Political diversity in your social circle might come with a surprising trade-off

A new study published in Political Research Quarterly</em> shows that Americans who report more political diversity in their social networks tend to trust political information on social media more than others. But this heightened trust applies to both true and false content, raising concerns about the unintended consequences of network diversity in the digital age.

Social media has transformed how people access political news, but the shift from traditional media has come with new challenges. Platforms like Facebook and Twitter often present a mix of verified facts and misinformation, and many users struggle to tell the difference. Against this backdrop, researcher Paul B. Platzman set out to explore how the political diversity of a person’s social network shapes their trust in political content they encounter online.

Platzman’s interest stemmed from previous work suggesting that people’s social environments—especially exposure to differing opinions—can shape political knowledge, tolerance, and behavior. While past research has examined echo chambers and polarization, little was known about whether having politically diverse friends and acquaintances affects how people judge the truth of what they see on social media.

“Prior to this study, I had been interested in how the composition of Americans’ social networks influences their political attitudes, beliefs, and behaviors,” explained Platzman, who earned a PhD from Columbia University. “Although scholars had already demonstrated an array of important relationships between individuals’ network attributes and measurable political characteristics, no one had yet explored the relationship between an individual’s political network diversity (i.e., how evenly both parties or political ideologies are represented within an individual’s personal network) and the likelihood that an individual accepts at face value the political content they encounter online.”

“It seemed to me that existing theories often cited in the social networking literature, as well as broader and related disciplines within sociology and social psychology, suggested a number of causal mechanisms that might relate the two phenomena. So, I set out to determine whether a relationship between political network diversity and belief in the accuracy of online political content exists, and this paper is the result of that effort.”

Platzman conducted a multi-phase investigation. First, he analyzed nationally representative data from the 2020 American National Election Studies (ANES) Social Media Study. More than 3,000 participants were asked about the political makeup of their personal and Facebook friend networks, and how much they trusted political content they saw on Facebook and Twitter. Based on their answers, Platzman calculated a “network diversity” score, ranging from zero (no diversity) to one (equal mix of Democrats and Republicans).

He then examined whether higher diversity scores were associated with greater trust in social media content. The results showed that people with more politically diverse networks expressed more trust in political information they saw on Facebook and Twitter. This association held even when controlling for other factors like partisanship, ideology, and demographics. Importantly, the relationship was evident both across different individuals and within individuals over time.

However, these findings left an important question unanswered: Is this increased trust justified? To investigate further, Platzman designed a follow-up survey experiment. He recruited 550 U.S. adults to participate in a study where they were shown mock Facebook posts that included political headlines—some true, some false—along with images and layout features mimicking the look of real social media. Participants were asked to rate the accuracy of each headline and indicate whether they would share it.

Before seeing the headlines, participants answered questions about their political discussion networks—specifically, how often they talked about politics and how much disagreement or diversity they perceived in those conversations. Once again, people who reported more political diversity in their networks were more likely to believe the headlines were accurate, regardless of whether the information was true or false. They were also more likely to say they would share the content.

While those in politically diverse networks were better at recognizing accurate headlines, they were equally likely to believe and share false ones. In other words, their increased trust did not come with increased discernment. Rather than helping them spot the truth, network diversity appeared to make people more credulous in general.

“Americans whose social networks contain political diversity are more likely to believe the political content they encounter online, whether the information is true or false,” Platzman told PsyPost.

To explore whether people with diverse networks might respond differently to interventions aimed at improving accuracy, Platzman conducted a larger follow-up study with over 2,000 participants. In this experiment, respondents again rated the accuracy and shareability of mock Facebook headlines, but this time, some headlines were accompanied by fact-checking labels, and some participants saw a warning message reminding them that false information exists online.

This setup allowed Platzman to test whether accuracy nudges had different effects depending on a person’s level of network diversity. Surprisingly, he found no evidence of such differences. Fact-checking labels and warnings did influence how participants judged content, but they worked the same way for people with low and high network diversity. In other words, network diversity raised baseline trust levels, but it didn’t change how people responded to interventions meant to improve judgment.

“Americans in politically diverse networks are just as willing as others to change their beliefs about political content following exposure to fact-checking labels that authoritatively declare content to be true or false,” Platzman said. “These findings expand our knowledge of the types of individuals prone to believing both factual and fictional content online and suggest conditions under which fact-checking labels are more or less likely to produce anticipated and second order effects.”

Platzman notes that this trade-off complicates efforts to fight misinformation. Trust in news is generally seen as a good thing, especially in an era of declining confidence in journalism and democratic institutions. But when trust increases across the board, including for content that is false, it can backfire. Efforts to build public resilience to misinformation must take these nuances into account.

“Because possessing a political diverse network is neither inherently helpful nor harmful in ascertaining fact from fiction in online content, media outlets and technology platforms intent on deploying fact-checking labels and other accuracy nudging interventions would be wise to optimize their delivery of these interventions to users who are likeliest to reduce belief in false information without also reducing belief in factual content or generating unwarranted trust in factually dubious content,” Platzman said. “Harnessing the benefits of network diversity requires careful management, and accuracy nudging interventions should be deployed selectively and in creative ways.”

Like any study, the research has some limitations. All measures of network diversity were based on self-reports, which might not accurately reflect the true makeup of people’s social environments. Additionally, while the survey experiments used mock social media content, they can’t fully capture how people behave in real-world online settings.

“Virtually all research analyzing survey responses about social networking features rely on self reports of individuals’ network characteristics, which may not represent their true natures,” Platzman explained. “This study is no exception. Future research would benefit from highly tailored data collection efforts that directly measure attributes of Americans’ social networks.”

Even so, the study sheds new light on a largely overlooked factor in how people interact with information online. Political diversity in one’s social circle is often seen as a remedy to polarization and bias, but it may also come with unintended risks. When it comes to online content, being surrounded by a range of opinions may not make people more discerning—it may simply make them more trusting.

“The size of the relationship between reported political network diversity and belief in the accuracy of online political content actually exceeded the effect sizes of fact-checking labels on Americans’ beliefs about the accuracy of online content,” Platzman noted. “This suggests that a relatively obscure and underappreciated aspect of social identity has more influence over individuals’ beliefs about online content accuracy than overt technical interventions specifically designed to influence these beliefs.”

The study, “Political Network Diversity and Trust in Online Political Content,” was published April 23, 2025.