Quantifying the spread of racist content on fringe social media: A case study of Parler

Few studies characterize the diffusion of racist content on fringe social media platforms. We demonstrate how racism spread on Parler, a far right, un(der)-moderated social media platform, and that a single comment to a racist post increases the likelihood a person will generate and propagate new racist content. We found that racism on Parler was a social “contagion.” Using 50,375 posts from 2018 to 2021 that contained racist remarks, we quantified the spread of racism from the posts-to-comments (micro) and user-to-user (macro) levels. Comments on racist posts were 21 times more likely to be racist than comments on non-racist posts. On an average, Parler users posted 166% more racist content after an engagement with a racist post. At the posts-to-comments level, the spread of racist sentiment is alarming within ethnic subgroups (e.g. anti-Jewish-specific comments were 191 times more likely to appear on “anti-Jewish” posts; anti-Black-specific comments were 227 times more likely to appear on “anti-Black” posts). The spread of racist rhetoric between subgroups is also significant, albeit lower than within subgroups, suggesting a spillover effect. Aggregate user posting patterns also suggest that the spread of racist rhetoric between subgroups is lower, but significant. Our findings therefore suggest interventions should target users actively engaging with racist content. Furthermore, our study provides evidence that Parler served as a gateway platform that radicalized users toward racist content production, underscoring the urgency of intervention. Given the drawbacks of traditional moderation strategies, we propose additional interventions, policies, and structural changes for social media platforms to mitigate racism online.