Gang intervention programs have promised to solve America’s violence problem for decades. Politicians love announcing them. Communities desperately want them to work. But when researchers actually measure what happens after the press conferences end, the results tell a story that contradicts much of what the public has been sold.
Some gang intervention programs produce dramatic results: 63% fewer youth homicides, 70% fewer arrests among participants, violence drops that persist for years. Others do nothing. A disturbing few actually increase criminal behavior among the young people they were designed to help.[s]
The difference is not about funding levels or political will. It comes down to a fundamental question: does the program target actual drivers of violence, or does it just look like it should work?
The Boston Experiment
In 1996, Boston faced a youth homicide crisis. The city launched Operation Ceasefire, a gang intervention program built on a counterintuitive premise: instead of trying to arrest their way out of the problem, authorities would deliver a direct message to gang members. Continued violence would trigger overwhelming legal consequences for the entire group. At the same time, social services would be offered to anyone who wanted out.
The results became known as the “Boston Miracle.” Youth homicides dropped 63%.[s] Citywide gun assaults fell 25%. Shots-fired calls dropped 32%. These were not modest improvements. They represented a fundamental shift in how effective gang intervention programs could be when properly designed.
The approach worked because it targeted the right people. Research shows that in most cities, roughly half a percent of the population is responsible for up to 70% of homicides and gun violence.[s] Reaching that small group with credible threats and genuine offers of help proved far more effective than broad enforcement sweeps.
What Actually Reduces Violence
Since Boston, the focused deterrenceA strategy of preventing hostile actions by threatening credible retaliation that would impose unacceptable costs on an adversary. model has been replicated in dozens of cities. The typical result: a 35 to 60 percent reduction in community-wide homicides.[s]
New Orleans implemented the strategy in 2012. Monthly homicides dropped 18.6%. Gang member-involved homicides fell 30.1%. When researchers compared New Orleans to 14 similar cities that did not adopt the program, the difference was statistically significant.[s]
A separate approach, the Cure Violence model, treats violence as a contagious disease. Former gang members work as “violence interruptersFormer gang members or community members trained to identify and mediate conflicts before they escalate to violence, serving as credible messengers in gang intervention programs.,” mediating conflicts before they escalate. A systematic review of 27 program sites found that 68.7% showed reductions in shootings or killings. Outside of Baltimore, where implementation faced significant challenges, 95.8% of sites demonstrated violence reductions.[s]
Hospital-based gang intervention programs catch victims at a critical moment. Oakland’s Youth ALIVE! program, launched in 1994, found that participants were 70% less likely to be arrested and 60% less likely to have any criminal involvement compared to non-participants. Remarkably, 98% of participants were not rehospitalized for violence-related injuries.[s]
What Sounds Good But Fails
For every program that works, several do not. The pattern is consistent: approaches that seem intuitively effective often fail when subjected to rigorous evaluation.
D.A.R.E., the Drug Abuse Resistance Education program that once operated in three-quarters of American school districts, became a symbol of this problem. Study after study showed it had little to no impact on drug use. One study found use actually increased among participants.[s]
Scared Straight programs, which bring at-risk youth into prisons for “reality checks,” won Emmy and Oscar awards for a 1979 documentary. They also produced more criminals. Participants were more likely to end up behind bars than similar youth who never went through the program. The Justice Department eventually warned states they could lose federal funding if they continued using an approach the evidence proved harmful.
Heavy-handed suppression tactics, the most common response to gang activity, fare poorly as well. A Justice Policy Institute review found “no evidence that gang enforcement strategies have achieved meaningful reductions in violence.” Worse, aggressive suppression can increase gang cohesion by creating an external enemy, while alienating communities whose cooperation is essential for public safety.[s]
Why Implementation Matters
Even proven gang intervention programs can fail if poorly implemented. The Cure Violence model showed mixed results in Baltimore while succeeding elsewhere. The difference was not the model itself but the context: insufficient funding, interagency conflict, and departure from core program elements undermined effectiveness.
The G.R.E.A.T. program, a school-based curriculum delivered by law enforcement officers, shows how mixed results require careful interpretation. Evaluations found it reduced gang membership by 39% at one-year follow-up and improved attitudes toward police. But it did not reduce general delinquency.[s] A program that achieves some goals while missing others is not a failure, but neither is it a complete solution.
The Cost of Getting It Wrong
Ineffective gang intervention programs do not just waste money. They waste lives and erode public trust in solutions that might actually work.
Cities that invested in evidence-based approaches, including Boston, Chicago, Indianapolis, New Orleans, Oakland, and Stockton, experienced declines of more than 30% in shootings resulting in injuries.[s] For every dollar invested in Cure Violence, cities saved up to $18 in reduced medical and criminal justice costs.
The evidence is clear enough that the question is no longer whether effective approaches exist. It is whether communities will choose proven methods over politically convenient ones.
Gang intervention programs represent one of the most extensively studied areas of criminal justice policy. The evidence base now includes multiple systematic reviews, meta-analyses, and rigorous experimental evaluations. What emerges is a clear hierarchy: some approaches produce consistent, statistically significant reductions in violence, while others show null effects or, in documented cases, iatrogenicCausing harm or adverse effects through medical treatment or intervention, especially when the treatment itself creates problems it was meant to solve. harm.
Understanding why requires examining both the interventions themselves and the methodological rigor applied to their evaluation.
Focused DeterrenceA strategy of preventing hostile actions by threatening credible retaliation that would impose unacceptable costs on an adversary.: The Evidence Base
Operation Ceasefire launched in Boston in 1996 as a problem-oriented policingA policing strategy that identifies and analyzes specific crime problems to develop tailored responses, rather than relying on standard reactive enforcement methods. strategy targeting gang violence. The evaluation by Braga and colleagues used a one-group time-series design analyzing monthly data from January 1991 through May 1998.
Key findings: youth homicide victims decreased from an average of 3.5 per month to 1.3 per month, a 63% reduction that was statistically significant. Citywide gun assaults dropped 25%. Shots-fired calls for service fell 32%. All reductions achieved statistical significance after controlling for employment rates, youth population changes, citywide violent crime trends, and drug market activity.[s]
The program earned an “Effective” rating from CrimeSolutions.gov based on multiple studies, meaning implementation is likely to produce intended outcomes.
Replications in other cities demonstrated the approach was transferable. The National Network for Safe Communities reports typical impacts of 35 to 60 percent reductions in community-wide homicides.[s] Specific results: Indianapolis saw a 34% reduction in homicide; Stockton achieved 42% reduction in gun homicide; Cincinnati reduced group member-involved homicide by 41%.[s]
New Orleans: A Quasi-Experimental Test
The New Orleans Group Violence Reduction Strategy evaluation by Corsaro and Engel employed a two-phase quasi-experimental design comparing the city to 14 urban areas with similarly high historical homicide rates.
Results from interrupted time-series analysis: overall monthly homicides decreased 18.6%; firearm-related homicides fell 17.4%; firearm assaults dropped 16.2%; gang member-involved homicides declined 30.1%. All reductions were statistically significant.[s]
Critically, the researchers controlled for the simultaneous implementation of Cure Violence in Central City to rule out confounding. The GVRS effects remained significant, and New Orleans showed significantly greater homicide rate declines compared to the 14 comparison cities.
Cure Violence: Systematic Review Findings
A 2025 systematic review published in Inquiry examined 13 papers analyzing 27 Cure Violence program sites, yielding 83 findings on shootings or killings. The review followed PRISMA guidelines and excluded programs that deviated from core model elements.[s]
Overall, 68.7% of findings indicated reductions in shootings or killings, with 32.5% achieving statistical significance. Geographic variation was substantial: Baltimore accounted for a disproportionate share of null or negative results. Outside Baltimore, 95.8% of sites demonstrated violence reductions, with 54.2% reaching statistical significance.
Notable effect sizesStatistical measure that quantifies the magnitude of difference between groups, indicating whether research findings are practically meaningful beyond statistical significance.: Chicago showed 52% reduction in killings; New York City achieved 63% reduction in shootings; Cali, Colombia, recorded 74% reduction in killings.
The review identified key implementation factors: targeting highest-risk individuals, employing credible messengers, real-time conflict mediation, sustained case management, structured monitoring with rapid response, and sufficient sustained funding.
Hospital-Based Gang Intervention Programs
Hospital-based violence intervention programs (HVIPs) represent the oldest and most researched CVI model, originating in Oakland in 1994. The theoretical basis: traumatic injury creates a “teachable moment” when victims may be receptive to intervention.
A 2004 evaluation of Oakland’s Youth ALIVE! Caught in the Crossfire program used a control groupIn research, the group of participants that does not receive the treatment being tested, used for comparison with the treatment group. design. Participants were 70% less likely to be arrested and 60% less likely to have criminal involvement. The program saved hospitals $750,000 to $1.5 million annually by reducing rehospitalization.[s]
Baltimore’s Violence Intervention Program evaluation (2006) found 5% rehospitalization among participants versus 36% among non-participants. Participants were half as likely to be convicted of any crime and four times less likely to be convicted of violent crime. Estimated savings: $600,000 in healthcare costs and $1.25 million in incarceration costs.
San Francisco’s Wraparound Project reported a 400% decrease in reinjury rates over its first six years. Giffords Law Center analysis determined prevention of 3.5 injuries per year renders the program cost-neutral.
Ineffective and Harmful Gang Intervention Programs
D.A.R.E. underwent multiple rigorous evaluations, including randomized controlled trialsA research study where participants are randomly assigned to receive either the treatment being tested or a control condition.. Meta-analyses consistently found little to no effect on drug use. At least one study documented increased use among participants. The program’s persistence despite this evidence illustrates how political appeal can override empirical findings.[s]
Scared Straight programs produced the strongest evidence of iatrogenic effects. Multiple evaluations found participants exhibited higher rates of subsequent offending than control groups. The mechanism appears to be counter-normative: exposure to prison environments may have increased rather than decreased identification with criminal lifestyles. The Justice Department classified these programs as harmful.
Gang suppression tactics alone, the most widely employed strategy, show consistently poor evidence. A Justice Policy Institute review stated: “Our review of the research found no evidence that gang enforcement strategies have achieved meaningful reductions in violence.”[s] Heavy-handed approaches can strengthen gang cohesion through external threat, damage police-community relationships essential for cooperation, and trap youth in criminal justice systems through practices that target former members long after active participation ends.
G.R.E.A.T.: Mixed Results Require Nuance
The Gang Resistance Education and Training program underwent a National Institute of Justice-funded evaluation following more than 3,800 students in seven cities through six waves of data collection.
One-year post-treatment results: G.R.E.A.T. students showed 39% lower odds of gang joining, greater resistance to peer pressure, less positive attitudes about gangs, and improved attitudes toward police. However, the program did not reduce general delinquency.[s]
This represents partial success: gang intervention programs may achieve specific targeted outcomes without producing broader behavioral changes. Policy implications depend on whether reduced gang membership alone justifies program costs.
Economic Analysis
Cost-effectiveness data support investment in evidence-based gang intervention programs. The Vera Institute reports that for every dollar invested in Cure Violence, cities save up to $18 in reduced medical and criminal legal system costs.[s] Sacramento’s Advance Peace program saved between $18 and $41 per dollar spent across emergency response, healthcare, law enforcement, and corrections.
These figures do not capture broader social costs of violence: lost productivity, community trauma, reduced property values, business disinvestment. Full economic analyses would likely show even greater returns.
Methodological Considerations
Evidence quality varies across gang intervention programs. Randomized controlled trials remain rare due to ethical and practical constraints. Most evaluations use quasi-experimental designs with comparison groups or interrupted time-series analyses. Selection bias remains a concern in HVIP studies, as participants who accept intervention may differ systematically from those who decline.
Despite these limitations, the convergence of evidence across multiple methodologies, cities, and research teams supports confidence in the core finding: targeted, relationship-based interventions outperform suppression-only approaches.
