On January 2 of this year, the Baltimore Police Department sent out its annual analysis of the city’s homicides: in 2018, there were 309. The majority of the victims were killed in the street and almost sixty percent died from a gunshot wound to the head. Ninety-four percent were black men.

The analysis also included data on the criminal records of the victims. The vast majority of homicide victims had been arrested before. One in four were either on parole or probation when they died.

“We get it, bad people killing bad people, right?” the Baltimore Sun’s editorial board wrote, criticizing the police department’s report. “The department seems to be assigning some blame to the victims rather than assessing its own inability to bring the violence under control.”

But such victim-blaming from the police department hints, unwittingly, at a more fundamental, and paradoxical, problem: The police are uniquely unqualified to protect the people dying violent deaths in Baltimore because they are often also out to arrest them. Cooperating with law enforcement is often morally and practically unthinkable for the city’s most vulnerable, and it’s similarly absurd to think police will safeguard the same people at whom they regularly point their guns.

It’s not just Baltimore, either. Around the world, people who commit crime tend to be crime victims as well. Policing can’t solve that riddle.

The pioneering victimologist Hans Von Hentig’s 1948 book The Criminal and His Victim is a dubious artifact, complete with lengthy analyses of the relationship between misbehavior and hair color (red is bad), handedness (lefties seem to be bad, but more study is needed), and twins (bad in general). But the book continues to be cited because it marks the start of modern scholarship on how victims of crime and those who commit them are “intimately connected.” There wasn’t a simple division between “evildoer” and “evil-sufferer,” Von Hentig theorized. Instead, a “nefarious symbiosis” between criminal and victim explains much of the world’s crime.

Ten years later, the criminologist Marvin Wolfgang’s research on homicide victims in Philadelphia gave empirical heft to Von Hentig’s theory — 47 percent had arrest records, and 17 percent had been arrested for aggravated assault. In the decades since, studies in places as disparate as Iceland, New Mexico, and Colombia have confirmed that there is a large “overlap” between people who commit crime and people who are crime victims. This phenomenon is now a foundational fact of criminology.

But the overlap is stronger for people who live in poor areas. People who hurt people are much more likely to get hurt themselves if they live somewhere with a high poverty rate. Some scholars cooked up racist theories to explain this: there was more of an overlap in poor, black neighborhoods because their residents felt violence was an acceptable way to resolve disputes. Life, to them, was cheap.

Most people in violent places strongly disapprove of violence, but they also often disapprove of the only people with the legal power to confront it.

Except this wasn’t true. Researchers have since found that people across the socioeconomic spectrum have similarly negative attitudes toward violence — if anything, some studies found that black and Latinx people in the U.S. are actually less tolerant of violence than whites.

In 1998, Dawn Bartusch and Ralph Sampson developed a new concept, “legal cynicism,” that has helped guide crime scholarship. Most people in violent places strongly disapprove of violence, but they also often disapprove of the only people with the legal power to confront it. That tension produces the phenomenon that Bartusch and Sampson describe (though “legal realism” might have been a more apt term for it).

Violence is typically framed as an act of anarchic pleasure, but it’s often how people address disputes. After all, the large majority of crime victims are hurt by someone they know. A murder can be impulsive and horrifying and also moralistic, as the sociologist Donald Black explained in his 1983 paper “Crime as Social Control.” That doesn’t mean moralistic murders are a social good, but Black’s influential work helps show that a lot of violence isn’t born of sadism or nihilism. Instead, it’s often a kind of desperate “self-help” — like reclaiming your stolen property — produced by the justice system’s chronic failure to render justice.

“Lower-status people of all kinds — blacks and other minorities, the poor, the homeless — enjoy less legal protection,” Black wrote. “Crimes of self-help are more likely where law is less available.”

Police seldom solve crimes of self-help — they rarely even find out about them. The majority of overall violent crime isn’t reported, and people who have criminal histories are even less likely to call the police if they get victimized. In 2018, only 30 percent of the Baltimore Police Department’s homicide investigations led to an arrest. But homicides are actually among the crimes the police have the best chance at solving simply because they know the crime happened. Rapes, robberies, and assaults often go unreported, but dead bodies are hard to hide.

Still, police often say they can’t solve cases unless witnesses come forward, and many don’t. In 2004, an underground documentary called Stop Fucking Snitching hit the city’s streets, featuring impromptu speeches and raps by a couple dozen Baltimoreans. Produced by Rodney Bethea, a local entrepreneur, and Skinny Suge, a gang leader, the movie was only distributed by DVD, and it was arguably dangerous — people named alleged snitches and brandished guns. Baltimore-bred NBA star Carmelo Anthony made an appearance while hanging out in his old neighborhood, making the documentary into a national story.

“Hope they get hit by a milk truck and just be creamy as a motherfucker and have no foam on top,” Suge declares in the movie’s opening minutes. “Snitchin’ bitch.”

The city’s criminal justice apparatus used the film to push through a new law to combat witness intimidation. People in Baltimore have been murdered for talking to the police, which obviously deters some who might otherwise cooperate — leading the police to solve fewer crimes, which means people are even less apt to turn to them, creating a downward spiral of mistrust.

But the justice system can be just as vindictive. In 2013, Rashaw Scott survived a shooting that killed his one-year-old son, who had been sitting behind him in his car in an infant car seat. Scott didn’t want to cooperate with prosecutors, so a Baltimore judge had the grieving father arrested, jailed, and hauled onto the witness stand in an inmate’s jumpsuit with his hands cuffed to his waist.

Illegal police violence is a durable feature of law enforcement in black communities, and it has measurable consequences.

Police also regularly reaffirm the idea that they’re an occupying force. For example, on the morning of April 12, 2015, a 25-year-old black man named Freddie Gray was standing on a Baltimore street corner when he saw three police officers. He started running. The police chased him down, arrested him, and loaded him into a police van. An hour later, he was taken out of the van with a broken neck. He died after seven days in a coma. The next month, Baltimore City State’s Attorney Marilyn Mosby charged six officers in relation to Gray’s death. Police opposed the prosecutions, and many began to do their jobs passively in protest. Crime spiked in the city around the same time, and many attributed it to “the pullback” by the police. (Higher police presence does lower street crime, studies show.)

But Gray’s death likely also provoked a different kind of pullback. Illegal police violence is a durable feature of law enforcement in black communities, and it has measurable consequences. In 2004, a mob of off-duty police in Milwaukee savagely beat a black man named Frank Jude. Three months later, after the case went public, 911 calls from black people to report crimes in the city dropped sharply and stayed down for more than a year, according to a 2016 study.

“Publicized cases of police violence not only threaten the legitimacy and reputation of law enforcement,” the study said. “They also — by driving down 911 calls — thwart the suppression of law breaking, obstruct the application of justice, and ultimately make cities as a whole, and the black community in particular, less safe.”

In May 2013, Carlos Wheeler shot a young black man named Ronnie Thomas III, who got hit in the leg but survived. In April 2014, Thomas III was arrested for shooting at someone. In June 2015, Thomas III was shot and killed. The following month, Carlos Wheeler’s brother, Nathanial, was shot and killed — in retaliation for Carlos shooting at Thomas III back in 2013, prosecutors said.

Skinny Suge, who produced Stop Fucking Snitching, is legally named Ronnie Thomas — Ronnie Thomas III was his son. Addressing Baltimore’s epidemic of violence means reaching Baltimoreans like the Thomases. Police might not be the best people for that job. There are dozens of anti-violence organizations in Baltimore unaffiliated with the criminal-justice system. Perhaps the biggest is the city government’s Safe Streets program, which has about 80 employees working out of nine sites across the city. The program follows the Cure Violence model pioneered in Chicago back in 2000, and recently repopularized by the 2018 documentary Charm City. (It’s not yet clear exactly how well the model works. A 2015 paper found that it has “promise.” But researchers still don’t have good data because of its inconsistent implementation, and the difficulty of disentangling the various factors that affect how much gun violence happens in a given neighborhood.)

The model is best known for its “violence interrupters,” people who work on the ground, mediating conflicts that might turn violent and defusing tensions. They typically have shared experiences with the Baltimoreans getting shot and doing the shooting, which gives them credibility — along with the critical fact that they don’t work with the police or any other part of the justice system. Violence interrupters solve problems like getting stolen property returned or helping kids to resolve who had the right to sell drugs on a certain corner.

They work towards justice outside of the justice system, which is a fragile place to be. Baltimore’s police “hate” Safe Streets, the Washington Post reported in 2016. Imhotep Fatiu, a program manager for the organization, politely declined to answer questions about the dynamic between Safe Streets and the police department.

“Part of our training is to steer clear of conversations surrounding the police,” Fatiu told me. “We have a job to do and they have a job to do.”

Grassroots organizations in the city are small operations that generally rely on unreliable and inadequate funding. The city even forces its own Safe Streets program to help pay for its operations with private donations, the Baltimore Sun reported in 2017.

Meanwhile, the Baltimore Police Department spent about half a billion dollars of public money last year, the single largest portion of the city’s $2.9 billion budget. Safe Streets is a rounding error compared the department’s 3,000 employees. The city effectively grants the police a monopoly on the crime control business.

Police are supposed to use their immense power to protect people. But those who need protection most are often the same people the police want to punish. Law enforcement is premised on an imaginary world in which victims are innocent and criminals have nothing to fear but the law. No reform can fix that. Public safety for all means tearing down that binary fantasy and all the institutions that uphold it.