Baltimore’s homicide clearance rate is rising, but arrests for murder remain flat

This story was produced in partnership with The Baltimore Brew. 

With homicides declining significantly in the last two years, Mayor Brandon Scott has lauded his administration’s role in the trend, calling Baltimore “the model for violence reduction” at his State of the City address in April.

In 2024, there were 201 murders in Baltimore. It was the second consecutive year of significant homicide reductions in the city, following a drop from 333 murders in 2022 to 261 in 2023.

These declines followed trends nationwide. As of early May, year-to-date homicides are down 30% in New Orleans, 29% in Philadelphia, and nearly 24% in Chicago. At the end of May, Baltimore had 54 homicides for in 2025, the fewest murders in the first five months of any year on record.

City officials have also pointed out that the Baltimore Police Department’s homicide clearance rate increased from 42% in 2020 to 68% in 2024.

This statistic puts the department at “over 10% better than the national average,” Scott said during his State of the City address.

Police Commissioner Richard Worley said these numbers show “what our men and women can do when they’re given time to do an investigation and aren’t running from shooting to shooting or homicide to homicide.”

But an analysis of Baltimore crime and arrest data by The Garrison Project and The Baltimore Brew shows that while the BPD’s homicide clearance rate has significantly increased, the number of arrests for murder has stayed about the same.

As far back as 2005 the number of arrests BPD made for murder has largely remained in the 100-130 range each year. Over a few years, the number of murder arrests dipped below 100. In 2019, for example, there were 348 homicides and just 89 arrests for murder, leading to a clearance rate of 31%.

When BPD had a 36.3% clearance rate in 2022, it made 103 arrests for murder. By 2024, it had achieved a nearly 70% clearance rate, but the number of arrests for murder increased to just 127. Arrests for murders ticked up slowly because police had far fewer murders to solve: homicides declined from 333 in 2022 to 201 last year.

What does it mean that arrests for murder have remained stagnant for so long as violence has declined so rapidly? Looking back over the past two decades, when homicides rose and fell dramatically, arrest numbers varied little.

“Clearance rate is tricky”

BPD leaders and the Mayor’s Office of Neighborhood Safety and Engagement (MONSE) downplay murder arrest numbers, arguing that clearance rates are the more meaningful metric.

A police department’s clearance rate is determined by dividing the total number of murders cleared by police by the total number of murders recorded in that given year.  The FBI defines a murder as “cleared” when the person that police believe committed the murder is arrested, charged or referred to court for prosecution.

A homicide can also be cleared by “exceptional means” when the offender has been identified but cannot be located and arrested for a number of reasons. This includes death.

National clearance rates for homicides have fallen for decades, from 91% in 1965 to 57.8% in 2023. But the high clearance rates of the past may have been inflated by shoddy police work that led to wrongful arrests and convictions.

Baltimore has paid out millions in legal settlements stemming from bad cases by BPD’s famed 1980s-era “murder police.” In 2023, the city approved a $48 million settlement to the “Harlem Park Three,” a group of Black men wrongfully convicted at age 16 for the 1983 shooting death of a 14-year-old.

According to BPD, its clearance rate is the metric they most closely monitor.  Spokesperson Lindsey Eldridge said that clearance rates “allows us to measure our progress and success compared to the national average and similar sized city averages. BPD has made significant strides over the last year, achieving a 28 point increase in the clearance rate compared to 2023.”

But data analyst Jeff Asher warns that clearance rates can be “manipulated intentionally or accidentally.” Asher has also written about how fewer crimes usually means higher clearance rates: often, the clearance rate rises because there are fewer murders.

“You can make the case, ‘Hey, look, we’re solving more murders. Look at all the great work we’re doing,’ when in reality, you’ve simply lowered the denominator,” Asher said. “It’s more complicated than that and it’s not always the case, but that’s what you frequently see with a lot of cities.”

Criminologist John K. Roman, a senior fellow at the National Opinion Research Center at the University of Chicago and director of its Center on Public Safety and Justice, said the “clearance rate is tricky” because the definition of a cleared homicide is broad.

“What does it mean for a case to be cleared? It could be cleared because somebody’s arrested. It could be cleared because the investigating officers are convinced they know who committed the crime but do not pursue and arrest that person for a variety of reasons,” he pointed out.

In short, a clearance does not always mean an arrest. Murder arrests as a metric is also somewhat limited because murders can have more than one perpetrator, and one perpetrator could commit more than one murder.

But unlike most crimes, homicides are usually reported to the police, making it a reliable metric to measure violence and crime rates. Roman said that “arrests for murder is one clean way to understand a performance measure for law enforcement.

Two decades worth of 100-130 arrests for murder in Baltimore, regardless of the number of murders in a specific year, he observed, is a sign of “state capacity.”

“What is the ability of a state or local agency to perform its duties, provide its resources and do all of the things that we want it to do?” Roman asked. “If I see a number that’s really stable, that tells me that Baltimore’s capacity to close more cases will require additional resources.”

Staffing and “capacity”

BPD’s failure to substantially increase the number of arrests for murder over two decades can also mean that its resources are not used effectively.

Conversations around department capacity primarily focus on police staffing. Since 2000, when there were around 3,000 full-time sworn personnel, BPD staffing has declined to approximately 2,500 sworn members.

U.S. District Judge James Bredar, who oversees the federal consent decree on the BPD, said the department is in a “staffing crisis” and should hire 500 more officers. He said the city is “playing with fire” if it does not hire more officers. BPD’s police union has said the same.

Data provided by BPD to The Garrison Project shows that the homicide unit is staffed at around 50 detectives, though it has sometimes dipped to under 40. The caseload per detective has, at times, risen to as many as 10 cases per detective.

In 2024, when the clearance rate soared to 68.7%, the homicide unit had 39 detectives, the smallest number since 2016, when there were 37 detectives and the clearance rate was 38.7%.

BPD says that its investigative capacity increased thanks to the hiring of nearly 30 civilian investigators since 2023.

“The on-boarding of professional civilian investigative specialists has been a game-changer,” BPD’s Eldridge said. “These specialists provide critical support to detectives and have taken on a significant share of unattended death investigations.”

BPD says it has also been assisted by an array of new surveillance technologies including license plate readers and stingrays, which mimic cell tower signals and force cellphones to communicate with them.

“There is a difference between deterrence and prevention,” said Roman, the criminologist. “All the police technology and tools fall into the deterrence bucket: The idea that if people think that their likelihood of being caught committing a crime is going up, they’re less likely to commit that crime.”

Detective bias?

Defense attorney Karyn Meriwether, who has been trying murder cases for over 20 years with the Maryland Office of the Public Defender, says “detectives have the ability to identify suspects much more quickly because of the CCTV cameras and other technology. I’d say as recently as 2020, it really was unusual to have any [surveillance and security camera] footage in a case. Now it is everywhere.”

Meriweather cautioned that technology can also often exacerbate confirmation bias in homicide cases.

“My experience, including very recently, is that it is a devil to get a prosecutor or a detective to move away from what their original thinking about a suspect is – often to the exclusion of all other possibilities,” she said. “I respect my SAO colleagues in the homicide unit and the quality of their work in general. But I think the detective bias in conducting investigations is ingrained into their practice.”

Meriwether added that the discovery process – information and evidence handed over to defense attorneys from police and prosecutors in a criminal case – has greatly improved. “As a defense attorney,” she said. “I can get investigation materials that I couldn’t get before that we are entitled to receive.”

But Meriwether said that what she sees in discovery sometimes troubles her.

“I see more conscientiousness and transparency now for sure—and an ability to identify a suspect more quickly using technology,” she said.

“But in terms of how the police connected the dots in their investigation and how that led them to my client? That hasn’t improved. Old-fashioned, boots-on-the-ground investigation? That has not improved. I wish I could say otherwise.”

Deterrence and prevention

Mayor Scott’s comprehensive violence prevention plan includes a Group Violence Reduction Strategy that he has credited for recent declines in Baltimore homicides. It uses a “carrot and stick” model, promising access to social services while charging people in alleged criminal groups.

Stefanie Mavronis, director of MONSE, said that GVRS, which is in place at four of the city’s nine police districts, has helped police take a more targeted, intelligence-driven approach to homicide cases.

“There are lots of things that we’re doing now that help us be much more intentional about who is even being looked at for an arrest,” Mavronis said.

“With our Group Violence Reduction Strategy, we saw great reductions in homicide, non-fatal shootings, carjackings,” she continued. “But we didn’t see a commensurate increase in arrests that went along with it, so as that enforcement action was happening, it was happening in a much more targeted and intentional way.”

A 2024 analysis of GVRS by The Garrison Project shows that the program’s arrests are overwhelmingly for non-violent crimes, with few for shootings or murder.

Mavronis believes that the city’s combined deterrent and preventive resources could help move BPD out of its range of 100-130 arrests per year for murder.

“I wouldn’t peg success on the number of arrests alone. I think because there’s probably so many different factors working at the same time to push that and kind of keep it in this area that it’s in,” she said. “I’d be curious to see, does this trend hold in the next five years if we continue to drive down the number of homicides and shootings?”

Entrenched policing culture

Roman identified victim’s services as an important place where deterrence and prevention intersect. Treating shooting victims and other victims of crime better can lead to greater trust between police and the community, which increases the likelihood of crime-solving.

“The person most likely to shoot in the future is somebody who themselves have been shot, and we do a terrible job with these people,” Roman said. “Part of the reason we do a terrible job with these people is that they tend to know who shot them, and they tend not to want to tell the police about it, because they’re afraid. They just got shot. They’re afraid of being shot again. It’s completely reasonable. But when the investigators go to see them, the investigator’s upset that they can’t clear the case.”

The approach to shooting victims that Roman describes is documented in a report released in 2022 from the U.S. Department of Justice’s National Public Safety Partnership about BPD and its victim services.

“One of the important themes of the report is around the ways people are treated, in particular in the aftermath of some of the most traumatic events someone can experience,” said Heather Warnken, the report’s lead author.

“We heard repeatedly that BPD was treating people like pieces of evidence and not calling people back. Especially when they were done with them or didn’t think they’re useful. But they were a mom whose kid had just been shot to death who wanted an update on their dead child.”

The report led to some reforms, including MONSE expanding victim services. But Warnken said her “understanding is that many of the dynamics we laid out in that report persist.”

For Warnken, BPD’s behavior described in the report represents an “entrenched policing culture.” She said that culture is dismissive of community concerns and “denigrates the community relationship piece [of crime reduction] and sees that as separate from the people solving crimes.”

Warnken said that this “bifurcation” between community relations and crime solving might explain how increased resources to BPD have not moved the needle on arrests for murder.

“Commissioner Worley and others talk about how successful the consent decree has been and how BPD is this unrecognizable, improved department,” she said. “There has been some really important progress. But if BPD made the strides that they claim to have made in transforming their relationship with the community, that would translate to better, more communicative relationships that play a big role in actually solving cases.

“We should be paying attention to the clearance rate. It is important for a lot of reasons. But so is connecting the dots and really building cases.”