Inside the architecture of a strategy reversing the trajectory of one of America’s most dangerous cities
At the end of 2021, Baltimore’s streets told a grim story — candlelight vigils in the cold, police tape in familiar corners, a weary city in its seventh straight year with more than 300 homicides. Alongside St. Louis, it was regularly cited as the most dangerous city in America, with a homicide rate 10 times higher than New York City’s, even amid its COVID-era surge in violence.
Three-and-a-half years later, Baltimore is making news again, this time for reducing violence. In 2024, Baltimore saw its lowest homicide toll in a decade. The first half of 2025 closed with the fewest killings ever recorded in the city.
Behind the decline is Baltimore’s most effective iteration of a cohesive strategy to prevent and reduce violence — a deliberate shift from reacting to shootings to actively managing the dynamics that drive the city’s most lethal violence, built on identifying the small number of people most likely to shoot or be shot in Baltimore, delivering them a clear message of support and accountability, backing those words with meaningful services and focusing law enforcement precisely on people who continue to commit violence. It’s the first time Baltimore has built systems with the staying power to reduce gun violence, not just endure it as a crisis. Known as the Group Violence Reduction Strategy (GVRS), the city’s integrated framework for managing serious violence is not just bending Baltimore’s violence curve — it has begun to reshape how the city manages and governs.
How we got here
Baltimore’s current strategy grew out of a pilot launched in early 2022 in the Western District, the city’s most violent district and one of the most dangerous in the nation, a part of the city that will be familiar to viewers of “The Wire.” By midyear, violence in the Western District declined faster than anywhere else — an early sign, amid persistently high citywide numbers, that something might be starting to shift.
Baltimore has tried versions of a group violence reduction strategy before: Operation Safe Neighborhoods in the late 1990s and Operation Ceasefire in 2014 (not to be confused with the later community-led Baltimore Ceasefire 365 movement). Built on the same theory, each showed early promise before ultimately unraveling amid missing components, fractured leadership and poor coordination. This time, on the third attempt, city leaders set out to avoid those pitfalls, working with a team from the University of Pennsylvania’s Crime and Justice Policy Lab (CJP) to design, launch and scale an integrated and sustainable approach.
To ensure an unbiased evaluation, CJP brought in two outside researchers — Max Kapustin of Cornell University and Aaron Chalfin of the University of Pennsylvania (both Vital City contributors) — who were unaffiliated with implementation or city funding. Their independent evaluation, completed in early 2024, confirmed that the Western District’s reductions were a direct product of the pilot. In the strategy’s first 18 months, homicides and nonfatal shootings fell by roughly one-third, with no rise in low-level arrests or signs that violence was simply being pushed to other parts of the city.
The city began expanding the new GVRS beyond the Western District in early 2023. Not long after, Baltimore began to see historic citywide declines. By the end of 2024, the city recorded 201 homicides — the first time the annual toll fell below 300 since 2014. The first half of 2025 saw just 68 killings, the city’s lowest six-month total ever. As of mid-August, homicides are down roughly 25% year-to-date, on pace for another record low.
Plenty of other cities have seen post-COVID declines, but Baltimore’s drop stands out for its scale and consistency. We can’t credit this strategy for all of it, but evidence strongly suggests that it drove the Western District reductions — and the broader citywide drop closely tracks with its expansion. From our front-row vantage, it’s clear something is working in Baltimore. It would be a mistake not to examine what’s driving the change — and what lessons other cities might draw from it.
The theory behind the results
The strategy in Baltimore builds on decades of practice and research often referred to as “focused deterrence” — a term that has been applied to so many different approaches that it’s lost much of its clarity. While the policing component relies on deterrence, the strategy also mobilizes community leadership, applies procedural justice in how city agencies engage with residents and partners closely with outreach workers and service providers.
Measuring the true impact of strategies like this one is challenging. Because they operate at a community or city level, randomized controlled trials (RCTs) are often infeasible, making causality hard to determine. A 2021 survey showed that crime researchers generally believe the strategy reduces gun violence, but the gap in rigorous research at that time left lingering doubts.
Some of that uncertainty may be out of date, as the evidence base around GVRS has grown considerably in the last few years. Recent studies — including several RCTs — show increasingly consistent evidence of impact. Evaluations in cities like Oakland and Stockton, for example, have demonstrated that this approach can produce substantial violence reductions.
Still, one of the most enduring reasons for skepticism is how rarely cities get implementation right — and sustain it. The theory only matters if a city can translate it into disciplined, day-in, day-out practice — something few have managed to do. The “Boston Miracle” of the mid-1990s was the first widely recognized example of an effective GVRS. But even Boston eventually stepped away, saw violence rise and later returned to the strategy. Baltimore’s previous attempt collapsed on multiple fronts: leaning too heavily on enforcement, failing to deliver services and, ultimately, damaging community trust.
The architecture of a turnaround
A key starting point we and others built into Baltimore’s approach: GVRS was never meant as a one-off program, but as a citywide systems change. The goal was to move away from reactive and fragmented government toward focused, coordinated action across partners working on violence intervention and reduction. Scaling from a single-district proof of concept to a citywide strategy forced the city to manage and govern differently — building durable systems, scaling capacity and earning legitimacy through sustained coordination and deep institutional change. The following elements define that shift — a suite of complementary and reinforcing changes to government operations.
Baltimore conducted a data- and intelligence-driven analysis to understand who was most directly affected by gun violence — as both victims and suspects. Across districts, the findings were consistent: fewer than 2% of residents were linked to up to 75% of shootings and homicides. Many were older men (over 30) with extensive criminal histories, often appearing as both victims and suspects. The violence didn’t stem so much from drug markets as from retaliatory disputes among criminally active groups. In short, a small number of high-risk individuals — shaped by shifting group dynamics — were at the center of harm, both experiencing and inflicting it. This shared understanding helped align city agencies, law enforcement and community partners around a single focus — the people at the center of the violence.
The city established a cycle of real-time management structures to identify those at very highest risk and coordinate responses as violence unfolds. These include the Baltimore Police Department’s weekly shooting reviews — law enforcement-only meetings where investigators, analysts and proactive enforcement units assess shootings, flag retaliation risks and identify those needing immediate attention. State and federal partners join to align investigations and prosecutions. Intelligence from those meetings flows into the Mayor’s Office for Neighborhood Safety and Engagement’s (MONSE) coordination meetings, where front-line workers from law enforcement, community violence intervention (CVI) — the city’s network of credible outreach workers and service providers — and community partners come together to share information, align on cases and develop action plans with the explicit goal of intervening to save lives. These meetings are carefully structured and decision-oriented, designed to translate intelligence into targeted action.
This intelligence-driven focus triggers a tightly organized intervention to deliver the strategy’s core message: the community wants the violence to stop; help is available; and if the violence continues, consequences will be swift and certain. Baltimore built the capacity for custom notifications — these are small, face-to-face meetings that take place shortly after someone surfaces as a likely shooter or victim. Unlike the large, infrequent “call-ins” other cities rely on, custom notifications make the message personal, immediate and hard to ignore. Each intervention team is a deliberate partnership of authority, legitimacy and help — typically a law enforcement and city representative, a service provider and a credible community “moral voice” such as residents directly affected by violence, members of the clergy and other trusted leaders who speak with authenticity and moral credibility.
The power of that message rests on the city’s investment in outreach and services. Baltimore built a new partnership with Youth Advocate Programs to engage older men (25 and up) while bolstering Roca’s long-standing work with high-risk young men and boys ages 16–24. These professional outreach teams work exclusively with the small number of people at highest risk, offering a range of supports: relocation, stipends, housing assistance, trauma counseling and intensive mentorship. This is CVI fully integrated with system partners — not a parallel effort, but the infrastructure that makes an effective violence reduction strategy possible. At the core of the strategy is trust: in the messenger, in the consistency of the offer and in the integrity of the system behind it. That trust takes time to build, and can be lost quickly.
These support structures set the stage for enforcement. When someone rejects the community’s message and continues to engage in violence, the strategy shifts to focused enforcement and prosecution. Police Commissioner Richard Worley, building on the groundwork laid by former Commissioner Michael Harrison, prioritizes violence-focused precision policing and partnership, rejecting outdated and harmful practices such as “zero-tolerance” and indiscriminate sweeps. That focus may sound simple, but it rests on deep reforms inside BPD, which is integrating intelligence, outreach, enforcement and investigations in ways it has never done before. When done this way, enforcement can actually make services more effective — it’s difficult to focus on change if you remain at immediate risk of violence. The fact that the evaluation of the strategy found no increase in arrests suggests that BPD was serious about restraint and precision in their operations.
Keeping the system honest
Many cities stumble because their systems are fragile, ad hoc or simply nonexistent. Baltimore confronted this by intentionally designing meetings to foster consistency, trust and focus among partners who had rarely shared the same table. Principal leaders in City Hall, BPD and the State’s Attorney’s Office all committed to meeting monthly to review performance, make decisions and keep the strategy on course — exhibiting strong leadership even amidst occasional public disagreements. City leadership has also exercised discipline. Mayor Brandon Scott made a rare political choice: resisting calls to expand too quickly, insisting, “We’re going to do this the right way.” In a city impatient for change, that patience was a gamble — and it’s paying off.
All of this raises a major question: Can other cities succeed at this type of violence reduction without the political leadership and institutional commitment seen in Baltimore? Both are essential — without them, a strategy rarely survives. But as more cities show results, we now have a clearer picture of the specific capacities needed to reduce violence. We also see more clearly how those capacities can be built. Some cities may not be ready now, but capacity can be developed — and with the right analyses, leaders can be educated and aligned around the sustained work the strategy demands.
Today, the strategy is active in five of Baltimore’s nine police districts, covering roughly two-thirds of shootings and homicides, with the strategy’s reforms already shaping government citywide. Shooting reviews now inform police and prosecutors’ priorities, and the mayor’s office is engaging community and service partners more cohesively. The plan is to scale citywide by early 2026.
There’s still work to do. Group conflicts persist. Progress is uneven. But the drop in gun violence is hard to ignore — and it’s no accident. Baltimore didn’t get lucky; it got serious. And in doing so, it’s helping to rewrite the playbook for how other cities can make lasting reductions in violence possible.