Eric Thayer/Reuters

Police Commissioner Bill Bratton and Mayor Bill de Blasio at a news conference about the police department’s use of stop-and-frisk practices, Brownsville, Brooklyn, January 2014

The death of Eric Garner at the hands of his arresting officers in New York on July 17 has provoked a public debate about the so-called broken windows style of policing that has been, in various incarnations, the New York Police Department’s guiding tactic since 1994. The phrase “broken windows” is a metaphor that neatly illustrates the policy, as first put forth by James Q. Wilson and George L. Kelling in a 1982 essay of that name in The Atlantic. If a window in a building is broken and left unrepaired, the rest of the windows will soon be broken as well, because the unrepaired window signals that no one cares. This explains why the police should make arrests for panhandling, public drunkenness, loitering, and other minor infractions that have long been considered unavoidable by-products of urban street life: if allowed to flourish, they foster an atmosphere of disorder that causes law-abiding citizens to feel fearful and wary, as if the streets of their neighborhood have been invaded and are not theirs.

Believing that this general atmosphere of disorder reduces their chances of being caught, the theory goes, violent criminals feel emboldened. Since disreputable minor offenders create this atmosphere in which violent crimes are more likely to be committed, they should be swept off the streets as if they were violent criminals themselves, and physically roughed up, if necessary, even if they may not be breaking the law.

Had it not gone awry, the Eric Garner case would have been a typical example of the policy at work. His offense, by all accounts, was that of selling loose cigarettes in a park near the ferry on Staten Island, and then verbally protesting policemen’s attempts to arrest him. Three hundred and fifty pounds, asthmatic, forty-three years old, and black, Garner was put into a chokehold and died, according to the New York medical examiner, “of compression of chest and prone positioning during physical restraint by police.” Garner was a frequent presence in the park and there is no doubt that the arresting officers from the 120th Precinct knew that he wasn’t a violent threat. The medical examiner has ruled his death a homicide and a grand jury has begun hearing evidence to decide whether criminal charges should be brought against the policeman who employed the chokehold, which is banned by the NYPD but continues to be used by some officers when subduing suspects.1

The debate about the broken windows method of policing unavoidably turns around the question of racial injustice. By an overwhelming majority, New Yorkers who are arrested for low-level infractions—“rule-breaking” may be a better term—are young black and Hispanic men in poor neighborhoods. Often these arrests have been for possessing tiny amounts of marijuana, pushing these men into the criminal justice pipeline for a drug that is sold legally for recreational use in two states and for medical use in twenty-one others.

Currently, in New York, possession of less than twenty-five grams is not a crime unless you are caught lighting up in public or, in the language of the law, the drug is “open to public view.” A beat cop on foot patrol, instructed to enact the policy, may approach a person he deems to be suspicious. He orders the suspect to empty the contents of his pockets, which may contain a couple of grams of pot. The suspect has now publicly displayed the drug and is arrested according to the letter of the law. Black and Hispanic men make up 86 percent of these busts.

The stop-and-frisk tactic that was heavily employed during the Bloomberg administration is an example of the logic of the broken windows theory taken to an extreme: beat cops were told to dispense with the pretense of minor infractions to identify suspects; simply being on the street became sufficient cause for a frisk, and then a search, to take place. Eighty-seven percent of those stopped were black or Hispanic.

In 2011, two years before Judge Shira Scheindlin of the US District Court ruled that stop and frisk was a form of “indirect racial profiling” and that it violated the Constitution’s equal protection clause, more than 50,000 New Yorkers were arrested for “displaying” tiny amounts of pot. After the state court complained about the flood of cases, Bloomberg’s police commissioner, Ray Kelly, scaled back the arrests. But in 2013, small pot busts still numbered about 28,000. As of June of this year, with Bill de Blasio as mayor and a federal monitor watching over the NYPD as a result of the stop-and-frisk ruling, small pot busts were on track to be at least as numerous.

The inequity is glaring. With the aim of maintaining order in poor, high-crime neighborhoods, police saddle thousands of young men with criminal records for an offense that the state has largely decriminalized and that white people regularly commit with impunity. Penalties imposed by the courts for possession are usually minimal—dismissal of the case after six months if the person has no further run-in with the law—but the damage can still be considerable, taking the form of rejected job and housing applications or being banned from joining the military and attending certain colleges.


William Bratton, the current police commissioner, continues to endorse the busts, but marijuana possession is not a “feeder crime,” as he has often maintained. In 2012, Human Rights Watch found that only 3 percent of those arrested for low-level possession went on to commit a violent felony.

New York’s legacy of using drug possession to create a criminal class that had not existed before, and that disproportionately singles out blacks and Hispanics, can be traced back to the Rockefeller Drug Laws of 1973. Pushed through the state legislature when Nelson Rockefeller was governor, they were the toughest laws of their kind in the country, mandating, for example, a minimum of fifteen years to life in prison for possessing 113 grams of pot. The mass incarceration that resulted is universally recognized to have been a catastrophic failure, and most of Rockefeller’s mandates have been amended. But the psychology of the practice is still ingrained, both in the criminal justice system and in the public’s way of thinking about crime.

Kenneth Thompson, the newly elected Brooklyn district attorney, who is black, has made it clear that he is fed up with the pot busts and the burden they place on the borough’s poor. In July, he announced that he would not pursue cases against people arrested with less than twenty-five grams who had no or “a very minimal” prior record. The issue for Thompson is racial fairness, not whether pot should be legal, and his announcement reflects the determination of some of the progressive officials that came to power this year to take an active part in changing the way the criminal justice system operates in New York.

I saw for myself some of the effects of these low-level arrests during an unplanned visit I made, in July 2013, to the “Tombs”—the windowless holding pens in the basement of the 100 Centre Street courthouse in Manhattan. I counted four white men out of hundreds of prisoners who were waiting to be arraigned. One was there for allegedly slugging his girlfriend, another for buying cocaine in an upscale night club. The other two were accused of driving while intoxicated. (I was one of the latter; the charges against me were eventually dismissed.)

This was a large summer weekend crowd, men tightly crammed in the cells, agitating for a few inches of bench space. A neatly dressed seventeen-year-old boy had staked out a spot on the floor, where he sat with his head between his knees in what appeared to be a state of silent despair. The single overflowing toilet that served the thirty or forty men in the cell seemed to bring him close to tears.

The boy had made the mistake of asking a rider who was exiting a subway station to swipe him through with her MetroCard. “I was thirty-three cents short for a single fare,” he told me. He neither jumped the turnstile nor harassed the woman, who obligingly swiped him through. A policeman witnessed the exchange, arrested the boy, and let the woman off with a stern warning, though what law she had broken is unclear. The policeman now had cause to search the kid and found the remnants of a joint in his pocket—crumbs of pot. Though he had no prior arrests, he was now facing two charges: marijuana possession and theft of services, a class A misdemeanor punishable by up to a year in jail. He wouldn’t do time, most likely, beyond his current incarceration, but he feared, with good reason, that the financial aid a college in Pennsylvania had granted him for his freshman year would be rescinded.

There was the sense among some of the prisoners that they were living a permanently restricted existence, one that shot down their on-again, off-again efforts to gain a foothold in “the respectable world,” as one of them called it. There were, of course, lifetime lawbreakers in the Tombs (minor drug dealers, perennial street fighters, career panhandlers and thieves) but a notably large number faced nothing more than loitering and trespassing charges—among the most vague and discretionary charges available to police, and typical of the broken windows focus on small-bore violations. “The cost of hanging out,” one young man called it. If he’d had the money to hang out at a restaurant or club, the police probably would have left him alone, he said. As he saw it, his arrest turned being broke into a crime.


A long-standing argument in favor of these kinds of aggressive shakedowns is that they lead police to criminals on the run from more serious charges. But in the cases I’ve mentioned, and in several others I encountered that weekend in the courthouse, there was no record of a criminal past.

These instances of injustice aren’t the only effects of the policy, and attempts to condemn it completely run aground on some facts. A reasonable way to assess the broken windows style of policing is to ask: What have been its benefits to New Yorkers over the past twenty years? And if some have benefited, do they include the poor who bear the brunt of violent crime? According to the NYPD, there were 333 murders in New York City last year, a drop of more than 85 percent from 1990 when 2,262 New Yorkers fell victim to homicide. Armed robberies are down 81 percent from 1990, rape 55 percent, burglary 82 percent, auto theft 92 percent.

It’s worth noting that these numbers come from reports that the NYPD periodically supplies to the FBI, and in light of what is known about the pressure from top brass to keep the official number of felonies down, they are probably not entirely accurate. Eli Silverman of John Jay College and John Eterno, a former NYPD captain, interviewed hundreds of NYPD supervisors during the Bloomberg administration and found “unethical manipulation of statistics” to be common practice.2 In 2008, Adrian Schoolcraft, an officer at the 81st Precinct in Bedford-Stuyvesant, Brooklyn, tape-recorded roll calls at the station house that revealed threats from precinct bosses if cops didn’t make their arrest quotas. At the same time, they were pressured to downgrade felony charges to misdemeanors.


New York Daily News/Getty Images

Eric Garner during his arrest on Staten Island, right before he was put into the chokehold that killed him, July 2014. His wrist is held by a police officer behind him.

Still, the number of murders would be hard to manipulate. The fact is, 89 percent of murder victims and 86 percent of murder suspects are either Hispanic or black. Impoverished neighborhoods are where the decline in homicides is felt most keenly, and gratefully.

How much of this decline can be attributed to the broken windows policy is impossible to know. A conclusive, long-term study of the policy has never been conducted; it isn’t even clear what such a study would consist of or how accurate one could expect it to be, given the range of potential factors involved. In 2004, James Q. Wilson told an interviewer that broken windows “was a speculation.” “I still to this day do not know if improving order will or will not reduce crime.” But there is plenty of circumstantial evidence that points to the policy as an important factor in reducing crime. Kelling has said, “The burden’s on the other side to [show] there is no link between disorderly conditions and serious crime,” and the various rather far-flung attempts by its critics to discount its influence on what in the late 1990s was known as the “New York Miracle” have been neither quantifiable nor convincing.3

Definitive explanations for the ebb and flow of violent crime remain stubbornly elusive. The downward trend over the past two decades has been nationwide, but in no major city has it been as dramatic as in New York. We might expect that in the years of recovery from the recession of 2008, crime would rise. An accelerating march of gentrification has been severely shrinking the city’s stock of affordable housing and leaving tens of thousands homeless. There has been appallingly high unemployment among young black men, and an exponentially rising income gap between New York’s rich and poor. How are we to explain that with all these circumstances, there has not been even an uptick in violent crime? No single theory comes close to telling the whole story.

The paradox is that high-crime neighborhoods make young people vulnerable to violence, but a police policy of zero tolerance in those neighborhoods makes them vulnerable to police arrogance and unnecessary incarceration. They are squeezed from two sides. I know parents in such neighborhoods who keep their kids on domestic lockdown to protect them from both of these forces, a depressing battle for all involved, and a losing one since no one—young or old—can be healthily confined inside a crammed apartment.

Might other police strategies be more effective than random maintenance of order, both in reducing crime and in protecting victims and their families? About half of shootings occur in just eleven of the city’s 123 precincts. The precincts include those in the South Bronx (which in 2010 had the highest percentage of its residents living below the federal poverty line in the United States), East New York and Brownsville in Brooklyn, and parts of Rockaway in Queens.

New Yorkers who neither visit nor live in these and a handful of other neighborhoods experience none of the intimidation on the street that Kelling and Wilson described in 1982. They are the misery centers of “the other New York” that Mayor de Blasio spoke of during his campaign, and they have more in common with small, isolated, post-industrial ghetto cities like Newburgh, Camden, and Troy than with the glamorous metropolis of Manhattan and large swaths of Brooklyn and Queens. The extent of the isolation of these neighborhoods is startling. The police and a handful of emissaries from health and social services are virtually the only outsiders who regularly set foot in them.

The shootings largely involve youth “crews”—the word young people themselves use for local gangs—in public housing developments that are like discrete villages within the city—ungentrifiable settlements of the indigent and underclass. (Though for people with less stable housing, the New York City Housing Authority apartments are a bonanza, however neglected and crime-ridden the developments may be. As of March 17 there were 247,262 families on the NYCHA waiting list, a grim indication of the extent of the city’s housing crisis.) There are more than three hundred youth crews in New York—178 in Brownsville alone—and they have names like Adore My Beauty, Money Comes First, Goons On Deck (GOD), Hood Barbies and Gun Clappin Divas (both all-female crews), Ape Gang, Vicious Villains, and Fresh and In Charge.

Police surveillance of the crews is intense, and not always oppressive; many residents depend upon the police to keep the crews from threatening them. Crews, it should be noted, are different from gangs such as the Bloods and Crips, who engage in organized crime and have extensive economic interests. Crews tend to belong to a single building or block; theirs are local turf wars. Children sometimes have no choice but to identify with a crew in order to be able to walk safely from their apartment to the subway or to school.

A therapist I know who works with juvenile offenders and their families has told me of the nature of policing she’s witnessed since Judge Sheindlin’s ruling on stop and frisk forced the NYPD to develop different tactics. Officers from Operation Crew Cut, as one initiative is called, meet with parents to teach them to recognize hand signals and hat angles that may indicate their children are affiliating with a crew.

Sometimes, police will use the pretext of minor infractions, such as truancy or riding a bicycle on the sidewalk, to lock up crew members on the day of a planned shoot-out—a selective employment of the broken windows tactic that may actually save kids’ lives. Many of these cops are trained in community relations. “I need help, and if the cops are the ones giving it, that’s fine by me,” a mother in the Albany Houses in Brooklyn told me.

It isn’t rare for parents to plead with police to take their children into custody, in order to protect them. In households with domestic abuse, a crime that has not decreased in recent years, police repeatedly check in, paying follow-up visits after an arrest or a complaint, to see how the family is getting on.

Ray Kelly, who started the Crew Cut initiative toward the end of his tenure as commissioner, said, “If I had to point to one reason why the murders and shootings are down, it is this program. And I can tell you that there is a lot of positive feedback from cops.” The remark is as close as he has come, to my knowledge, to questioning the relative effectiveness of stop and frisk, whose main purpose was to confiscate guns; it’s also an indirect acknowledgment of the widespread dissatisfaction among the NYPD’s rank and file during the Bloomberg and Kelly years. Beat cops particularly disliked stop and frisk, and sometimes would write up “ghost 250s”—fake stop-and-frisk reports with no names—in order to meet quotas.

It’s telling that the most progressive black and Hispanic members of the City Council support Bratton’s request for more money to fund detailed anti-crew projects in which police collaborate with parents and some get to know the crew members themselves. This is the kind of order that they and their constituents say they need, not order that comes from random arrests for victimless offenses.

When “Broken Windows” was published, in 1982, tax revenues in New York were shrinking at an alarming rate and the city’s ability to maintain itself was in doubt. In 1980, the population had fallen to 7,071,639, a drop of about 800,000 from ten years earlier and around where the city’s population had been in 1930. Crime by blacks—not the collapse of local manufacturing or the flight of middle-class families to the suburbs—was popularly perceived to be the primary cause.

This racial perception is no less prevalent today. The most comprehensive study to date on the roots of crime found that the central factor in how people perceive the safety of a neighborhood is not disorder or even the presence of boarded-up stores and abandoned buildings, but the number of African-Americans (and to a lesser extent Hispanics) who live there. This perception was true for blacks and whites alike.4 The link is ingrained in the American psyche. When we criticize the police for racial prejudice, we are decrying a condition that is bigger than the police, a prejudice that we may share ourselves.

In their original presentation of the theory, Kelling and Wilson worried that in implementing broken windows, “skin color or national origin or harmless mannerisms [would] become the basis for distinguishing the undesirable from the desirable.” They offered “no wholly satisfactory answer to this important question” beyond the vague hope that police would be trained to understand “the outer limit of their discretionary authority.” But this trust in individual discretion was never realistic. Jennifer Eberhardt, a social psychologist at Stanford, has demonstrated the deeply coded visual association that Americans—and especially policemen—make between race and crime. When she flashed images of faces across a screen, Eberhardt found that her subjects were better able to detect weapons when the faces were black.

With its emphasis on what was seen as unacceptable public behavior—“untended behavior,” “obstreperous” behavior, “strange” and “unpredictable” behavior, in the words of Kelling and Wilson—broken windows seemed to be more of a psychological than a policial theory, a brainchild out of the lab of B.F. Skinner. The image it presented was of beat cops patrolling the streets on foot as feared and respected moral watchmen, ridding neighborhoods of addicts, beggars, obnoxious loiterers, and the mentally ill. It seemed to elevate policing to the status of a social force in its own right, comparable to, if not on par with, education and housing.

At the time “Broken Windows” appeared, reducing crime was generally seen to be beyond the reach of law enforcement and most other government efforts. The primary occupation of police in the 1970s and 1980s was responding to 911 calls reporting crimes and emergencies, and for years the scene in New York and other cities was of shrieking, lit-up squad cars hurtling away from streets boiling with crime toward some unknown destination where, presumably, even worse transgressions were taking place. Police made arrests in less than 3 percent of 911 calls involving serious crimes. Mostly, they appeared to be isolated within their cars or hiding inside the fortress of the station house itself.

When Mayor Rudolph Giuliani appointed Bratton to be his police commissioner in 1994, to implement the theory in New York, he took to calling it “zero tolerance,” a phrase Bratton rejected because it implied “zealotry and a lack of cop discretion.” But zero tolerance was an accurate description of the practice, as arrests for the most minor infractions began to soar.

The broken windows theory was never meant to be the arrest machine that it became in practice. The objective wasn’t law enforcement, but order enforcement. What Kelling and Wilson did not want was for police to be “governed by rules developed to control relations with suspected criminals,” because the police actions they advocated “probably would not withstand a legal challenge”—apparently they were referring to unwarranted searches and roughing up those who resisted. In other words, show them whom the streets belong to and let the niceties of constitutionally protected civil liberties fall by the wayside—and do it on the street itself. But don’t haul them in, if you don’t have to. “I’ve never been long on arrests as an outcome,” Kelling recently told The New York Times. This may not be an especially humane style of policing, but it’s very different from incarceration as a first resort.

In the wake of Eric Garner’s death, Mayor de Blasio has stood fast in his support of Bratton. As a candidate he made clear than he believed “in the core notions of the broken windows theory.” More recently, the mayor said that, no matter their class or color, New Yorkers want their police to respond to “small acts of vandalism and threatening behavior…in a just manner.” He may well be right about what families in poor districts want. But how does a policy that at its inception pushed aside the question of fairness arrive at “a just manner”?

There are signs that de Blasio and Bratton are planning meaningful reforms. Bratton has said that he will stop judging officers by the number of arrests they make, as his predecessor Kelly did—a change that would immediately reduce incarceration and relieve the anxiety officers feel to produce numbers at any cost. After the Eric Garner tragedy, Bratton announced that 35,000 officers would be retrained in proper arrest procedures. He has already expanded Operation Crew Cut; and Kelling, who at seventy-eight remains one of his closest advisers, has floated the idea of having officers team up with social workers as a different way of addressing minor infractions. Whether that happens or not, it indicates an important shift in Bratton’s thinking about the tactic.

At an oversight hearing before the City Council on September 7, Bratton agreed that police have been too aggressive in poor neighborhoods, and he acknowledged the “need for a fundamental shift in the culture of the department.” On September 3, he announced the introduction of a pilot program in which officers on patrol would wear video cameras in five high-crime precincts. The program was ordered by Judge Scheindlin as part of her stop-and-frisk ruling last year, but rather than appealing the ruling, as Kelly and Bloomberg had planned to do—Bloomberg called the cameras “a nightmare”—Bratton appears to be embracing it as “the next wave” of police accountability.

Significant reform will take time. Kelly ran the NYPD for twelve years, the longest tenure of any police commissioner in New York’s history. About half of the current force graduated from the Police Academy while he was commissioner, and many of their attitudes were formed under his regime. Putting emphasis on misdemeanors while soft-peddling arrests for violent crime had become a measure for the career advancement of officers under Kelly; the books became bloated with hundreds of thousands of small-bore offenses. Michael Palladino, head of the Detectives’ Endowment Association, has described the “demoralized” state of his members during the Bloomberg administration. And in March, two months after he became commissioner, Bratton lamented the “awful” morale of the police force: “The public didn’t understand that, politicians didn’t understand it, but [the NYPD] was a very dispirited organization…beat down over several years, beaten up by the political establishment.”

De Blasio is New York’s first liberal mayor in twenty years. Ninety-six percent of black voters cast their ballot for him in November, a higher percentage than Mayor David Dinkins received in 1989. At the same time, voter turnout was the lowest on record—24 percent—making the black vote all the more crucial to his mandate. One of the main reasons for this support was de Blasio’s unequivocal opposition to stop and frisk.

The trust minorities feel for de Blasio was evident at a demonstration on Staten Island on August 23 to protest Eric Garner’s death. The crowd, the rhetoric, the atmosphere, were remarkably tame, despite the echoes of unfolding violence in Ferguson, Missouri. The overriding sense was that of a disparate group of New Yorkers making a reasonable point about overaggressive policing with the expectation that it would be heard. Visible police presence was scant, consisting mostly of Community Affairs officers in their cloth caps and high-ranking officers in white shirts who looked on with the air of spectators at a mildly interesting show. Blue-clad officers in riot gear were stationed out of sight of the protesters, ready to swoop down, if called, but no call was needed. If anything, the demonstration highlighted how far New Yorkers are from the sheer hatred by the police and of the police that we saw in Ferguson.

De Blasio has posed his belief that “the great threat to the city is inequality” as an urgent moral issue. He has staked his progressive ambitions, in part, on the creation of a more just police force. Bratton’s avowed long-term ideal has always been for police work to be conducted as a collaboration with community residents. Yet he is the police chief most identified with linking violent crime to street-level disorder. Both officials have a unique opportunity to define how a big-city police force can operate in an era of low crime. The actions they take will be an important test of the limits, and possibilities, of liberalism in the twenty-first century.