"EBK": Using Crowd-Sourced Social Media Data to Quantify How Hyperlocal Gang Affiliations Shape Networks and Violence in Chicago

Abstract Purpose Test recent gang theory derived from ethnographic research which suggests that gang dynamics in Chicago's South Side have evolved, with decentralized micro-gang "set" factions and cross-gang interpersonal networks marking the contemporary landscape. Importantly, this is theorized to have produced "EBK" dynamics, where between-faction conflict is irrelevant to driving the emergence of violence. Methods A parts-of-speech based natural language processing strategy is used to analyze text from Reddit pages about Chicago to create a social network dataset of 271 individuals across 11 gang sets. Using datasets collected and shared by Reddit users, we identify the geographic boundaries associated with each set, which are validated against official police data. Louvain community detection is utilized to identify sub-communities within the network, and hierarchical linear models are used to evaluate whether gang affiliations or network positionality are more salient in explaining mortality risk. Results Overall, results provide quantitative evidence for EBK hypotheses. Community detection analyses suggest that gang-affiliated individuals in the South Side often connect with gang-affiliated peers from several gang sets, particularly those operating in spaces nearby their own gang. Hierarchical logistic regression revealed that individuals with ties to homicide victims and central positions in the overall gang network were at increased risk of victimization, regardless of gang affiliation. Conclusions This research demonstrates that utilizing crowd-sourced online information can enable the study of otherwise inaccessible topics and populations, including gangs. Future research should continue to investigate the strengths and limitations of using crowd-sourced information about gangs and crime to conduct scientific research.