Analyzing network exposure and news media sharing among hundreds of thousands of US Twitter users in the summer of 2022, the authors document an ideological asymmetry in news exposure and sharing: liberal media were shared across party lines, while conservative outlets were overlooked by liberals and circulated mostly within a tight core of strongly conservative accounts. The phenomenon is striking because it points to an apparent paradox: although conservative users primarily engage with their own media, liberal outlets attract a broader audience, including many conservatives.
One explanation would point to differences in homophily—the tendency to form ties with ideologically similar others. This behavioral explanation suggests that conservatives would need to actively choose broader networks to maintain greater exposure to cross-cutting content. The authors find no support for it. The key insight, by contrast, is structural. Twitter before the Musk acquisition was a predominantly liberal platform, which placed conservative users in the minority. As the sociologist Peter Blau famously argued, minority members inevitably form a larger share of outgroup ties simply due to the asymmetry in opportunity structure, even when ingroup preferences are equal across groups. This means that being in the majority, liberal users can afford to be narrow networkers, while conservatives are compelled to form more outgroup ties that expose them to cross-cutting content.
Much like minority students who form more outgroup friendships in schools, or individuals of the minority gender who are drawn into more outgroup interactions in workplaces, the observed asymmetry in media exposure is driven by structural constraints inherent to skewed group sizes. By combining their digital trace data with computer simulations that artificially vary homophily levels, the authors show that partisan differences in networking behavior are not needed to reproduce the observed asymmetry in exposure to news content. It is not that users of different ideologies form ties in starkly different ways, but rather Twitter's ecosystem at the time that explains why liberal news spread broadly and conservative news narrowly.
The authors note that their analysis of 2022 Twitter captures a particular moment in time. Yet the structural mechanism is general and extends beyond a particular platform: wherever an ideological group is in the minority, it can rarely insulate itself fully and thus becomes disproportionately exposed to outgroup content. As a result, a surprising number of minority-group members follow and engage with majority content. As user compositions shift across emerging platforms, exposure asymmetries follow in predictable ways.
