Association Between School Closure and Neighborhood Firearm Violence: Evidence from Chicago and Implications for Prevention
Justin S Hatchimonji, Renee I Tristano, Diane N Haddad, Michael Poulson, Andrew J Benjamin, Danielle R Hatchimonji, Selwyn O RogersBackground:
Firearm violence remains a leading cause of death in the US, yet the neighborhood consequences of large-scale public school closures are poorly understood. We evaluated whether Chicago's 2013 closure of 49 public elementary schools was associated with subsequent firearm violence in surrounding communities.
Study Design:
Chicago Police Department shooting data (2010–2024) were geocoded using ArcGIS and linked with Census-derived neighborhood covariates. Census tracts with centroids within 0.5 miles of a closed school were classified as exposed. Shooting trends were compared within and outside the exposure radius, and difference-in-differences Poisson regression evaluated post-closure changes for all-age and youth victims, stratified by schools that reopened (n=32) versus those remaining vacant (n=17).
Results:
Among 47,112 shootings, neighborhoods outside the exposure radius experienced stable shooting rates (1.7 to 1.5 per 1,000 population, 2010–2024). Rates increased from 2.5 to 3.1 per 1,000 near schools that reopened and from 3.0 to 4.7 per 1,000 near schools that remained vacant. Post-closure increases were greater within exposed neighborhoods (reopened schools: IRR 1.17, p=0.009; persistently vacant schools: IRR 1.51, p<0.001). Among youth, associations were strongest near persistently vacant schools (IRR 2.00, p<0.001), whereas increases near reopened schools were not statistically significant (IRR 1.18, p=0.122).
Conclusions:
The 2013 mass closure of Chicago public elementary schools was associated with increased neighborhood firearm violence, particularly where school buildings remained vacant. These findings suggest that long-term school vacancy may contribute to community firearm violence and should be considered in education and urban policy decisions.