A decade ago, Baltimoreans became lab rats in a fateful experiment: their elected officials decided to treat the city’s long-running crime problem with many fewer cops. In effect, Baltimore began to defund its police and engage in de-policing long before those terms gained popular currency.
 

This experiment has been an abject failure. Since 2011, nearly 3,000 Baltimoreans have been murdered—one of every 200 city residents over that period. The annual homicide rate has climbed from 31 per 100,000 residents to 56—ten times the national rate. And 93 percent of the homicide victims of known race over this period were black.
 

Remarkably, Baltimore is reinforcing its de-policing strategy. State’s Attorney for Baltimore Marilyn Mosby no longer intends to prosecute various “low-level” crimes. Newly elected mayor Brandon Scott promises a five-year plan to cut the police budget. Both justify their policies by asserting that the bloodbath on city streets proves that policing itself “hasn’t worked”; they sell their acceleration of de-policing as a “fresh approach” and “re-imagining” of law enforcement.
 

The motivation for de-policing traces to the city’s botched response to an earlier crime epidemic in the 1990s, when it averaged 45 homicides per 100,000 population, up 55 percent from the previous decade. So in 1999 Baltimoreans elected a mayor, Martin O’Malley, who promised to apply New York’s successful crime-fighting approach, where homicides had plunged by two-thirds over the decade (to one-ninth Baltimore’s rate) thanks to an expanded police force and innovative, proactive policing strategies.
 

O’Malley’s first commissioner, NYPD veteran Ed Norris, initially showed promise. By 2002, Baltimore’s homicide rate was 20 percent below its 1999 level. As O’Malley pressed for more, however, relations soured, and Norris departed (and some financial shenanigans eventually earned him a stint in federal prison). His successor, Kevin Clark, another NYPD import, also became embroiled in personal and professional controversy; he was fired and succeeded by a Baltimore PD holdover. By the time O’Malley moved to the Maryland governor’s mansion in 2007, Baltimore’s homicide rate was back to its 1990s average.
 

The problem was not just turmoil among BPD leadership and meddling (or worse) by O’Malley, but a fatal misunderstanding of what had worked in New York. There, the broad spectrum of criminal activity was addressed efficiently and with community engagement. Detailed data helped guide resources to crime hot spots. Chief William J. Bratton implemented the Broken Windows theory-inspired community-policing methods pioneered by social scientists George Kelling and James Q. Wilson, who understood how small manifestations of disorder could grow to larger ones. Minor offenses that made residents feel unsafe or hinted at acceptance of violence were addressed in order to improve quality of life, strengthen communities, and prevent serious crime.
 

In Baltimore, however, Broken Windows was misunderstood and misapplied. It mutated into a malignant variant, “zero tolerance” policing—and BPD conduct became not just intolerant but unfocused and excessive. As David Simon, a veteran Baltimore crime reporter and creator of HBO’s The Wire, summed things up, O’Malley “tossed the Fourth Amendment out a window and began using the police department to sweep the corners and rowhouse stoops and [per Norris] ‘lock up damn near everyone.’” That sometimes even included Wire crew members on their way home from a long day of filming.
 

True Broken Windows policing, in Kelling’s words, creates “a negotiated sense of order in a community” and involves collaboration between cops and residents. As one BPD vet put it, “You go to a community—before we come in, [we should ask], ‘What are the main things you all can’t stand?’ Everybody playing music at 11:30 at night, kids sitting on the corner, the prostitutes using the little park over there to work their trade. Now, ‘What don’t you care about?’ See the old guys sitting down at the corner playing cards every night? They could stay there all they want. . . . Then the police come in and do what the neighborhood wants. You just don’t go out and lock everybody up.” But, he concluded, “we went overboard.”
 

Kelling had warned that “If you tell your cops, ‘We are going to go in and practice zero tolerance for all minor crimes,’ you are inviting a mess of trouble.” That’s exactly what Baltimore got: stratospheric arrest rates (over 110,000 in 2005, in a city of 600,000), no meaningful reduction in homicides, an ACLU lawsuit, and an erroneous but widely shared feeling that Broken Windows was bunk and policing was not the answer to the city’s crime problems.
 

Then came a respite. O’Malley’s successor, Sheila Dixon (the city’s first female and third black mayor), defied her staff’s recommendations and named as commissioner Frederick Bealefeld, a BPD lifer with no college pedigree. “It was something in my gut that felt he was the best person,” Dixon explained. “I could just feel his passion.”
 

Bealefeld understood community policing better than the New York imports, addressing disorder and crime efficiently. He attended community meetings tirelessly to find out what residents wanted done; got cops out of their cars and walking patrols more often; invested in better training; and supported cops’ work with kids. Partnering with a savvy federal prosecutor, Rod Rosenstein, he targeted known dealers and shooters, emphasizing quality arrests—including of cops on the take. It worked. Even as arrest totals fell (to 70,000 by 2010), so did the homicide rate, to a low of 31 per 100,000 residents by 2011.
 

But that’s when progress stopped and the de-policing experiment began.

Dixon had embezzled gift cards meant for the poor—petty corruption is a Baltimore tradition—and in 2010 was succeeded by Stephanie Rawlings-Blake. The Oberlin-educated former public defender was more liberal than Dixon, personally lukewarm to Bealefeld, and sympathetic to those embittered by O’Malley’s “zero tolerance” policies. And she faced budget problems. De-policing, then, seemed to tick all the right boxes—and, with the homicide rate at a 23-year low (though still almost seven times the national average), there would be little outcry against it.

First came some defunding, with a 2 percent pay cut to help address a recession-related budget pinch; cops’ contributions to their pension funds also were raised to help address shortfalls there. The new mayor’s first proposed budget actually cut the BPD’s request by 10 percent, though the difference eventually was split. Demoralized, experienced cops started retiring in numbers.
 

Rawlings-Blake did not replace them, and she trimmed staffed aggressively. BPD budgets had consistently authorized about 3,900 positions through the O’Malley and Dixon years. Rawlings-Blake took that down by 5 percent in her 2012 budget and another 6 percent in 2013. Bealefeld called the cuts “unconscionable” and retired. As he’d told the head of the police union at one point, “you can only beat down your horses for so long before they give up.”
 

So even before Freddie Gray died in police custody in 2015 and Baltimoreans rioted, the BPD had 460 fewer budgeted “horses” than under Mayor Dixon—with 300 fewer on patrol, conducting investigations, or targeting violent criminals. Not surprisingly, the homicide rate surged 20 percent by 2013. And after the city’s newly elected prosecutor, Mosby, criminally charged six uniformed officers in Gray’s death—though she failed to convict any—proactive policing essentially ceased. The city’s annual body count jumped and has remained tragically high since.
 

Criticized for her handling of the riots—somewhat unfairly—Rawlings-Blake decided not to run for reelection, but in her last two budgets she shaved another 345 personnel from BPD’s budget, nearly halving its investigative staff. (Real BPD expenditures, however, grew about 4.5 percent per year in her term because of mandated pension contributions and ballooning overtime outlays.)
 

Today, then, the BPD patrols the city’s 81 square miles with 18 percent fewer staff than a decade ago. Post-Ferguson, of course, it has become common to point to intuitively plausible but difficult-to-quantify reductions in the level of police effort to explain localized surges in crime; the evidence for this claim, though tentative, is supportive. In Baltimore, the “Ferguson Effect” has intensified an established pattern of diminished policing resources contributing to rising bloodshed.
 

And now Baltimore is among the national vanguard in a new trend: de-prosecution. While it was widely perceived that early in her tenure Mosby put the brakes on prosecution of many “low-level” crimes, once the pandemic began she made that policy explicit (nominally to ensure that overcrowded prisons not become Covid spreaders). She dismissed over 1,400 pending criminal cases and quashed as many warrants for possession or “attempted distribution” of controlled dangerous substances, prostitution, trespassing, public urination or defecation, minor traffic offenses, and more.
 

A year later, she revealed that this policy was not just a Covid palliative but an experiment with human subjects; declaring it a big success, she proclaimed that “the era of ‘tough on crime’ prosecutors is over in Baltimore.” She pointed to a 20 percent reduction in violent crime and a 35 percent decline in property crime in the first quarter of 2021 compared with the same period last year. With all the confounding variables at work during the pandemic, of course, no social scientist worth her salt would proclaim such a complex experiment complete—much less successful—with just a year’s worth of data (or a subsample thereof).
 

When you’ve got data you like, however, “the science” or logic can be overlooked. So Mosby claimed that a 33 percent decline in 911 calls mentioning drugs and a 50 percent decline in calls mentioning sex work during her experiment proves that “there is no public safety value in prosecuting these offenses.” To the contrary: with drug use and prostitution de facto legal in Baltimore, many residents still wasted their time calling the cops about the dealers, junkies, hookers, or johns on their block.
 

Then there is Mosby’s spin that focusing “the limited law enforcement resources we have” on murder, armed robbery, and carjacking will magically lead to a safer Baltimore. Yet it is Mosby who has been running the State’s Attorney’s office for over six years, during which time her staff has grown by 14 percent (with 50 added positions) and her real budget by 27 percent. A cynic might suggest that the resource limits she imagines are a byproduct of her active travel schedule or other distractions.
 

A simpler explanation is that Mosby is just not very good at her job. Pre-pandemic, violent crime surged on her watch; homicides (averaging 55 per 100,000 residents) have run one-fifth higher than in any prior administration. Conviction rates fell as soon as she took office. According to Sean Kennedy of the Maryland Public Policy Institute, in 2017 only 12 percent of murder, attempted murder, or conspiracy-to-commit-murder cases resulted in a guilty plea or verdict for the murder charge. In 2018, only 18 percent of gun-crime defendants were found guilty.
 

It’s true, of course, that BPD resources—measured in actual boots on the ground—have been increasingly scarce in recent years. Mosby ignores the rather obvious implications of that trend while drawing dubious conclusions from her own too-brief, badly designed test. A dispassionate look at Baltimore’s decade-long experiment with de-policing seems fairly clear: people die.
 

De-prosecution is likely to amplify this tragic tendency. Now that sellers and buyers of drugs and sex face lower risk (or no risk) of prosecution in Baltimore, these markets will expand and become more profitable. The gangs that supply these products often compete for market share by violent means; their customers sometimes fund their habits with muggings and assaults.
 

As Kelling and Wilson taught and many cities’ histories have demonstrated, disorder and crime can be contagious, but policing these problems efficiently and with community involvement can yield major improvements in public safety and quality of life. Baltimore is simply ignoring these lessons. Other cities should not follow suit.
 

Stephen J. K. Walters is the author of Boom Towns: Restoring the Urban American Dream, chief economist at the Maryland Public Policy Institute, and a fellow at Johns Hopkins University’s Institute for Applied Economics, Global Health, and the Study of Business Enterprise.