What Can Cities Learn From Baltimore’s Extraordinary Decline in Violence?
When Baltimore Mayor Brandon Scott proposed his signature violence reduction strategy in 2021, he was asking community members to believe that an effort that had already failed — twice — could work if the city tried again.
Baltimore had tried versions of the strategy, known as focused deterrence, in the 1990s and in 2014. The approach — often oversimplified as a “carrot-and-stick” tactic for violence prevention — aims to identify the small number of people most likely to be involved in shootings and offer them both help and the promise of legal consequences if violence continues. Focused deterrence had a long history, including successes in other cities, and a strong research pedigree. But in Baltimore, both efforts fizzled, undermined by impatience, uneven follow-through on enforcement, and a failure to deliver the services that officials promised.
That history hung over Scott’s plan as his administration prepared to launch what became known as Baltimore’s Group Violence Reduction Strategy, or GVRS, in early 2022. The city was coming off seven consecutive years with more than 300 homicides. Community meetings meant to build support for the plan drew sparse attendance. Some residents were weary of new crime plans. Others doubted City Hall could keep its promises.
Nearly five years later, Baltimore is in a very different place.
Through the first six months of 2026, the city logged 50 homicides — the fewest in Baltimore during the first half of a calendar year since at least the 1960s, according to the city. Homicides were down 23 percent compared with the same point last year, while nonfatal shootings were down 2.5 percent. From 2019 to 2022, Baltimore had already surpassed 150 homicides by the midpoint of the year. In 2023, the city recorded 142 by this time of the year. But in 2024, Baltimore fell below 100 homicides in the first six months for the first time since 2014, and the number has been declining since.
“We are making progress on public safety that many thought was impossible,” Scott said in a statement announcing the numbers last week.
It’s unlikely that a single program caused Baltimore’s almost miraculous decline in violence. Instead, a growing body of research suggests that the city’s approach — a coordinated system combining focused deterrence, community violence intervention, victim services, improved investigations, and sustained leadership from City Hall — may be one of the clearest examples yet of a city reducing gun violence at scale.
The strongest evidence comes from a new evaluation of GVRS by researchers affiliated with the University of Pennsylvania’s Crime and Justice Policy Lab. The strategy launched in January 2022 in the city’s Western District, then Baltimore’s most violent police district, before expanding elsewhere. The study found that within 18 months, GVRS was associated with a roughly one-third decline in shootings and homicides in the Western District compared with similar areas of Baltimore that had not yet received the strategy. Carjackings also fell by roughly 40 percent.
Ben Struhl, executive director of the Crime and Justice Policy Lab, said one reason Baltimore’s effort worked better than past attempts is that the city treated GVRS not as a stand-alone program but as an approach to governing.
“It’s not a program,” Struhl said. “It is a city strategy that starts with the mayor as one of the key political authorizers, and then tries to arrange government through a lot of different management means and some theory-driven activities to actually bring down violence at scale.”
That distinction is important, Struhl said, because cities often adopt pieces of focused deterrence — call-in meetings, community messaging, or enforcement actions — without building the full system needed to reach community members with resources and support to make the strategy work. Baltimore’s earlier attempts suffered from that problem. Using the “carrot-and-stick” simplification, Baltimore haphazardly wielded the stick and forgot the carrot. The current version, he said, was built around data, weekly reviews of shootings, a real service component, community messengers, and credible follow-through.
“In our experience, it’s never seemed to work anywhere with just the stick,” Struhl told me. “There always needs to be a genuine offer of support that you can follow through on.”
That was one of the central lessons Baltimore officials say they took from past failures.
Stefanie Mavronis, director of the Baltimore Mayor’s Office of Neighborhood Safety and Engagement, said the city had previously tried to scale focused deterrence too quickly and had not invested enough in the service side of the strategy. This time, she said, the city launched in one district and spent a full year there before beginning a measured expansion.
“We’ve had to be really intentional and have candid conversations and collaboration with our law enforcement to make sure that we’re not moving in such a way that outstrips capacity,” Mavronis said.
Today, GVRS is operating in six of Baltimore’s nine police districts, with plans to expand citywide. The strategy brings together the neighborhood safety office, police, prosecutors, probation, the Department of Juvenile Services, community messengers, and community-based service providers, including Roca and Youth Advocate Programs. After shootings or other serious incidents, the city identifies people at elevated risk of being involved in retaliation or future violence. Some receive direct communications warning them to stop shooting and offering help. Others are referred to service providers.
Palermo said each partner has a defined role. “I’m not a police officer. A police officer is not Roca staff. We have our own expertise, and we do it,” he said. “That’s what works. We meet on a weekly basis, and we’re talking about referrals.”
That coordination, he said, is a major shift from the Baltimore of a few years ago.
“How did Baltimore get from 300 homicides for almost 10 years to 200 to 100?” Palermo said. “It didn’t happen this year. It didn’t happen in 2025. It really is the last five years, starting with the mayor’s violence reduction plan.”
Baltimore’s progress also cannot be separated from its community violence intervention ecosystem. Safe Streets Baltimore, the city’s Cure Violence-style violence interruption program, has operated since 2007 and now has 10 sites. A recently released Johns Hopkins University evaluation found that Safe Streets was associated with a 42 percent reduction in homicides among youth ages 15 to 24 and a 21 percent reduction in nonfatal youth shootings.
The findings come with a caution: The estimates were not statistically significant, and effects varied from site to site. But Shani Buggs, a University of California, Davis researcher who co-authored the Safe Streets evaluation, said the results are “suggestive of substantial violence reductions” in areas where Safe Streets has operated, especially longer-running sites.
Safe Streets occupies a difficult lane in Baltimore’s public safety strategy: Frontline workers, who often come from the same communities most affected by violence, are asked to intervene in some of the city’s most volatile conflicts. But the effort has become a political and media punching bag when something goes wrong, including the arrest last month of an off-duty Safe Streets worker in connection with a nonfatal shooting.
“What they chose to do was obviously incompatible with the work of Safe Streets,” Mavronis said, adding that the incident raised fair questions about oversight, trust, and training.
“We’re all committed to maintaining that community trust,” she said. She said staff responded by increasing their visibility in the community, supporting affected workers, and recommitting to checking on outreach workers’ mental health and emotional state before shifts. “We’re spending a lot of time thinking about how we strengthen our data tracking, professionalizing the work, and investing in our frontline staff.”
Buggs, who has studied violence prevention efforts nationwide, said support for violence prevention workers is still an area for improvement in Baltimore, and it could make the program more effective in the future.
“Safe Streets continues to be plagued by challenges in adequate staffing, adequate pay, adequate mental health and therapeutic support, and adequate benefits,” she said. “There has been substantial improvement, and more needs to be done. That impacts, from site to site, how successful the sites can be.”
No one program explains Baltimore’s homicide decline, and strengthening each component strengthens the city’s strategy overall, Buggs said. “A takeaway from Baltimore is that all of those pieces are important,” Buggs said. “Without a piece of the puzzle, the puzzle is still incomplete.”
Aside from Safe Streets and GVRS, she said, grassroots groups, philanthropic support, the Mayor’s Office of Neighborhood Safety, and several other mayoral offices focused on children, families and African American male engagement are also part of that puzzle.
Daniel Webster, a Johns Hopkins researcher who co-authored the Safe Streets evaluation and advised Scott’s transition committee on public safety, said Baltimore’s story also includes a less discussed piece: improved accountability for serious violence.
Webster said homicide clearance rates have improved significantly, and arrests for serious violent crimes such as aggravated assaults and robberies have increased even as violence has declined. Year-to-date, Baltimore’s homicide clearance rate is 60 percent, up from 54 percent compared to this time last year. At the same time, he said, weapons arrests have gone down — suggesting that the city is not simply casting a wider enforcement net.
“They’re actually going after people who are committing violence,” Webster said. “The big change is that more people who are committing serious violent crime are getting arrested than they used to, even while the incident rate is declining.”
The GVRS evaluation found a similar pattern in the Western District. Despite targeted law enforcement being a key component of the strategy, overall arrests did not increase, and arrests for nonviolent and drug offenses did not rise. Instead, arrests for serious violent crimes increased, suggesting that enforcement became more precise rather than overreaching.
That finding is crucial to understanding why Baltimore’s current approach appears to have avoided some of the pitfalls that have undermined focused deterrence efforts elsewhere. In cities where the strategy devolved into aggressive policing, community trust suffered. In cities where officials failed to deliver services or consequences, the strategy lost credibility. The researchers said Baltimore’s version appears to have worked, at least in part, because it hewed closely to its plan and tried to do both — offer help and follow through — while narrowing the focus to the people and conflicts driving violence.
There are still reasons for caution. The strongest GVRS evidence comes from the Western District during the strategy’s first 18 months, not from every district where it has since expanded. The Safe Streets evidence is promising but not definitive. And Baltimore’s decline likely reflects other factors, including improved victim services, reductions in ghost guns, and broader shifts that researchers are still trying to disentangle.
But Baltimore’s recent trajectory is hard to dismiss. Unlike many cities now celebrating post-pandemic declines, Baltimore did not experience a major COVID-era homicide spike. Its violence had already surged years earlier, after the 2015 death of Freddie Gray and the unrest that followed, and remained stubbornly high for nearly a decade. In other words, the city’s current decline is much more than a return to normal.
Mavronis said the drop is beginning to register in a city long accustomed to violence that once seemed immovable. “I think even for the most hopeful of us five years ago who were engaged in this work, I don’t think anyone could have guessed or would have believed that we could be here,” she said.
Mavronis said the changes came after years of building a system designed to identify risk more precisely, respond faster, offer credible support, hold people accountable for serious violence and keep community partners at the table. That may be one of Baltimore’s most important lessons for other cities, Mavronis said. Still, she warned against treating the progress as permanent.
“There’s a real risk in this moment of violence going down that we start to take our foot off the gas and switch to the next thing,” Mavronis said. “For us in Baltimore, we’ve seen different times where violence has gone down for a year or two and then it’s gone right back up. We wanted to make sure that this is sustainable.”
Chip Brownlee is a staff writer at The Trace covering federal policy related to violence prevention and firearms. He is also the author of The Trajectory newsletter, which spotlights the people, policies, and programs grappling with America’s gun violence crisis. Before joining The Trace as an investigative fellow in June 2020, Chip worked as a reporter and the editor-in-chief of his collegiate newspaper, The Auburn Plainsman. He also covered the state legislature, governor, courts, and elections for the Alabama Political Reporter.
At The Trace, we’re building the only team of journalists exclusively dedicated to reporting on our country’s gun violence crisis. We use the power of journalism to improve public understanding, increase accountability, and identify solutions that can lead to safer homes and communities for all Americans.