NYPD Watchdog Shatters Bratton’s ‘Broken Windows’ — Now What?


When sixteen-year-old Rhamar Perkins escaped police officers who were trying to arrest him for jumping a subway turnstile in Brownsville last week, the NYPD scrambled an all-night manhunt, complete with helicopters. The mobilization was eventually called off when Perkins turned himself in to the precinct four and a half hours later, but the incident — an all-out effort to catch a boy who skipped out on a $2.75 fare — makes for a pretty tidy illustration of what policing in New York City looks like in 2016.

Two days later, the city agency tasked as police watchdog released an 88-page report that landed like a grenade tossed into the swimming pool of New York politics. It challenged not only the logic used to justify the dragnet for Perkins, but also twenty years of policing theory, the reputation of the city’s police commissioner, and the state of the city’s social contract.

The finding at the heart of the report from The Department of Investigation’s Office of the Inspector General for the New York City Police Department is this: Looking at six years of NYPD summons and arrest data, the IG found zero evidence to suggest that arresting and issuing criminal summonses to people for so-called “quality of life” offenses — riding their bikes on the sidewalk, drinking beer in public — leads to a drop in more serious felony crimes. On its face, that conclusion contradicts the central tenet of what has come to be called “broken windows policing”: that by coming down hard on minor offenses, police can set a tone that discourages serious violent crime.

It’s difficult to overstate what a big deal that finding is. Coming down hard on the small stuff has been, in one form or another, the operating philosophy of New York City’s police since the 1990s; in many quarters, it is taken as axiomatic that that strategy was responsible for pulling New York out of its crime-ridden “bad old days” and keeping it from backsliding ever since. Bill Bratton, the current police commissioner, is regarded as the foremost disciple of the broken windows school, having introduced it to the city during his first stint as commissioner, under Rudy Giuliani. When Bill de Blasio needed to telegraph that — progressive campaign rhetoric notwithstanding — he wouldn’t be letting up on aggressive policing, he hired Bratton and declared himself a firm believer in broken windows. (For more on the tortured Bratton–de Blasio relationship, see Ross Barkan’s piece on de Blasio and Bratton’s unnatural symbiosis.)

But even as it has become law-and-order orthodoxy, the theory of broken windows has always been dogged by critics who argue that there’s no empirical basis for its central conclusion; that as practiced it leads to excessive and racially discriminatory arrests and summonses; and that it legitimates a toxic mindset that divides society into honest citizens on the one hand and the disorderly and disreputable on the other. “Order-maintenance policing helps create the category of the disorderly and this, in turn, facilitates the very policy of aggressive arrests for minor disorderly conduct,” Bernard Harcourt, now a law professor at Columbia University, wrote in 1998. “The category triggers an aggressive response, even absent evidence supporting the broken windows theory.”

Eighteen years later, crime in New York is at historical lows and the political leadership is committed — if we are to take them at their word — to bridging the widening gap between the city’s haves and have-nots. In that spirit, and assuming the Inspector General is right that arresting and summonsing people for low-level offenses doesn’t make us safer, the question is inescapable: Should this controversial approach simply be set aside once and for all?


Just what constitutes “broken windows policing” is itself the subject of debate. The term originated in a 1982 Atlantic article by George Kelling, a criminologist, and James Wilson, a political scientist. Kelling had spent some time with Newark police officers on foot patrol, and watched how they maintained order in the neighborhood.

“As Kelly [an officer] saw his job, he was to keep an eye on strangers, and to make certain that the disreputable regulars observed some informal but widely understood rules,” Kelling and Wilson wrote. “Drunks and addicts could sit on the stoops, but could not lie down. …Talking to, bothering, or begging from people waiting at the bus stop was strictly forbidden.”

Kelling and Wilson wedded these Newark observations to an earlier study, from which their article took its name: In 1969, Stanford psychologist Philip Zimbardo left a pair of abandoned cars, hoods up, denuded of their plates, on the street — one on a block in the Bronx, the other in Palo Alto. The Bronx car was stripped almost immediately. The Palo Alto car was left alone — until Zimbardo smashed a window with a sledgehammer. In short order, vandals turned the car upside down and destroyed it too. The lesson, Kelling and Wilson concluded, is that signs of disorder produce a vicious cycle: Neighborhood residents, rightly or wrongly, will associate these signs with a rise in violent crime, and will retreat from the public sphere, paving the way for an actual “criminal invasion.” To prevent this inexorable slide, then, it was the role of the police, like those foot-patrol cops in Newark, to prevent indications of disorder.

Despite its squishy grounding in anecdote and analogy, the article became wildly popular in certain policy circles. The prevailing norm up until then — police patrolling from cars, waiting to respond to 911 calls after a crime had already been committed — had already proved an evident failure. “For me, broken windows was about community policing,” George Kelling told the Voice. “It was about bringing the police and the community closer together.” But as police departments began to put the precepts of broken windows to work on the streets, improved community relations were often not the result. “The extent to which the practice has matched the theory has at times been not so good,” Kelling said.

As New York police commissioner under Giuliani, Bratton identified his own crime-fighting strategy with broken windows. But there were crucial differences between his interpretation and that detailed in the Atlantic piece. Kelling and Wilson had described a style of policing that reflected the individual sense of order of each neighborhood. Bratton’s NYPD imposed a uniform standard of order for the whole city. And where the Newark foot-patrol officers generally asserted their presence without arresting or summonsing people, the NYPD was told to take a much harder line, establishing quotas of low-level arrests each officer must meet. Often conflated with broken windows, this strategy is probably better known by a more precise name: Zero Tolerance Policing. (Bratton denies that arrest and summons quotas are in effect today, but a lawsuit filed this spring by twelve minority officers alleges otherwise.)

Kelling, who has a formal consulting relationship with the NYPD, declined to speak specifically about the Inspector General’s report but told the Voice that where his theory has historically gone wrong is when it becomes equated with summonses and arrests. “That was never what Jim and I had in mind,” he says. “I’m not backing away from summonses or arrests — if people are saying, ‘Screw you, officer,’ that’s a challenge to authority that has to be responded to with forceful action. But you want to be relying on the minimum amount of authority necessary.”

But in New York, the rise of broken windows coincided with the advent of statistically informed policing, particularly in the form of the CompStat program Bratton inaugurated in the 1990s. The more nuanced forms of community interaction Kelling says are at the heart of a true broken windows strategy aren’t always easy to capture in statistics; arrests and summonses are.

“But if we only focus on those, we’re not paying attention to the other goals of community policing,” says Kelling. “Not just to reduce crime, but to ensure that there’s also justice in neighborhoods.”

Policing in New York has fallen short of that goal. Under Mike Bloomberg’s police commissioner, Ray Kelly, the NYPD relied on an especially aggressive form of contact with the community: Every year, police would stop, question, and frisk hundreds of thousands of New Yorkers, often on the weakest of pretenses. In 2011, at the practice’s peak, police carried out more than 685,000 stop-and-frisks. The idea, Kelly repeatedly explained, was that if someone was thinking about carrying a gun, they might think twice if they knew there was a good chance they’d be randomly stopped, questioned, and patted down.

One problem with this method, of course, was that in this country the Constitution protects citizens from unreasonable search and seizure. Another problem was that black and brown New Yorkers were bearing an overwhelming share of these stops — in 2011, 87 percent of those stopped were black or Latino — and the Constitution also guarantees citizens equal protection under the law. What’s more, even in those rare cases — 6 percent — that a police stop did lead to an arrest, nearly half the time it didn’t result in a conviction, and the program started to look like a campaign of racist harassment against basically law-abiding citizens. In 2012, staring down the barrel of a federal class-action suit, the NYPD radically drew down its rate of stop-and-frisks. Despite a drumbeat of doomsaying from stop-and-frisk supporters inside and outside the NYPD, who predicted civic chaos and blood in the streets, crime continued to go down.

By 2013, when de Blasio campaigned for office tilting at the inequities of stop-and-frisk, the policy was essentially a lame duck. But to reassure the constituencies most invested in the sort of order that makes the city safe for business investment and ever-widening waves of gentrification, he brought Bratton back as commissioner. The result: Even as stop-and-frisks declined, they were replaced by the low-level “quality of life” arrests and criminal summonses called for under Bratton’s absolutist interpretation of broken windows.

And yet here, too, the burden falls disproportionately. As last week’s Inspector General’s report found, some precincts had thirty times the rate of quality-of-life arrests per resident that others did, and “the distribution of quality-of-life enforcement activity in New York City was concentrated in precincts with high proportions of black and Hispanic residents.” Defenders of broken windows argue that if the police are focusing on black and Hispanic neighborhoods, it’s because that’s where the crime is. That explanation doesn’t cut it, says the Inspector General’s report: “In many, but not all, instances, the rates of enforcement remained high even after adjusting for crime rates.”

City leaders congratulating themselves for the wind-down of the stop-and-frisk regime might want to reassess, says Harcourt, the Columbia law professor who authored the first critique of the cult of broken windows nearly two decades ago. “You’re replacing one racially biased police technique with another.”

To no one’s surprise, Bratton has been hitting back hard on the report since its release. At a press conference last week he called it “useless” and “fatally flawed,” promising a substantial rebuttal within ninety days. In the meantime, the man hailed for bringing CompStat’s statistics-driven policing to the NYPD waved off the report’s empirical conclusions with vague anecdote. “Every day, you in the media report when my officers make a fare-evasion arrest and find a gun on that individual, make a fare-evasion arrest and find they’re wanted on warrants for murder,” Bratton told the Voice at the press conference, before pulling rank on the Inspector General’s Office: “I’ve been doing this a long time,” he said. “I have a lot more expertise than they have on this issue.”

Bratton also questioned the report’s methodology, noting that it looked at only six years of recent data and did not address a wider time period, including the miraculous turnaround years of the 1990s. But it’s worth noting that the NYPD hasn’t made comparably detailed statistics from that period publicly available; the report calls on the NYPD to “release historical incident-level and geographic data.” Asked by the Voice whether the IG’s office had asked for data going back further, and whether it had received all the data it requested from the NYPD, an IG spokesperson declined to answer.

Sitting next to Bratton at the press conference, de Blasio rubber-stamped his dismissal of the IG’s analysis. “I agree with the commissioner’s assessment of the substance of the report 100 percent,” the mayor said. “The core findings, we don’t see merit in.”

Bratton’s response was to be expected, says Chris Dunn, associate legal director of the New York Civil Liberties Union. “I’m more surprised by the mayor’s, given that he has long professed a lot of concern about overaggressive policing, particularly as it affects poor and minority New Yorkers. I wouldn’t have thought he’d be quite so quick to dismiss it quite so aggressively.”


The IG’s report is hardly the first study to ask whether broken windows reduces serious crime — it’s a question that’s been hotly debated for nearly as long as the theory itself. As a Daily News editorial noted after the report was released, some researchers have found evidence to suggest broken windows does have an effect on crime. Others have found evidence that it does not. The IG’s report will no doubt be subjected to further empirical scrutiny, but it’s a sharp reminder that the NYPD’s organizing dogma isn’t anything like settled science. And as the IG notes, with broken windows, there is the other side of the ledger to consider: “Issuing summonses and making misdemeanor arrests are not cost free. The cost is paid in police time, in an increase in the number of people brought into the criminal justice system and, at times, in a fraying of the relationship between the police and the communities they serve.”

Challenging New York’s broken windows strategy has implications that go far beyond policing to include how our society is ordered. Broken windows — and the practices it gave rise to in New York — functioned as a sort of ideological sleight of hand, conflating the two traditional functions of police: preventing crime and maintaining order. The first mission finds widespread natural support; no one wants to be mugged or murdered. The second goal, however, is more complicated. Notions of order and disorder are subjective and contextual; the version of order sought by the gentrifying developer might not map cleanly to that of the man selling loose cigarettes on the corner.

The prices we are prepared to pay for safety and for order may not be the same either. When the public is told that problems of “disorder” are a slippery slope to violent crime, it is likely to have a far greater tolerance for a solution that involves arresting the homeless guy who’s bothering the homeowner, or launching an all-night manhunt for the kid who hopped a turnstile. But if the Inspector General’s report is correct, and disorder doesn’t lead inevitably to mass muggings, murder, and rape, it raises the question of whether police, courts, and jails are really the most effective solution for addressing “quality of life” issues.

“Saying that there’s no connection between violent serious crime on the one hand and disorder on the other hand is not to say, well, fine then, let’s just let homeless mentally ill people wander around pissing themselves and yelling at people on the street,” says Alex Vitale, a professor of sociology at Brooklyn College and a critic of broken windows. “That’s a problem. That’s a symptom of a society that’s headed down the wrong track. But what the broken windows theory does is turn social problems into policing problems. There’s no reason to assume that the solution to these sorts of problems are best delivered by the police. In fact there’s lots of research to show that police efforts are incredibly counterproductive, and even if they’re not, they produce considerable collateral consequences — criminalizing youth, criminalizing homeless people, et cetera.”

Almost all quality-of-life problems have solutions that don’t rely on the police. Instead of arresting people for public urination, we could build public bathrooms. Instead of arresting the mentally ill homeless, we could build more public housing and a robust mental healthcare infrastructure. The problem, of course, is that addressing social problems in a holistic way, rather than simply criminalizing them, would involve an expensive expansion of government services that is unlikely to find favor with the sorts of people who finance election campaigns.

Moving beyond broken windows, says Harcourt, will require a dramatic paradigm shift. “We’d have to radically revisit the way that we conceptualize harm in our city, to get a different measure of what it is that is hindering urban life,” he says.

The question now is whether the mayor and City Council are prepared to rise to the occasion presented by the Inspector General’s report. “This is not a report that’s coming from the crazy NYCLU,” says the NYCLU’s Dunn. “Given that this philosophy has had such devastating consequences in minority communities, this report should be forcing everybody to re-examine the wisdom of that policy.”