Unflagged Factual Content More Damaging Than Flagged Misinformation, Study Finds

New research from the University of Pennsylvania and MIT challenges conventional wisdom about what drives vaccine hesitancy online, revealing that factually accurate but potentially misleading content has a far greater impact than outright fake news.

According to a groundbreaking study published in Science, factual but selectively presented information about COVID-19 vaccines that promotes skepticism was found to be 46 times more impactful in discouraging vaccination than content flagged as misinformation.

“Since the 2016 U.S. presidential election, many thousands of papers have been published about the dangers of false information propagating on social media,” explained Duncan J. Watts, Stevens University Professor in Computer and Information Science at Penn Engineering and a senior author of the study. “But what this literature has almost universally overlooked is the related danger of information that is merely biased.”

The research team, led by Jennifer Allen, a 2024 MIT Sloan School of Management PhD graduate, analyzed thousands of survey results and used artificial intelligence to estimate the impact of more than 13,000 individual Facebook posts during the height of the COVID-19 pandemic.

Their findings paint a concerning picture of how public health information spreads online. While deliberately false content about vaccines—such as claims about tracking microchips—can be persuasive when encountered, these posts represent only a tiny fraction of what users actually see on social media platforms.

Instead, the researchers found that factual but “vaccine-skeptical” content from mainstream sources had a much larger overall impact due to its widespread visibility. One striking example was a true story about a doctor who died shortly after receiving the COVID-19 vaccine—a statistical anomaly that nonetheless received tens of millions of views on Facebook.

“Vaccine-skeptical content that’s not being flagged by Facebook is potentially lowering users’ intentions to get vaccinated by 2.3 percentage points,” Allen noted. “A back-of-the-envelope estimate suggests that translates to approximately 3 million people who might have gotten vaccinated had they not seen this content.”

This revelation raises difficult questions for social media platforms and their content moderation policies. While flagging or removing demonstrably false information has become standard practice, addressing factually accurate but potentially misleading content presents a much more complex challenge that pits free expression against public health concerns.

“When you actually look at the stories people encounter in their day-to-day information diets, fake news is a minuscule percentage,” Allen explained. “What people are seeing is either no news at all or mainstream media.”

The impact of these findings extends beyond the pandemic. With the 2024 U.S. Presidential election approaching, the researchers point to similar dynamics at play in political information. Allen cited an example from the 2020 election about military ballots found in the trash that were mostly votes for Donald Trump—a factual story, but one whose headline failed to mention that it involved just nine ballots total.

The study highlights a fundamental tension in our information ecosystem: how should platforms handle content that is technically accurate but presented in ways that may mislead the public or promote potentially harmful behaviors?

Allen suggests that community-based solutions might offer a path forward. “Crowdsourcing fact-checking and moderation works surprisingly well,” she said. “That’s a potential, more democratic solution.”

The research team’s methodology represents a significant advancement in understanding social media’s impact on public opinion. “What makes our paper really unique is that it allows us to break open Facebook and actually understand what types of content are driving misinformed-ness,” Allen explained.

As vaccination rates for COVID-19 boosters remain low among U.S. adults, the study provides valuable insights into why factual but selectively presented information may be more effective at swaying public opinion than outright falsehoods—a finding with profound implications for how we approach information literacy and public health communication in the digital age.