Want to find some new conspiracy theories? Social media is a great place to find them. But will it make you believe them?

That’s the question asked by a new working paper looking at conspiracy theories surrounding the Donald Trump assassination attempt on July 13. (Gotta love it when academia moves quickly!) It suggests that, while social media is a productive source of conspiracy discovery, you’re more likely to believe theories you hear from friends and families.

The paper is by Katherine Ognyanova (Rutgers), James N. Druckman (Rochester), Jonathan Schulman (Penn), Matthew A. Baum (Harvard), Roy H. Perlis (Harvard), and David Lazer (Northeastern). Here’s the abstract:

Conspiracy beliefs can lead to maladaptive and, in rare cases, even violent behaviors. Focusing on conspiracies about the assassination attempt on former President Trump, this report takes the rare step of differentiating exposure to conspiracy theories from belief in them.

It finds that there is considerable exposure through social media. Yet, reliance on social media does not correlate with belief in conspiracy theories. Rather, interpersonal relationships play a larger role. Conspiratorial thinking and political motivations also significantly relate to holding a conspiracy belief.

The results suggest that corrective interventions would face substantial communicative and psychological hurdles.

So what conspiracy theories did you come across in the wake of the assassination attempt at a Trump rally in Bethel, Pennsylvania? Did you hear that it was a Democratic plot to take out the Republican nominee — ordered by Joe Biden himself? Or maybe that it was staged by Republicans to generate sympathy and a swell of support for their candidate?¹

The authors ran a survey in the days following the attempt, from July 17 to 21, asking 2,765 people if they had heard of the two main left-leaning and right-leaning conspiracy theories and whether they believed them. (“Only” 93% of those surveyed said they had heard about Trump being shot. The other 7%…must have interesting media diets.)

Within the sample, 38% said they had heard the theory that Democrats operatives were behind the assassination attempt, while 50% said they’d heard the attempt was staged and not real. And similar percentages of people said they believed the two theories: 12% said it was “very likely” a Democratic plot, while 11% said it was “very likely” staged.

(If you include people who said each conspiracy theory was only “somewhat likely” rather than “very likely,” those numbers grow to 30% and 29%, respectively.)

Interestingly (for me, at least), the usual divides in American politics aren’t as profound here as they often are. People with and without college degrees — an increasingly important source of polarization² — heard and believed the conspiracy theories at about the same rate. It’s clear that a lot of Americans heard these conspiracy theories in a relatively short period of time.

These descriptive results contrast with reports by Budak et al.³ and other⁴ scholars⁵ who argue that relatively few Americans – mainly highly motivated partisans – are exposed to most false information online. We find quite the opposite: more than half of Americans rapidly learned about one or both of these conspiracy stories, primarily through social media and interpersonal connections. And, nontrivial percentages of those who were familiar with the stories believed them.

This suggests there is an important distinction between “most” false stories of the sort emphasized by Budak et al. and the most high-profile, salient false stories, like the “stop the steal” movement or, in this case, the attempted assassination of a presidential candidate.

Ognyanova et al. ran the numbers to see what personal characteristics correlated with both hearing and believing either conspiracy theory. Surprising no one, being a Republican and a Trump fan made you more likely to hear that Democrats were behind it all, and people with low approval of Trump’s presidency were more likely to hear the it-was-staged theory. (Partisans love having their priors affirmed, after all.) Having a broadly conspiratorial mindset also correlated with hearing about both.⁶ People who relied on social media for their news were also more likely to hear both conspiracy theories; no other source of news had the same effect.

When it comes to conspiracy belief, partisanship and conspiratorial thinking were again factors. But when it came to where people get their news, relying on social media didn’t correlate with believing conspiracy theories — only with hearing them. But you know what source of news did correlate with conspiracy belief? Friends and family.

Republicans more likely to believe the right-leaning conspiracy and Democrats more likely to believe the left-leaning one. Conspiracy thinking again mattered in the expected direction.

Importantly, hearing about a conspiracy theory from social media was not significantly related to believing it. The only source consistently linked to higher conspiracy beliefs was friends and family members.

In short, partisan leaning and conspiratorial thinking relate to both exposure and belief, but reliance on social media only connects to exposure. This accentuates the limitation of looking only at digital trace data.

In other words, you can come across a lot of crazy ideas on Facebook or Twitter — but hearing them from your best friend seems to make them more likely to take up residence in your brain. The combination of conspiratorial thinking, partisanship, and real-life social connections is a potent one.

While social media were a key vector for exposure to conspiracies, finding those theories plausible was associated with hearing about them from social ties. This result aligns with classic ideas about the persuasive role of mass and interpersonal communication…⁷⁸

While it may well be the case that most false information does not reach the majority of Americans, there are important exceptions. The most high-profile false stories can quickly reach large portions of the public and be believed by a nontrivial percentage…

Finally, the results show why altering conspiracy beliefs is difficult: they tend to be held by people inclined to such thinking, motivated to hold a given belief, and who received information via social ties.

1. It was the reptilians. We all know it was the reptilians. [↩]
2. For more on this, check out the new book (out Sept. 5) from Matt Grossman and David A. Hopkins, Polarized by Degrees: How the Diploma Divide and the Culture War Transformed American Politics. [↩]
3. “Misunderstanding the harms of online misinformation,” by Ceren Budak, Brendan Nyhan, David M. Rothschild, Emily Thorson, and Duncan J. Watts. [↩]
4. “Fake news on Twitter during the 2016 U.S. presidential election,” by Nir Grinberg, Kenneth Joseph, Lisa Friedland, Briony Swire-Thompson, and David Lazer. [↩]
5. “Exposure to untrustworthy websites in the 2016 US election,” by Andrew M. Guess, Brendan Nyhan, and Jason Reifler. [↩]
6. This was measured using the American Conspiracy Thinking Scale, created by Joseph E. Uscinski and Joseph M. Parent. It asks subjects to agree or disagree with four statements: Much of our lives are being controlled by plots hatched in secret places. Even though we live in a democracy, a few people will always run things anyway. The people who really “run” the country are not known to the voters. Big events like wars, recessions, and the outcomes of elections, are controlled by small
   groups of people who are working in secret against the rest of us. [↩]
7. “Through the Grapevine: Informational Consequences of Interpersonal Political Communication,” by Taylor N. Carlson. [↩]
8. “The Incidental Pundit: Who Talks Politics with Whom, and Why?” by William Minozzi, Hyunjin Song, David Lazer, Michael A. Neblo, and Katherine Ognyanova. [↩]