Skip to main content Skip to footer
← Back to Community Violence Intervention
Card 05

Does It Work?

Three tiers of evidence address the question: randomized controlled trials, longitudinal evaluations, and city-level outcome data. What follows presents each tier with its documented strengths and documented limitations.

The Gold Standard: Randomized Controlled Trials

Three RCTs have evaluated CVI or CVI-adjacent programs.

The University of Chicago Crime Lab’s 2024 READI evaluation enrolled 2,456 participants in a randomized design. The results, published in the Quarterly Journal of Economics: a 65% reduction in shooting and homicide arrests among participants. The outreach-referred subgroup — those identified through street intelligence rather than criminal justice referrals — showed a 79% reduction in arrests and a 43% reduction in victimization. The effect on gunshot victimization, a 44% decrease, was sustained at 18 months. An important statistical note: the study’s pre-specified primary outcome (a combined index of three violence measures) did not reach conventional statistical significance, and the 65% shooting arrest reduction carries a p-value of 0.13 after multiple-testing adjustment. An independent review at EvidenceBasedPolicy.org characterizes the violence effects as “suggestive under established scientific standards” rather than definitive. The cost-benefit ratio (4:1 to 18:1) is statistically significant (p=.03). A 40-month follow-up is forthcoming.

The 2024 Denver GRID evaluation enrolled 143 participants in a randomized design. Per the published NIJ findings by David Pyrooz of CU Boulder, participants were “nearly 70 percent less likely to perpetrate violence” compared to the control group.

The University of Chicago Crime Lab’s Choose to Change (C2C) evaluation enrolled 2,074 youth and found approximately a 50% reduction in violent-crime arrests at two years, with effects persisting through 36 months. C2C targets a younger population than READI and uses cognitive behavioral therapy delivered in schools rather than comprehensive wraparound services.

Longitudinal Evidence: Baltimore Safe Streets

Johns Hopkins University tracked Baltimore’s Safe Streets program from 2007 through 2022. Daniel Webster, then director of the Johns Hopkins Center for Gun Violence Solutions, concluded that there is “clearly less gun violence” when programs like Safe Streets are properly implemented and adequately funded. The finding: 22% fewer homicides than predicted in Safe Streets program areas, and 23% fewer nonfatal shootings.

The Baltimore data also produced a cost-benefit estimate: $7.20 to $19.20 in economic benefits for every dollar invested, per Johns Hopkins. The returns come from prevented emergency department visits, avoided trauma surgeries, reduced criminal justice costs, and maintained economic productivity.

Two caveats from the methodology: First, the 15-year timeframe includes periods of both strong and weak program implementation — Baltimore’s Safe Streets has experienced site closures, management problems, and funding disruptions. The 22% finding averages across varying levels of fidelity. Second, the methodology is observational, not experimental — Johns Hopkins used statistical modeling to predict what homicide levels would have been without the program. This is rigorous but not as definitive as an RCT.

City-Level Outcomes

Multiple cities have reported violence reductions in areas where CVI programs operate. These are primarily program-reported or city-reported data — they have not been independently verified by outside researchers and should be read with that caveat.

Detroit’s Force Detroit program, per the City of Detroit and Detroit Free Press reporting, documented a 72% drop in homicides and shootings in targeted neighborhoods from November 2023 to January 2024, compared to the prior year. The citywide decline during the same period was 37%.

Lansing, Michigan saw a 52% decline in fatal shootings in a March 2025 Michigan State University evaluation, per The Trace. One caveat the evaluation’s own authors flagged: the decline was greater in non-program areas than in program areas, raising questions about specific program attribution versus citywide or national trends.

Orlando’s Advance Peace program was evaluated by UC-Berkeley researchers in 2024, who found a 20% reduction in firearm homicides and a 36% reduction in nonfatal shootings in the first year. By the third year, program neighborhoods showed an 88% reduction in gun homicides and a 71% reduction in shooting victims, per the Florida Rights Restoration Coalition. Orlando recorded just 10 total homicides in 2025, its lowest since 1971. The first-year evaluation estimated $8.3 to $8.9 million in taxpayer savings.

A March 2025 NYC Comptroller’s report found the Crisis Management System produced a 21% shooting reduction across multiple precincts, per the Comptroller’s office. An independent NYC Council Data Team analysis found a consistent 17% reduction.

Northwestern University’s Center for Neighborhood Engaged Research and Science (CORNERS) found that Chicago’s Communities Partnering 4 Peace prevented at least 383 homicides and shootings between 2017 and 2021. Peacekeeper hotspots experienced a 41% overall reduction in victimizations from 2023 to 2024, per a Northwestern University CORNERS evaluation published in April 2025.

The growing evidence base also includes replication results. Orlando’s Advance Peace model produced results in its UC-Berkeley evaluation, and the same model in Lansing produced a 52% decline in fatal shootings in Michigan State University’s evaluation — two-city results from the same model, each with independent university evaluations.

Philadelphia reported in November 2023 a 26% homicide reduction since 2021, during a period of significant CVI investment, per the City of Philadelphia.

In 2023, Dallas police reported that what had been the city’s number-one violent crime hot spot for decades had dropped off the list, per WFAA, following a violence reduction strategy that included CVI partnerships.

A February 2024 White House estimate credited CVI programs nationally with contributing to the 12.4% national homicide decline, though this is a modeled attribution across a broader strategy that included policing and gun legislation.

Hospital-Based Outcomes (HVIP)

Cleveland’s University Hospitals tracked more than 600 participants over five years through 2025 and documented reinjury rates dropping from 29% to 19%, per Ideastream Public Media. Indianapolis’s Prescription for Hope program at Eskenazi Hospital found, in a 2018 long-term evaluation, a 3% repeat injury rate among participants compared to 8.7% without the program — what evaluator Thomas Stuckey described as a “two-thirds reduction.” A 2008 Johns Hopkins RCT focused on youth aged 10 to 15 presenting with peer assault injuries found participants were three times less likely to be arrested for a violent crime.

Atlanta’s Grady Hospital program, led by trauma surgeon Dr. Randi Smith, reported in January 2024 a 1% reinjury rate in its first year with 600 consultations and a 98% acceptance rate among eligible patients, per 11Alive. By August 2025, the program had expanded to nearly 1,000 participants and the reinjury rate had risen to approximately 3% in years two and three — still dramatically below the national average of 30-40%, per Atlanta News First. Minneapolis’s Next Step program documented reinjury dropping from a historical baseline of 41% to 3% after one year.

A 2025 St. Louis evaluation of the Life Outside of Violence HVIP program, published in the Annals of Internal Medicine, found no significant difference in reinjury between program participants and a comparison group — the most prominent null result in the CVI evidence base.

Cost-Benefit Evidence

Each shooting generates emergency response costs exceeding $100,000 when police investigation, ambulance, emergency surgery, ICU stay, and follow-up are combined, per multiple economic analyses. Prosecution and incarceration cost $30,000 to $60,000 per year per person. Medical costs for a single gunshot injury range from $30,000 to $100,000.

Against those costs, the documented returns: Baltimore’s Safe Streets returned $7.20 to $19.20 per dollar invested, per the Johns Hopkins evaluation covering operations from 2007 through 2022. Orlando’s Advance Peace saved taxpayers $8.3 to $8.9 million, per UC-Berkeley’s 2024 evaluation. The 2024 READI RCT produced $182,000 to $916,000 in social savings per participant, representing a 4-to-1 to 18-to-1 return on investment (this ratio is statistically significant at p=.03).

The comparison baseline for all cost-benefit estimates is the status quo: the cost of emergency response, hospitalization, criminal justice processing, incarceration, and productivity losses associated with violence.

What the Evidence Does Not Show

Five evidence gaps matter for decision-makers.

No multi-site randomized trial exists. The three RCTs evaluated specific programs in specific cities. Whether READI’s results generalize to a different CVI program in a different city with different implementation quality is an assumption, not a finding.

Most programs have not been independently evaluated. The field has strong evidence from a small number of programs and limited evidence from the majority, per a February 2026 systematic review of Cure Violence evaluations published in INQUIRY, which found that 68.7% of 83 findings showed violence reductions but only 32.5% reached statistical significance.

The evidence is concentrated in large cities. Chicago, Baltimore, Detroit, Orlando, New York, and Philadelphia are not representative of every jurisdiction that might consider CVI. Whether the model produces equivalent results in rural communities, smaller cities, or politically conservative jurisdictions is largely untested.

Long-term outcomes beyond two years are documented only in Baltimore (Johns Hopkins, 2007-2022) and READI (44% gunshot victimization decrease sustained at 18 months). Whether violence reductions observed in other cities persist, grow, or decay over time is not documented.

The mechanism has not been independently tested. The READI evaluation tested the combined package — outreach, employment, cognitive behavioral therapy, case management — not the individual components. Whether the credible messenger relationship, the employment, or the CBT drives the results is an open question.

Not every evaluation has produced positive results. The Life Outside of Violence HVIP program in St. Louis found no significant difference in reinjury, a reminder that program design, implementation quality, and local context affect whether the model produces its intended effects.

The Invisible Success Problem

CVI workers describe a measurement challenge unique to prevention. As one violence interrupter put it: “We squashed it. But nobody knows about those kinds of stories.” A shooting that was prevented leaves no trace in the data. The mediations and interventions that produce population-level shooting reductions are, by definition, events that did not occur and therefore do not appear in any outcome dataset.

Bottom Line

Three randomized controlled trials anchor the CVI evidence base: READI (65% reduction in shooting arrests among 2,456 participants, though the pre-specified primary outcome did not reach conventional statistical significance), Denver GRID (70% reduction among 143 participants), and Choose to Change (50% reduction among 2,074 youth). Johns Hopkins tracked Baltimore Safe Streets from 2007 through 2022 and found 22% fewer homicides than predicted. City-level data from nearly a dozen jurisdictions shows consistent patterns of shooting reductions, with Orlando’s three-year results (88% gun homicide reduction) providing the most dramatic trajectory. Cost-benefit estimates range from 4-to-1 to 18-to-1 returns. The evidence gaps are real: no multi-site RCT, most programs not independently evaluated, evidence concentrated in large cities, long-term data limited, and one prominent null result (Life Outside of Violence, St. Louis). The honest summary: the model has produced impressive results when well-implemented in the cities that have been studied. Whether those results translate to every jurisdiction that might adopt CVI is the question the evidence cannot yet fully answer.

Source Appendix

  1. READI RCT — 2,456 participants, 65% reduction, p=.13 primary outcome, 4:1-18:1 cost-benefit (p=.03). Heller et al., QJE 139(1), 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC10898100/ Also: EvidenceBasedPolicy.org review. https://www.evidencebasedpolicy.org/study-reviews-1/readi-chicago

  2. Denver GRID — 143 participants, ~70% less likely. Pyrooz, D.C., NIJ, 2024. https://nij.ojp.gov/library/publications/multidisciplinary-teams-street-outreach-and-gang-intervention-mixed-methods

  3. Choose to Change — 2,074 youth, ~50% reduction at 2 years. University of Chicago Crime Lab. https://crimelab.uchicago.edu/

  4. Johns Hopkins 15-year evaluation — 22% fewer homicides, $7.20-$19.20 ROI. Daniel Webster: “clearly less gun violence.” Johns Hopkins Bloomberg School of Public Health. https://www.thebanner.com/community/criminal-justice/safe-streets-penn-north-JLZH6TGFJFAYHCIIMMTT4J4OD4/

  5. Northwestern CORNERS — CP4P prevented 383 victimizations 2017-2021; 44% gunshot victimization decrease sustained 18 months. Northwestern University CORNERS. Also: SC2 initiative. https://www.scalecvichicago.org/

  6. Peacekeeper — 41% reduction 2023-2024. Northwestern University CORNERS (Andrew Papachristos), April 2025. https://news.wttw.com/2025/04/17/chicago-has-seen-significant-gun-violence-declines-under-peacekeepers-program-new-study Also: Report PDF: https://cdn.prod.website-files.com/634dd45091db1de63b7112d9/67ffdcbe728fe830be5886c0_Peacekeeper%20Eval%20Report_FINAL.pdf

  7. Detroit Force Detroit — 72% targeted, 37% citywide, November 2023-January 2024. City of Detroit. https://detroitmi.gov/ Also: Detroit Free Press.

  8. Orlando Advance Peace — Year 1: 20%/36%, $8.3-$8.9M (UC-Berkeley 2024). Year 3: 88%/71% (FRRC 2026). 10 homicides in 2025. ClickOrlando, August 2024. https://www.clickorlando.com/news/local/2024/08/01/gun-violence-reduced-in-orlando-neighborhoods-due-to-intervention-program-heres-how/ Also: Florida Rights Restoration Coalition, February 2026. https://floridarrc.com/cvi/

  9. Lansing — 52% decline, March 2025 MSU evaluation. Authors’ caveat on non-program areas. The Trace, May 2025. https://www.thetrace.org/2025/05/advance-peace-lansing-michigan-shootings/ Also: Michigan Advance, June 2025. https://michiganadvance.com/2025/06/03/shootings-dropped-in-lansing-but-untangling-why-is-complicated/

  10. NYC Crisis Management System — 21% (Comptroller March 2025), 17% (Council Data Team). NYC Comptroller, March 2025. https://comptroller.nyc.gov/reports/the-cure-for-crisis/ Also: NYC Council Data Team. https://council.nyc.gov/data/cure/

  11. Philadelphia — 26% since 2021, reported November 2023. City of Philadelphia. https://www.phila.gov/2023-11-29-kenney-administration-progress-report-our-violence-prevention-efforts/

  12. Dallas — hot spot removed, 2023. WFAA, August 2023. https://www.wfaa.com/article/news/crime/one-of-the-top-crime-hot-spots-in-dallas-has-dropped-off-the-list/287-bffcac71-d9f6-4722-9f13-4565880031b6

  13. Federal 12.4% estimate. White House Fact Sheet, February 2024. https://bidenwhitehouse.archives.gov/briefing-room/statements-releases/2024/02/06/fact-sheet-the-biden-%E2%81%A0harris-administration-advances-equity-and-opportunity-for-black-americans-and-communities-across-the-country-2/

  14. Cleveland — 600+ participants, 29%→19%. Ideastream, February 2025. https://www.ideastream.org/health/2025-02-11/university-hospitals-program-reduces-repeat-violence-among-young-gunshot-victims

  15. Indianapolis — 3% vs. 8.7%, 2018. Bell et al., Journal of Trauma, 2018. https://pmc.ncbi.nlm.nih.gov/articles/PMC5739956/

  16. Johns Hopkins youth RCT — 2008, ages 10-15. Cheng et al., Pediatrics 122(5), November 2008. Also: AAMC. https://www.aamc.org/news/can-hospitals-help-reduce-violence

  17. Grady Hospital — 1% first year (January 2024), ~3% years 2-3, ~1,000 participants by August 2025. 11Alive, January 2024. https://www.11alive.com/article/news/local/grady-hospital-atlanta-ivyy-program-tackling-gun-violence-one-year-progress/85-6546fc09-ea58-469f-a4e8-0725878c65d3 Also: Atlanta News First, August 2025. Also: Everytown $100K grant, 2025.

  18. Minneapolis Next Step — 41%→3%. CBS Minnesota (WCCO), “Next Step Program Aims To Help Gun Violence Victims Avoid Retaliation,” November 17, 2020. https://www.cbsnews.com/minnesota/news/next-step-program-aims-to-help-gun-violence-victims-avoid-retaliation/ Also: PBS Frontline, 2021.

  19. Life Outside of Violence — null result, St. Louis. Annals of Internal Medicine, 2025.

  20. Cure Violence systematic review — 68.7% showed reductions, 32.5% significant. Ransford, Williams, Slutkin, INQUIRY, February 2026. https://journals.sagepub.com/doi/10.1177/00469580251366142

  21. Cost of violence — >$100K per shooting, $30-60K prosecution/year, $30-100K per gunshot. Multiple economic analyses. See Q11 for detailed cost-benefit.

  22. Violence interrupter — “We squashed it.” ProPublica, MacGillis, May 2023. https://www.propublica.org/article/are-community-violence-interruption-programs-effective

©2026 Safer Cities
Privacy Policy