Bipartisan Consensus Under Attack From Engagement-Driven Algorithms

Opposition to bipartisan cooperation has become a defining feature of divisive social media content, according to research analyzing what makes posts polarizing. Scientists found that content attacking political compromise and bipartisan consensus significantly contributes to increasing animosity between Democrats and Republicans, suggesting that algorithms amplifying such content undermine democratic functioning.
The study examined X posts during the 2024 presidential election, identifying several categories of divisive content. Alongside posts supporting undemocratic practices and partisan violence, researchers flagged content opposing bipartisan consensus as a key driver of polarization. When such posts were slightly amplified in users’ feeds, political animosity increased substantially.
This finding illuminates how social media may be changing democratic culture. Historically, bipartisan compromise was seen as a democratic virtue—opposing parties finding common ground to address shared problems. But engagement-optimized algorithms tend to amplify content that portrays compromise as betrayal, rewarding ideological purity over pragmatic problem-solving.
Over 1,000 participants experienced feeds manipulated to show more or less content attacking bipartisan cooperation. Those exposed to more anti-compromise content showed increased polarization equivalent to three years of natural societal change, all within one week. This suggests that algorithmic amplification of anti-compromise messaging may be systematically eroding democratic norms.
The implications extend beyond individual attitudes to institutional functioning. When citizens view any cooperation with political opponents as illegitimate, elected officials face intense pressure to refuse compromise. This can produce legislative gridlock and governmental dysfunction, undermining democracy’s ability to address collective problems. Algorithms that amplify anti-compromise content may thus be indirectly weakening democratic institutions themselves.