NYPD Tapes Update: Kelly’s Crime Stats Panel Finally Releases Its Report, Things Not All Rosy for the Commish


A committee formed by New York City Police Department Commissioner Ray Kelly to look into the accuracy of the city’s crime numbers finally released its report yesterday in an abruptly called press conference at police headquarters.

The committee was created in January 2011, and the report was initially supposed to be complete in six months. But over the ensuing two years and six months, until yesterday afternoon, not a word emerged about it, and Kelly was accused of delaying the report’s release on purpose.

Despite that cynical perception, the finished product is filled with interesting insights into the NYPD’s crime reporting, a process which may seem esoteric, but is actually the bedrock on which the city’s perception of public safety is built.

The document was written by former federal prosecutors David Kelley and Sharon McCarthy. Panel member Robert Morvillo passed away in December 2011.

While City Hall has long touted the decline in crime, questions about the accuracy of crime statistics have been raised for many years in published reports, including the Village Voice‘s 2010 “NYPD Tapes” series, which is cited several times in the committee report. But the department has always been able to fend off outside monitors.

The NYPD has called long called manipulation minimal. Kelley and McCarthy credit the efforts of the NYPD in the large declines in crime over the past 20 years and say the department does much right in preventing crime report manipulation.

But they also suggest that it is more widespread than the department admits. “A close review of the NYPD’s statistics and analysis demonstrate that the misclassification of reports may have an appreciable effort on certain reported crime rates.”

The authors add, “Patterns of misclassifying support the anecdotal evidence, including that certain types of incidents, may be downgraded as a matter of practice in some precincts.”

While the Police Department has been loath to allow outside auditors in, the authors of the report make external review their most important recommendation. “External review is indeed important and may help ease the tension between the NYPD’s need for accurate and timely crime reporting, and the pressure to report crime decreases.”

In a statement, Kelly credited the report’s authors and offered a tribute to Morvillo’s sense of public service. “While we are proud of the extraordinary efforts to which the NYPD goes to ensure the integrity of the process, no system is flawless or immune from weaknesses,” he said. “There is always room for improvement.

“The committee found many strengths in the robust auditing program we’ve developed over the last ten years or so. It was particularly heartening to learn that the committee was impressed by the quality assurance division and data integrity unit officers’ professional commitment and abilities.”

He acknowledged: “They also found a number of areas in which our efforts could be strengthened, and they produced sound recommendations to address them.”
Kelly said the agency will adopt tighter auditing standards, improve training, and certify its auditors. He did not explicitly say he would bring in any outside auditors.

Meanwhile, Molloy College criminologist John Eterno, a former NYPD captain who has written a book on crime report manipulation, pointed out that his work with colleague Eli Silverman found that manipulation increased sharply during the Kelly/Bloomberg era. The report, he said, underscored the need for an inspector general outside the department.

“The bottom line is that there’s a lot of fodder here that says the crime rates may be affected,” Eterno said. “Had this been done by a panel not so favorable to the department, the results may have been much worse. Here, we’re seeing a mixed bag.”

Among other committee findings:

  • Noted that Comp Stat, the NYPD’s signature crime fighting innovation, creates a tension throughout the department “between those officials desire to see continue crime reduction and their need for accurate statistics,” which can lead to manipulation.
  • Confirmed an unofficial but “widespread” practice of police supervisors looking over cops’ shoulders as they are taking reports. “This affords supervisors an opportunity to significantly alter the outcome of the classification at a point where the controls are among the weaker,” the authors write.
  • A crime analysis sergeant admitted to doing her own research to value stolen property, even though police are supposed to accept the victim’s value. The report also cites larcenies downgraded to lost property when complainants didn’t see their property stolen, and iPhones ripped out of people’s hands classified as larcenies.
  • Heard from a precinct cop who told them sometimes in order to make the following year statistically easier, crime reports would be upgraded so as not to show too much of a decline.
  • A detective admitted manipulating reports not on orders but because he thought he was supposed to do so.
  • Just 54 police officers since 2002 have been disciplined for crime manipulation, a number, the panel says, suggests the investigations are too narrow, and the department oversight overlooks a lot of other cases. “There did not seem to be substantial accounability as a result of QAD audits,” the report notes.
  • Fairly recently, the NYPD began auditing SPRINT, or computerized reports of radio runs. The committee found examples of crime victims being interviewed and officers failing to file a complaint. “The SPRINT audits resulted in higher rate of disciplinary action with investigations of 71 commands in 2011 and 57 commands in 2012 and sanctions against 173 officers,” the committee said. “SPRINT audits may be the best means to identify downgrading and suppression and ought to be expanded.”
  • The committee reports notes that internal police audits found 1,368 additional felony assaults in 2010, in addition to the 17,054 reported — meaning that felony assaults might have been 8 percent higher than what was reported in 2010.

Though the ranks of crime numbers all look very neat in the charts produced by the city, the authors spend a lot of time describing the deeply subjective nature of how crime reports are actually taken. From patrol officer to sergeant to desk lieutenant, up through the ranks, crime reports are subject to repeated second-guessing, which naturally can lead to manipulation.

It’s a reminder that New Yorkers should take them with a big grain of salt when crime stats are touted by politicians, police, and the media. Both the Bloomberg and Giuliani administrations compared the city’s number to the FBI numbers in other cities, claiming “New York is the safest big city.” (How many times have we heard that one?)

“The focus of the NYPD and the general public on declines in crime can serve to undermine the integrity of the statistics and compromise Comp-Stat as an effective law enforcement tool,” they write.

The authors go on to offer a mild rebuke. “This intense public focus is a product of the politicization of declines in crime that are growing ever smaller,” they write, adding, “Slight movements in the amount of reported crime are not sufficiently reliable to be a measure of whether crime is actually rising or falling.”

And finally here’s an interesting remark, given the administration’s resistance on this issue: “During the committee’s interviews, NYPD personnel indicated that recent press coverage of crime reporting issues had raised awareness and served as a deterrent to misconduct.”

A footnote: Even though the committee’s report cites Village Voice articles half a dozen times, Deputy Commissioner Paul Browne, Kelly’s top aide and the department spokesman, barred the Voice reporter who wrote those articles from attending the press conference.

The announcement of the press conference came shortly after the Voice filed a Freedom of Information Law request requesting documents related to the crime stats committee. In denying the request, the NYPD had claimed in a letter that it could find “no records” responsive to the Voice‘s request — a response which seems implausible.