Social media companies face a fundamental tension between business performance and democratic health, according to research quantifying the relationship between divisive content, user engagement, and political polarization. The study found that algorithms designed to reduce political animosity might decrease some engagement metrics, creating financial pressures that discourage healthier platform design.
Over 1,000 X users participated in an experiment during the 2024 presidential election. Researchers manipulated feeds to show some participants more anti-democratic and partisan content while others saw less. They then measured both psychological effects on political attitudes and behavioral effects on platform engagement.
Results confirmed that divisive content drives certain forms of engagement. Users exposed to more controversial posts spent marginally more time on the platform and viewed more posts overall. These metrics matter immensely to advertising-based business models because they directly determine revenue. Content that provokes strong emotional reactions keeps users scrolling, creating more opportunities to display advertisements.
However, the research revealed important nuance. While down-ranking divisive content slightly reduced time spent and posts viewed, it actually increased meaningful interactions. Users liked and reposted content more frequently when feeds contained less inflammatory material, suggesting that engagement quality improved even as raw volume decreased marginally.
This finding challenges assumptions about inevitable trade-offs between business performance and social responsibility. Platforms might be able to reduce political polarization while maintaining sustainable business models if they accept different engagement patterns—fewer total impressions but higher-quality interactions. Whether corporate decision-makers will voluntarily pursue this path or whether external pressure through regulation or public accountability might be necessary remains uncertain.