Like It or Not, Misinformation is (Still) a Thing

Like It or Not, Misinformation is (Still) a Thing

By Alex Worsnip, Director of the Applied Epistemology Project

In the fallout from last week’s US election, I’ve seen numerous commentators and friends on social media counsel against the narrative that Trump’s victory is due (at least in part) to misinformation. This narrative, they argue, is patronizing to the increasing numbers of working-class voters who have been gradually abandoning the Democrats for the Trump-era Republicans. These voters have well-founded economic anxieties, they argue, and the Harris campaign failed to offer them policies that would address those anxieties.

In assessing these claims, we should begin by sharply distinguishing the question of whether it is politically advantageous to stress the role of misinformation in the current political environment from the question of whether it is true that misinformation is playing such a role. It’s plausible enough that people do not like being told that that they have been duped. And perhaps political campaigns—which reach far more people than some academic somewhere on a blog or social media site does—should not tell them that. As students of the political environment, though, we should seek to understand it accurately whether or not the conclusions we reach line up with advantageous political messaging. Moreover, in the long run, a mistaken picture of the political environment is likely to leave us worse-off in our attempts to achieve our political goals.

Now, it is extremely difficult for anyone—let alone a philosopher like me, who is not a trained empirical researcher—to assess causal claims about why exactly Harris lost. Data about voters’ (self-reported) beliefs and attitudes is widely available, but it’s very hard to know which of those beliefs and attitudes actually account for the way they ultimately voted. We can’t just conduct the election hundreds of times in slightly different circumstances to statistically analyze which of those circumstances make a difference. So I don’t want to claim that misinformation was the cause of Harris’s defeat, or even that it was a leading cause. What is clear, though, is that factual misinformation continues to be utterly rife in the United States, and that denying its presence and trying to recast voters as only differing in their values and priorities is simply wishful thinking.

Let’s consider just a few major issues. First, the 2020 election, and whether it was “stolen”. As of last year, 30% of Americans, and 68% of Republicans, thought that Joe Biden “only won it due to voter fraud.” Second, the COVID pandemic. Again as of last year, only 63% of Americans thought that getting the COVID-19 vaccine is safer than getting COVID itself, while 26% thought that ivermectin is an effective treatment for COVID, with these numbers trending down and up respectively. Third, climate change. As of 2022, 24% of Americans, and 44% of Republicans, believed that human activity is contributing “not too much or not at all” to climate change; while only 46% of Americans, and 17% of Republicans, believed that it is contributing “a great deal.” In a more recent survey, only 23% of Republicans saw climate change as a major threat.

On all three of these issues, then, false beliefs are rife, and perhaps increasing. Moreover, these false beliefs concern politically important issues. If the Democrats really did cheat their way to the Presidency in 2020, subverting the democratic process, that would clearly be a very strong reason not to vote for them. Likewise, if climate change isn’t happening, isn’t affected by human behavior, or isn’t a threat at all, then the Democrats’ policies designed to address it would be sacrificing quality of life and current economic growth for no gain. And on the pandemic, it is Democrats who have been more supportive of lockdowns and (limited) vaccine mandates. If the pandemic were a hoax or the vaccine were unsafe, these would have been unnecessary at best and have put lives at risk at worst.

Even when we turn to the economy—the factor that those playing down misinformation tend to stress—we find a significant amount of false beliefs specifically among Republican voters, for example about the way inflation has been trending over the last year or about current stock market performance. Moreover, the perception that Trump’s economic policies, more than those of the Democrats, are aimed at making working-class Americans better off is itself at best extremely contestable. Trump’s 2017 tax cuts, for example, went overwhelmingly to the very rich (and even more disproportionately to the mega-rich), and barely decreased the tax burden on those in the bottom half of incomes at all. Meanwhile, as the journalist Rachel Cohen has recently discussed, Biden administration policies like the expanded Child Tax Credit, which put money into the pockets of the families of 65 million children and, by some estimates, kept 3.7 million children out of poverty, seem to have had minimal effect on the perception that Democrats are not interested in helping working people.

All of this matters for debates about what Democrats can do to win back support from the constituencies they are losing. Debates are already raging about what adjustments the Democrats should make to their policy platform in 2026 and 2028—with many commentators, predictably, claiming that Democrats will win if they move in the direction of their own preferred policy platform. But policy platform adjustments in any direction can only make a difference if voters find out about them, and more broadly if they have accurate beliefs about what the different candidates are selling. The depressing truth may be that, quite generally, the details of policy platforms have less capacity to shift electoral dynamics than any wing of the Democratic party would hope.

More broadly, as many others have pointed out, our contemporary epistemic environment is toxic for (small-d) democratic politics because it makes so many voters utterly unpersuadable. For example, the philosopher M. Giulia Napolitano has argued that what is distinctive of epistemically problematic conspiracy theories is the “self-insulated” way that their proponents believe them. A conspiracy-theoretic belief is self-insulated when there is no counterevidence that can shake it, because any counterevidence is explained away as planted by the conspirators or, on closer inspection, allegedly irrelevant to the conspiracy hypothesis. Many of those who hold the aforementioned false beliefs—that the 2020 election was stolen, that the COVID-19 vaccine is more dangerous than COVID itself, that climate change is a hoax—hold them in exactly this “self-insulated” way.

Indeed, we saw a poignant example of this immediately following the 2024 election results. If Democrats really had stolen the 2020 election, one would have expected them to try to do so again in 2024. So the fact that Kamala Harris (and every other prominent Democrat) immediately accepted the election result and conceded defeat seems like pretty good evidence that they didn’t win by rigging the election last time. At the very least, it’s evidence that they weren’t cheating this time, as many Trump supporters predicted they were or would. A couple of provocative political commentators—one from the left, one from the right—asked Trump supporters on X (a.k.a. Twitter) whether the fact that Harris conceded had changed their mind in any way. The replies are a truly dispiriting read. In short, virtually none of them seem to have changed their minds at all. Instead, they endorse a range of ad hoc explanations of Harris’s concession to allow them to hang on to their conviction that, in every other case, Democrats are trying to cheat and steal their way to victory.

How do you persuade such people to change their minds about politics? More acutely: how do you persuade them when you’re exactly the sort of person—a Democratic politician, or an academic at an elite institution—that they most deeply distrust? I don’t know. This is a deep problem, and it doesn’t admit of any easy solution. But to have any hope of tackling the problem, we have to acknowledge its reality. It is true that, like any paradigm for analyzing the political world, misinformation discourse can (as the philosopher Dan Williams has repeatedly argued) be taken too far. We should be wary of labeling just any claim that we disagree with as misinformation. And we should also be wary of labeling just anything that disrupts the initial consensus on a novel, fast-moving, politically-inflected scientific topic as misinformation. For example, it was arguably a mistake to confidently call the lab-leak hypothesis about the origin of COVID “misinformation”: as time has gone on, a reasonable number of experts have come to endorse this hypothesis or at least see it as a serious possibility, and the U.S. Department of Energy (under a Democratic administration) has come to see it as the most likely explanation. Whether they are ultimately right or not, it probably isn’t appropriate to call a view about an unsettled topic like this ‘misinformation’.

But some things—such as claims about widespread fraud in the 2020 US election, or the claim that human activity is not contributing toward climate change—simply are not supported by the available evidence, and I do not think we need hesitate about calling the spread of such views ‘misinformation’. We may wish that dissatisfaction with the current (soon-to-be former) administration could be fully chalked up to accurate perceptions of genuine policy failures, along with different priorities and values. And no doubt these play a role. But to leave misinformation out of the picture entirely is to refuse to face reality. Misinformation is still here, and it’s not going away any time soon.