A recent survey of retired New York City police officers suggests the department's culture has shifted toward data manipulation.
You can’t fault rank-and-file members of the NYPD for lacking a sense of drama. Over the past two weeks, the city’s police officers have turned their backs on New York Mayor Bill de Blasio at the funerals of two cops shot and killed in the line of duty. They’ve heckled him at the Madison Square ceremony where new officers graduated from the police academy.
The most stunning action, however, has been the department’s unprecedented work slowdown, which started after those officers were killed by a disturbed gunman as they sat in their patrol car on a Brooklyn street. According to the latest figures in the New York Times, in the week from December 28 to January 4, New York City cops wrote just 347 criminal summonses, as compared to 4,077 over the same week a year ago, and the number of arrests has similarly been cut in half. “Most precincts’ weekly tallies for criminal infractions,” according to the paper of record, “were close to zero.”
For a department that over the last generation has built a national and international reputation on its zeal for data and numbers, this sudden drop in arrests and summonses is a shocking about-face. It has even caused some observers to wonder whether the lack of chaos resulting from the action undermines the rationale for the city's overall approach to policing.
Since 1994, when current commissioner Bill Bratton took the job the first time around under get-tough-on-crime mayor Rudy Giuliani, the NYPD has been all about moving numbers—arrests and summonses going up, and crime going down. Starting back then, the department committed to a proactive policing strategy based on the “broken windows” philosophy and data-management tools such as CompStat. That strategy, the department regularly trumpets, led to an 80 percent reduction in crime between 1990 and today. The low number of crimes, the reasoning goes, is the direct result of the high numbers of summonses and arrests for minor violations.
Critics of broken windows say that it has led to racial profiling, civil-rights violations, and alienation within the community. The NYPD’s defense has been simple: it works. Commissioner Bratton and Mayor de Blasio both reaffirmed their support for the approach just a couple of days ago.
But has crime really gone down as much as the NYPD claims? And what has the data-driven approach done to the character of the police department over the last generation?
These are the questions behind an important study, “Police Manipulations of Crime Reporting: Insiders’ Revelations,” [PDF], recently published in Justice Quarterly that raises sharp concerns about the way this “proactive” policing approach has shaped the department’s relationship with New York’s citizens over the past generation.
The study is based on responses from 1,770 retired New York City police officers to an anonymous online survey about how real-life crimes are turned into publicly available statistics. Respondents were divided into three separate groups: those who retired between 1981 and 1993, before “broken windows”; those who retired during the advent of that policy under Giuliani, from 1994-2001; and those who retired under the data-centric, stop-and-frisk Bloomberg administration, from 2002 to 2012.
Among the first cohort, nearly 70 percent answered “No” to the question, “Based upon your experience do you have personal knowledge of any instance in which crime reports were changed to make crime numbers look better than they were?” In the second group, 65.5 percent answered “No.” But in the Bloomberg era, the majority of respondents—55.5 percent—answered “Yes.”
“We were floored. The extent of it blew our minds,” says Eli Silverman, one of the researchers. Professor emeritus of the John Jay College of Criminal Justice and Graduate Center at the City University of New York, Silverman has done years of research into the way the NYPD counts crimes. His most recent book is called The Crime Numbers Game: Management by Manipulation.
Silverman and his co-authors, John Eterno and Arvind Verma, allowed the survey subjects to submit anecdotal comments as well. The study quotes one respondent, who retired in 2005, as saying, “A [commanding officer] may ‘suggest’ that a burglary complaint was a criminal trespass.” Another wrote, “In some cases, larcenies became ‘lost property’ and values were skewed so they didn’t become grand larcenies. Also complainants were told they must go to the stationhouse or to the precinct of record to report a crime.”
The retired officers also expressed skepticism about the city’s much-vaunted crime reduction. When asked, “Are you confident that major crimes have declined by 80% in New York since 1990s?”, 58.2 percent answered, “No.” One cop who retired in 2007 submitted this comment: “Crime is a constant. To believe that crime has gone down 80% is a joke. I believe eventually if the Feds get involved the truth will come out.”
Now, it's not exactly surprising that individual cops, who deal every day with dangerous and unpredictable situations, are skeptical about the extent to which crime has declined over the last generation. The fact remains that the streets of New York, like those of many other large cities, are significantly safer than they were 30 years ago is undeniable. Are they 80 percent safer? Objective measurement remains elusive without outside analysis, which the NYPD has resisted.
Even more difficult to tease out is the question of how the change happened. Was it because of increased enforcement, as Bratton and de Blasio insist? Is it because of economic factors? Or is the cause something more unexpected, like the elimination of leaded gasoline and paint, as some researchers have suggested?
Regardless, Silverman says, cops on the street feel they bear the responsibility for pleasing the top brass by driving reported crimes down yet further. So he sees data manipulation, which has been alleged by many whistleblowers over the years, as not surprising given the difficulty of pushing numbers ever lower. The initial reductions may have been relatively easy to achieve, but the downward trajectory can’t go on forever. “The analogy I use, it’s like squeezing an orange,” he says. “When you first squeeze it, the juice flows freely. Then the more and more you try to extract the juice from it, the more difficult it becomes.”
The problem in crime measurement is not confined to New York. Recent investigations by the Los Angeles Times into the way crimes are classified by the LAPD led a department official to admit that a significant number of incidents had been downgraded in error.
Silverman points out that the NYPD has long resisted any outside scrutiny of its figures or independent analysis of its methodology. Academia, he suggests, has gone along for the ride, even as the "New York model" has been uncritically emulated around the globe. “A couple of people have talked about this phenomenon, but it in no way pervades the mainstream,” he says. “It hasn’t even pervaded the mainstream in criminology. Because if you think about it, if criminologists were to acknowledge that what we are suggesting is becoming increasingly the case, a lot of their work is based on official statistics. And if you’re challenging that, their foundation may be challenged.”
NYPD spokespeople did not respond to a request for comment on Silverman's findings.
Silverman wonders about the unhealthy long-term effect that an emphasis on numbers has on the way cops police the streets. In their paper, he and his co-authors cite social psychologist Donald Campbell, who wrote this in 1976: “The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.”
In New York, says Silverman, stats have been elevated into the holy grail of policing. That, he hypothesizes, has profoundly changed police behavior. “We talk about the pressures,” says Silverman. “We don’t talk about racist cops or anything like that. We talk about how the system compels cops to act in a certain way. The pressure on these guys to keep the numbers going down is incredible. Now no political leader or police leader can acknowledge, sometimes it’s going up.”
Bill de Blasio was elected mayor in part because he articulated New Yorkers’ frustration with the excesses of data-driven stop-and-frisk policing. The greatest crisis of his administration so far has resulted from the fallout from broken-windows policing in the Eric Garner case, and the resulting perception among rank-and-file NYPD officers that they don’t have his support. The long cycle of events that have gotten the city to this place may in fact have its roots in the single-minded chase for another set of numbers: ever-falling crime rates.
“I think there is a circular connection,” says Silverman. “If you just define policing in terms of curbing crime and don’t talk about constitutionality, if you don’t talk about professionalism, if you don’t talk about due process, if that’s not on your radar, it’s not going to be attended to. If you become single focus, you become myopic. That’s part of the problem. Unless it’s a broad approach and the agencies are working well together, and you can have confidence in the agencies, then you’re going to lose the people. And if you lose the community, then that circle is regenerated again.”