not change support for Marine Le Pen. In fact, supporters were more likely to support Le Pen and her position on immigration even after accepting that she had lied about people’s reasons for immigrating and the demographic make-up of immigrants. The authors speculate that this might be a result of exposure to true and false statements increasing the relevance of the issue, or its salience, to the audience. So, Le Pen supporters, who are already opposed to immigration, are likely to become more skeptical simply by thinking about immigration, regardless of whether they are well-informed or poorly informed about the issue.

Although a departure from the specific question of false claims and fact-checks, an earlier study about how people interpret policy proposals is also worth considering. In this study, Kahan et al. investigated whether exposure to scientific evidence about Outpatient Commitment Laws (OCLs) was likely to make people more or less likely to support OCL proposals. OCLs are laws that allow courts to mandate a person with a mental illness to undergo treatment while living in the community, with the threat of further evaluation and/or institutionalization if they refuse to comply with treatment orders. At the time of the study, the evidence supporting the effectiveness of OCLs was conflicted and advocates for people with mental illnesses were divided on whether the restrictions on freedom that OCLs impose are well justified.

What was particularly interesting about this study was not whether people updated their beliefs after exposure to evidence; it was that how people interpreted the evidence itself depended on prior beliefs and worldviews. In essence, people who are already likely to view the world through a hierarchical and communitarian lens are more likely to interpret the evidence as supporting OCLs. People who are more egalitarian and individualistic are more likely to read the same evidence as showing that OCLs are ineffective and therefore a poor justification for restricting people’s liberties.^vi This takes the findings of Barrera et al. and Nyhan et al.—that people can update their beliefs about political issues without changing their behaviour in relation to these issues—and amplifies it. Not only do people put distance between their political opinions and facts, they interpret the facts themselves according to their pre-existing belief structures.

By now, you’re probably getting a clear picture of the role of fact-checking. On the one hand, progressive and truth-loving organizations cannot allow false claims to go unchallenged. On the other hand, they need to recognize the limited impact that fact-checking has. It increases the audience’s likelihood of holding a true belief, but it does not necessarily influence their support for particular policies or politicians.

What does this mean in terms of political strategy?
This research points toward a strategy based on building up credibility. As Kahan et al. note, people interpret political and policy discussions through a ‘cultural credibility heuristic’: “Because individuals often lack the time and expertise to evaluate competing forms of empirical data, they rely on those whose judgment they trust to tell them what claims to accept. The people they trust, it turns out, are ones who share their defining group commitments.” ^vii In practical terms, the notion of cultural credibility means becoming both a reliable source of information and analysis and doing so in a way that is relevant to the target audience.

As Brown argues, the way experts talk about evidence often ignores the lived experience and the context of either the general public or a specific target audience.^viii This can leave experts looking condescending and the audience feeling shamed or stigmatized. It creates a divide between ‘knowers’ and ‘followers’ rather than fostering the kind of shared cultural commitments Kahan et al. suggest people rely on to interpret facts and evidence. As Brown puts it, “The danger of accepting a post-truth characterisation is that we abandon this empowering side of the evidence movement just as it’s winning through. Evidence and expertise have too often looked like counsel to the knowing, rather than what we could be making them: the means by which the less powerful can call the world to account.”

Combatting the forces of post-truth, then, means progressive groups need to collaboratively build up both their own credibility as evidence-based participants in public debates. At the same time, the presentation of evidence needs to connect facts to the lived experience of the intended audience. If we are sliding into a post-truth world—or if we’re already there—then we can’t rely on the power of truth itself to overcome that drift. It will take long-term investment in building capacity, credibility, and community.

ⁱ Daniel Dale, Daniel Dale: President Trump is a serial liar (CNN Politics, 2020).
ⁱⁱ Soroush Vosoughi, Deb Roy, and Sinan Aral, “The spread of true and false news online,” Science 359, no. 6380 (2018): 1
ⁱⁱⁱ Vosoughi, Roy, and Aral 2.
^iv Brendan Nyhan et al., “Taking Fact-Checks Literally But Not Seriously? The Effects of Journalistic Fact-Checking on Factual Beliefs and Candidate Favorability,” Political Behavior (2019/01/21 2019): 4-5.
^v Oscar Barrera et al., “Facts, alternative facts, and fact checking in times of post-truth politics,” Journal of Public Economics 182 (2020/02/01/ 2020).
^vi Dan M. Kahan et al., “Cultural cognition and public policy: The case of outpatient commitment laws,” Law and Human Behavior 34, no. 2 (2010).
^vii Kahan et al. 136.
^viii Tracey Brown, “Evidence, expertise, and facts in a “post-truth” society,” BMJ 355 (2016): 1.