Op-Ed: Why NYPD’s ‘Predictive Policing’ Should Scare You

Emerging from weeks of mass demonstrations and the killing of two cops, Mayor Bill de Blasio and NYPD Commissioner Bill Bratton showcased city crime stats for 2014 this month to a room full of journalists eager to prod the pair after a year of controversy. While touting the low crime figures, Bratton also spoke to the future:

“2015 will be one of the most significant years in the history of this organization. It will be the year of technology, in which we literally will give to every member of this department technology that would’ve been unheard of even a few years ago.”

While 2014 ended with historic protests and a political soap opera involving police unions, Bratton reminded everyone what he had been saying consistently during the first year of his second stint at One Police Plaza: He would be ushering the police department into a modern era of policing—one based on “predictive policing.”

While the mention of “predictive policing” might send chills down the backs of civil libertarians, no one can say they were blindsided. Bratton spoke very openly and candidly about it in an interview with the New York Times early last year. He also mentioned it to a law enforcement audience last Spring while sitting alongside military leaders on a panel discussing “21st Century Leadership” at the 92nd street Y. It was there that he reiterated his plans to get hi-tech tablets and smartphones into the hands of every cop on the force and into every squad car.

Rubber hit the road last October as Bratton, de Blasio and Manhattan District Attorney Cyrus Vance announced that the tablets would be paid for by asset forfeiture funds obtained by the City and the DA’s office.

“Christmas has come early”, Bratton declared.

Predictive policing has its roots in the work of Bratton and Manhattan Institute fellow George Kelling. Bratton, known widely as the champion of Broken Windows policing (the focus on low-level crimes to prevent more serious crime), also co-founded the COMPSTAT system at the NYPD in the 90’s. COMPSTAT became a numbers-crunching performance tool that was widely regarded as having revolutionized policing in New York, and then abroad. Its focus on crime stats, combined with the buffet-style policing approach that Broken Windows promotes, led to the oft-reported (and always officially denied) NYPD quota system. It also, by mapping crime stats, allowed authorities to aggressively police communities of color by saying they “put cops where the crime is.” These were the building blocks of a decades-long obsession with prevention down at 1 Police Plaza.

It was more recently, in an influential 2008 Oxford University paper titled “Police Performance Management in Practice: Taking COMPSTAT to the next level,” that Bratton’s devotion to preventing crime through low-level arrests made the leap towards prediction. Bratton, then head of the LAPD, forecast “predictive methods to create even more timely and successful intervention and crime reduction initiatives.”

“We will move from near real-time analysis to true real-time analysis and then to a ‘predictive policing’ posture wherein more accurate and reliable probability modeling will be utilized to forecast potential crime trends over an increasing time span.”

But Bratton’s role vis-a-vis predictive policing wasn’t just as a highly influential advocate. He had, in fact, laid much of the groundwork in Los Angeles from the moment he arrived. Shortly before his return as head of the NYPD was to be announced, the SF Weekly published a lengthy expose of a little-known California technology firm called PredPol that shed some light:

Interest in predictive policing spiked nationally in 2009 as the National Institute of Justice, the research and policy branch of the Department of Justice, published a series of white papers and doled out millions in grant money to seven police departments to undertake the task.

One of the grants went to the Los Angeles Police Department in 2009. When LAPD applied for the grant two years earlier, it was still under the leadership of Bill Bratton, who had championed CompStat’s introduction while serving as NYPD commissioner from 1994 to 1996. Bratton wanted LAPD to be a crime-fighting laboratory. He assigned then-Lt. Sean Malinowski, a former Fulbright scholar who had studied counterterrorism at the Egyptian National Police Academy in Cairo, to be the lead investigator on LAPD’s predictive policing grant.

Around the same time, researchers … were using grants from the Army, Air Force, and Navy to develop a series of algorithms based on earthquake prediction to forecast battlefield casualties and insurgent activities in overseas war zones. Army Research Office documents reveal that the work of anthropology professor Jeffery Brantingham, math professor Andrea Bertozzi, and math postdoc George Mohler was repurposed from its initial application of tracking insurgents and forecasting casualties in Iraq to analyze and predict urban crime patterns. This research would lead to the creation of PredPol.

In Los Angeles, LAPD Chief Charlie Beck, taking over after Bratton left, expanded on the theme in a 2009 piece for Police Chief magazine titled “Predictive Policing: What Can We Learn from Wal-Mart and Amazon about Fighting Crime in a Recession?” Beck explained that “intelligence-led policing” was largely a result of 9/11: “Homeland security increasingly has become hometown security.” And, as the title suggested, corporate America and law enforcement were melding their work together towards this end like never before. Beck, now working with PredPol, echoed Bratton’s assertions in that Oxford paper that “partnerships” with “business communities” could create “models of performance metrics” for police departments. Police leaders of large urban police departments were now turning to the corporate world for lessons on increased “efficiency and effectiveness” by forecasting trends.

Perhaps no other police department was as ambitious with predictive policing as Chicago’s police department, though. Chicago PD, recipient of over $2 million in NIJ funding, had developed a “heat list” of 400 of the city’s “most dangerous” residents based on predictive policing algorithms. One of the leaders of Chicago’s efforts was a professor and researcher who’d done this sort of work “since the 1980s when he worked with the U.S. military to recognize potential targets in the battlefield.” Like the UCLA researches did with the LAPD, Chicago was bringing key technological aspects of the war home.

Recently, they’ve come home to the city where COMPSTAT was born. Last year in Harlem, the city’s largest-ever gang raid resulted in 103 indictments stemming from two murders. The raid was buoyed by Operation Crew Cut, the NYPD program where social media interactions play a significant role in determining guilt and building cases oftentimes by mere association. NYPD detectives and intelligence analysts had monitored dozens of public housing residents for years, including the collection of more than a million Facebook posts, leading up to the military-style raid. Not to be outdone, District Attorney Vance played clean up in the sweeps by weaving together complex conspiracy charges that forced dozens of young men to plead to sentences up to 15 years for things they had alleged to have conspired to do—and in most cases had not actually yet done.

The raid framed the question that dogs predictive policing: In an effort get out ahead of crime were we locking people up for breaking the law—or for the future dangers they posed?

During the official “Predictive Policing Symposiums” hosted by the NIJ in 2009 and 2010, law enforcement bigs exhorted the value of getting “buy-in” from policymakers. They explained that predictive policing wasn’t something fundamentally new, rather that it would enhance current practices. This is most likely true. Cops who profile are in their own way predicting crime. But if you’re inclined to view profiling-based policing as moving in the wrong direction, these technological enhancements would seem to only grease the wheels of profiling and police departments hellbent on ever lower crime—at virtually any cost.

As Ingrid Burrington wrote after Bratton’s return to New York was announced, the legacy of Bratton and the turn towards predictive policing is also “about data—how governments think about data, how they use (and misuse) data… The techno-utopic dream of using data to identify crimes before they happen requires a tremendously low opinion of human beings and an unrelenting faith in algorithms.” This criticism, of course, echoes the national debate on data collection. But its being done by local police will mean increased face-to-face interactions between cops and the areas criminality is being predicted in. This poses a set of risks beyond mere privacy intrusions.

Predictably, social media would provide a growing harvest of information for tech firms and police departments looking towards the predictive era. The Economist boasted that “Firms that once specialized in helping executives measure how web users feel about their brands now supply products that warn police when civil unrest approaches … Cops in California admit to trawling social networks for early warnings of wild parties. ECM Universe, an American firm, offers software that crawls sites ‘rife with extremism’ to identify people who deserve closer attention.”

It’s hard to imagine a more Orwellian future for high-poverty, high-crime communities than one in which technology intensifies the criminal justice spotlight on their neighborhoods. Funneling more cops and surveillance into certain neighborhoods all but assures that you’ll find more crime there—especially the low-level type via Broken Windows. Indeed, if by being predictive police are looking to be prophetic, then it may be a self-fulfilling prophecy. Communities of color are most familiar with a cycle in which cops aggressively look for crime, arrest and document crime, then set out to aggressively look for more crime in a city with less and less crime. Technology, though, adds in a new element wherein it normalizes aggressive, racialized policing through a veneer of color-blind efficiency. As people become data plots and probability scores, law enforcement officials and politicians alike can point and say technology is void of the racist, profiling bias of humans.

As Forbes put it:

“This combination of predictive tools that direct officers into certain neighborhoods backed by an increasing number of channels to surveil those who live in areas predetermined to be criminal is not new, but an extension of a pathology of law enforcement that assumes a criminal pretext by its mere presence in a community. It’s not predictive policing that makes this phenomenon possible, but it takes it one step further toward being totally irrefutable, secured by computer objectivity that further distances people from any direct power within the structures around them.”

The stage seems to be set for this phenomenon in New York. PredPol’s website explains their technology can be “accessed securely through computers or mobile devices in the field.” Those new NYPD tablets will not only vacuum information, like fingerprints, into NYPD databases; they’ll also be linked to the NYPD’s Domain Awareness System (DAS), a joint project between the police department and Microsoft that provides access to 3,000 surveillance cameras, license-plate readers and untold amounts of crime stats and 911 call data. Gunshot-detection sensors that are slated to blanket communities of color for years to come have been contracted to SpotShotter, a firm connected to telecommunications giant Motorola. Bratton was a shareholder and board member for both. Even police-worn body cameras, ostensibly a reform, have the full-throated support of Bratton because of their value for surveillance—not accountability.

Perhaps most alarming, as law professor Andrew Guthrie Ferguson has pointed out, is the potential for predictive policing to “change the balance of suspicion” in a way that could affect the Fourth Amendment’s protections and essentially lower the reasonable suspicion bar in favor of the police. A computer algorithm that could serve the role of the anonymous tip in the eyes of the law could replace even the need for an actual human tip or complaint as a justification for a street stop. Entire neighborhoods fall under official, tech-sanctioned suspicion.

While New York city activists and reformers have fought for years the policies and ideas (Stop and Frisk, Broken Windows) that were the footprints of Bratton’s first tenure, it’s clear that any number of innovations going back to the 90’s still influence people’s lives today. Broken Windows has seen low-level crime arrests and summonses skyrocket over the years and is still the modus operandi of today. The fatal police interaction that took Eric Garner’s life last year is part of that legacy.

Still, predictive policing may cast the longest shadow. At 67 years old, Bratton didn’t return to New York to keep the seat warm. He’s back on American policing’s biggest stage to steer the department towards a policing era that could make Robocop and Minority Report seem prophetic.