The Intensity of Social Media Messages and Right-Wing Collective Action

Recent surges in right-wing activity have raised concerns about potential causal links to extremist social media platforms. This study examines the relationship between extremist online content and offline right-wing mobilization using discursive opportunity structure and social movement theories. We hypothesize that selection processes driven by intense online content motivate individuals primed for collective action. Using a large language model, we analyze video transcripts from social media platform Parler, measuring message intensity across four domains: hostility to outgroups, claims of conservative injustice, negative emotions, and violence. We link these to county-level right-wing event data from the Armed Conflict Location and Event Data (ACLED) project. Event-history analysis using proportional hazard models estimated the effects on right-wing activity rates in U.S. counties, controlling for spatial autocorrelation and other demographic factors. The results support our conjecture that counties with more intense postings across all four domains would have significantly higher rates of right-wing activity.