As political tensions continue to divide Americans, new research sheds light on a factor that may ease the discomfort of engaging across partisan lines. A series of three studies published in Social Psychological and Personality Science found that liberals and conservatives were more open to befriending someone from the other side when that person had a politically diverse group of friends. Participants expected to be viewed more positively, felt more interested in learning about their partner, and were less likely to dehumanize them—suggesting that diverse social circles can improve how people anticipate and experience political conversations.
The United States has seen a deepening political divide in recent years. Surveys suggest that very few Americans have many friends from the opposing political party, and online echo chambers and residential segregation only reinforce this pattern. Prior studies have shown that cross-group friendships—particularly across racial lines—can reduce intergroup tension and improve perceptions. But research had not yet examined whether this idea applies in political contexts.
The new research was inspired by earlier findings that people expect to be treated more fairly and with less bias when their interaction partner has a racially diverse set of friends. The authors of the current study, Kyle Anderson, Grace Flores-Robles, and Daryl Wout, sought to test whether similar patterns could be observed in political interactions. Specifically, they wanted to know whether being exposed to someone with politically diverse friends might reduce anxiety about being judged based on political identity—and, in turn, improve willingness to interact with them.
“Dr. Daryl Wout, my advisor and mentor over the past few years, invited me to pursue this topic,” said Anderson, who recently received a PhD from The Graduate Center, CUNY. “His work has primarily looked at the impact of friendship diversity on anticipated interactions. I have an evolving interest in intergroup relations, so exploring interpolitical friendship networks was a good fit for both of us.”
The researchers conducted three separate online experiments with a total of 696 U.S. participants, recruited through the online research platform Prolific. Participants self-identified as either liberal or conservative and were matched with a hypothetical interaction partner from the opposing political group. In each study, participants were given a brief profile of this person, which included details about their top friends’ political affiliations.
In the first two studies, White liberal (Study 1a) and White conservative (Study 1b) participants were shown one of three profiles: a political outgroup member with all same-party friends (homogeneous network), a mix of liberal and conservative friends (diverse network), or no political information about their friends. Participants were asked to imagine having a political conversation with the person and then rated how they thought the partner would perceive them (meta-perceptions) and how interested they were in becoming friends with that person.
In the third study (Study 2), the researchers expanded the sample to include people of multiple racial backgrounds and focused only on the diverse versus homogeneous friendship conditions. This time, participants were told to anticipate a conversation about a political topic and were asked to choose from several hot-button issues like abortion or gun control. They also responded to new measures, including their interest in learning from the conversation, how open they thought their partner would be, and how human they perceived the partner to be.
Across all three studies, participants consistently reacted more positively to outgroup members who had politically diverse friendships. In Studies 1a and 1b, both liberals and conservatives expected to be viewed more warmly and were more open to befriending someone from the other party if that person had a mix of liberal and conservative friends. These effects were stronger than in the homogeneous or no-information conditions.
For instance, White liberal participants who imagined a conservative with diverse friends had more positive expectations about how the person would view them, and showed greater willingness to befriend them. Conservatives in Study 1b showed a similar pattern when evaluating a liberal target.
“Political diversity in people’s friendship networks (having both liberal and conservative friends), as compared to only having friends from their political group, helps make people with different identities across the political divide feel more psychologically safe, and eases relations between liberals and conservatives in the United States,” Anderson told PsyPost.
Study 2 extended the findings to a more diverse sample of liberals and conservatives. This time, participants anticipated a political discussion and were told which specific hot-button issues might come up, such as abortion or voter fraud. Again, those who saw a political outgroup member with a diverse group of friends felt more positively about the interaction. They were more likely to expect to be seen in a favorable light, more willing to engage in the conversation, and more likely to believe the partner would be open to learning from the exchange. They also rated the outgroup member as less dehumanizing and immature.
These effects were not influenced by the participant’s own political identity—liberals and conservatives responded similarly. The effects of friendship diversity were largely explained by changes in meta-perceptions, or beliefs about how the outgroup member would see them. When people thought they would be perceived as warm and competent, they were more open to conversation and friendship.
The study also found that improved meta-perceptions helped explain why participants reacted more positively. Feeling that they would be seen as competent or friendly helped participants feel safer and more open to engaging with someone from the other side.
“Given our present political climate, the effectiveness of friendship diversity in creating identity safety for both liberals and conservatives was surprising,” Anderson said. “The heightened political rhetoric surrounding the 2024 U.S. presidential election could have negated any benefits of safety cues, such as friendship diversity. The fact that friendship diversity served to improve anticipated interactions across political lines speaks to the importance of creating interaction contexts that minimize people’s concerns about how individuals across the political divide would perceive and treat them.”
While the findings suggest that friendship diversity can improve interpolitical relations, the researchers acknowledge some limitations. The interactions were hypothetical or anticipated—no real conversations took place. Future studies could involve actual political discussions to better assess how these impressions play out in real time.
Another limitation is the demographic makeup of the samples. The first two studies only included White participants, and while the third was more racially diverse, the generalizability of the findings to other racial groups remains uncertain. The researchers noted that political identity might intersect with racial and historical experiences of marginalization, which could influence how safe or threatened someone feels during cross-political interactions.
“I want to explore this and related work across different political and racial identities,” Anderson said. “I am especially interested in using expressions of gratitude to create psychological safety. I also want to explore whether there is an additive benefit to people expressing gratitude and people having diverse friends in friendship networks.
“If people want to read a copy of this paper, they can access the accepted version on my website, kylemanderson.com, under the CV tab.”
The study, “How Liberal-Conservative Friendship Diversity Could Improve Interpolitical Relations in the United States,” was published online June 24, 2025.
