Interim evaluation of Baltimore's Safe Streets initiative: Effects on gun violence

The Safe Streets program was designed and implemented in selected high-crime neighborhoods of Baltimore in an attempt to replicate Chicago's CeaseFire program. The program is designed to prevent gun violence through outreach to the highest-risk youth, mediation of conflicts, referrals to services, and community mobilization. We analyzed monthly trends in murders and nonfatal shootings by police post with data obtained from Baltimore police for posts in the top quartile in shootings during a 4-year baseline period. For our interim evaluation, only 1 of 4 program neighborhoods had sufficient post-program data to derive reliable estimates of effect. We used negative binomial regression models to control for baseline means for murders and nonfatal shootings, arrests for illegal weapons, arrests for drugs (lagged), and calendar month. Safe Streets was associated with a significant reduction in homicides (p<.001) in the first neighborhood implementing the program. There were no homicides in the neighborhood for the first 19 months of implementation. Using model estimates, we estimate that, with no program, there would have been 5 homicides during this time. Nonfatal shootings declined in the target neighborhood by 7%; however, shootings declined more (-23%) in comparison areas. We will present up-to-date evaluation data for a total of 3 intervention neighborhoods and discuss possible explanations for disparate findings for homicides and nonfatal shootings.