The question the researchers set out to answer has its roots in one of the oldest and most robust findings in social psychology. Since the classic experiments of Donn Byrne in the 1960s, researchers have known that similarity breeds attraction. When people discover that a stranger shares their attitudes, they report more positive feelings toward that person, and the more attitudes they share, the stronger the attraction becomes. This similarity-attraction effect holds even when the shared attitudes are trivial, and it shapes who we befriend, who we sit next to, and even how long we are willing to spend in someone’s company. Decades of follow-up work have extended the effect to online friendships, exercise groups, and friendship maintenance over time.

Music preference turned out to be a particularly potent form of similarity. Previous studies showed that people feel warmer toward fans of the music they themselves enjoy, that shared musical taste predicts friendship formation both in person and on social media, and that even at the group level, learning that another group shares your musical preferences reduces intergroup bias. People also stereotype music fans systematically: fans of one’s own preferred genres are credited with positive traits, while fans of disliked genres are assigned negative ones. Music, in other words, is not just entertainment; it is a dense social signal.

That signaling power is exactly what made some researchers suspicious. If music taste reliably predicts personality, ethnicity, drug use, and even political orientation, then perhaps liking the same bands makes us like someone only because we infer they are like us in deeper ways. Earlier work by Boer and colleagues in 2011 suggested precisely this: shared values appeared to mediate the link between music preference similarity and social attraction. And the political correlations are well documented. Indie, alternative, and hip-hop lean liberal, while classical, country, pop, and religious music lean conservative. If music is just a proxy for politics, then a shared love of indie rock should matter only when it also signals shared politics.

The Goldsmiths team, led by Olivia Jewell with Lauren Stewart, Keon West, and Jacques Launay, designed an experiment capable of pulling these two threads apart. They recruited 255 British participants through the online platform Prolific, all of whom identified as either politically liberal or conservative and who had strong, opposing preferences for two genres identified in pilot testing: indie music, stereotypically aligned with the left, and country music, stereotypically aligned with the right. A power analysis indicated that 212 participants would be sufficient, and the final sample, aged 18 to 45 with a mean age of 31.6, exceeded that threshold. Participants were paid £1.40 for taking part.

The experimental design was elegantly simple. Each participant was shown a profile of a fictional target person named Kate, who was either a fan of indie or country music and either liberal or conservative. By matching participants to conditions, the researchers orthogonally manipulated musical similarity and political similarity, creating four groups: participants who shared both music and politics with Kate, shared only music, shared only politics, or shared neither. Because musical and political similarity varied independently, the design could distinguish a genuine independent effect of music from a mere political signaling effect. Participants then rated their social attraction toward Kate, a measure encompassing liking, positive feelings, and a sense of potential friendship.

The results were strikingly clear. A two-way analysis of variance revealed a significant main effect of political orientation similarity: when participants shared Kate’s politics, their social attraction scores were 0.52 points higher than when they did not, a substantial effect with a partial eta-squared of 0.21. But there was also a significant main effect of music preference similarity: sharing Kate’s musical taste raised attraction scores by 0.32 points, with a partial eta-squared of 0.09. Crucially, there was no interaction between the two factors. Music similarity boosted liking to the same degree whether or not it came bundled with political agreement. If music were merely a stand-in for politics, that interaction should have appeared, and music alone should have had no effect. It did not.

This outcome directly contradicts the mediation account offered by Boer and colleagues, who had argued that value orientations were the missing link between shared music and social bonds. The new findings suggest that values, at least as expressed through political orientation, are not the whole story. The authors caution that the relationship may simply be more nuanced than either study captured: perhaps values and political orientation are less tightly coupled than assumed, or perhaps music evokes values in ways it does not evoke political labels. Either way, the data indicate that shared music taste carries its own social weight, a conclusion consistent with the broader similarity-attraction literature showing that each additional point of similarity adds to attraction.

The practical implications may be the most exciting part of the study. The Contact Hypothesis holds that interaction between members of different groups can reduce prejudice, and previous research has shown that music exposure, from protest songs to unity anthems to cross-cultural music education programs, can soften intergroup hostility. But getting highly prejudiced individuals to engage with unfamiliar music or unfamiliar people is notoriously difficult. The new findings suggest an easier route: rather than exposing people to novel music from other cultures, organizers could use shared musical preferences as the common ground itself. Music festivals, where food, games, and performances cluster around a beloved genre, may offer an accessible setting in which people who differ politically, and might otherwise avoid contact, naturally come together around something they already love.

The study is not without limitations. The researchers used a single target stimulus, Kate, who represented only one gender, so gender could not be analyzed as a factor in attraction. Only two genres and two political orientations were tested, which limits how well the findings capture the messier, more pluralistic reality of most people’s tastes and views. Future work, the authors suggest, should control for variables such as gender, socioeconomic status, personality, age, and education, and could add other similarity dimensions, such as sports team allegiance or hobbies, to determine whether music is uniquely powerful or simply one instance of a general similarity-attraction mechanism. Even so, the core message stands: shared music taste is not just a shadow of shared values. It is a genuine, independent force in social attraction, one that might help divided communities find their common rhythm.

Subject of Research: The independent effects of music preference similarity and political orientation similarity on social attraction

Article Title: Music preference and political orientation have similar (but independent) effects on social attraction

Article References: Jewell, O., Stewart, L., West, K., & Launay, J. (2026). Music preference and political orientation have similar (but independent) effects on social attraction. Frontiers in Cognition, 5, Article 1838689. https://doi.org/10.3389/fcogn.2026.1838689

Image Credits: AI Generated

DOI: 10.3389/fcogn.2026.1838689

Keywords: music preference, social attraction, political orientation, similarity-attraction, social psychology, intergroup contact, prejudice reduction, experimental psychology, music and identity, friendship formation, values signaling, Frontiers in Cognition