the nypd’s compstat at 25

From nearly the moment CompStat was introduced in 1994, the New York City Police Department’s cutting-edge crime tracking and performance management system became an instrument of unparalleled influence in law enforcement.

CompStat’s implementation here coincided with an unprecedented drop in crime that saw rates fall by half by the end of the decade — a downward trend that continues with the historic low reported crime totals of today. As the NYPD evolved from the broken windows strategy of the 1990s to the modern neighborhood policing model, CompStat remained as an ever-present tool for judging police effectiveness.

CompStat (shorthand for “Compare Statistics”) was hailed as one of the key policing innovations of the era and adopted by departments around the country and world, with varying levels of success. But with its success came critiques that the system’s emphasis on numbers had helped produce unreliable crime data and unfair enforcement tactics.

So what is CompStat, and how has it shaped policing in New York City over the last quarter century?

At its core, CompStat is a management tool used by the NYPD to identify problems and measure the efficacy of its crime fighting strategies.

Central to the approach is the computerized real-time collection and mapping of crime and enforcement data across the city. But contrary to a common misconception, CompStat is more than statistical analysis software. Regular CompStat meetings at police headquarters bring together precinct commanders, bureau chiefs and other top executives to evaluate crime reduction strategies and results.

No More Silos The objective is to leverage the vast quantities of information gathered under the CompStat umbrella to effectively target police resources and ensure smooth information flow within the massive bureaucracy of the NYPD.

“The NYPD is a behemoth organization, and we have these specialized units that have different levers they can pull to address different crimes,” said Deputy Inspector Timothy Malin, the commanding officer of the 20th Precinct on the Upper West Side. “CompStat is the mechanism that ensures everybody with a stake in the game, who has the potential to affect violence, is doing what they need to be doing and communicating with each other.”

John Eterno, a professor at Molloy College and former NYPD captain who has studied CompStat extensively, said the system’s role as a conduit for information is one of its main positive features.

“It helps defeat what we call the silo effect, where each of the bureaus in the department were operating separately and distinctly,” Eterno said. “CompStat meetings helped bring together all of the various bureaus and even other agencies which were totally working in the blind from one another before CompStat began.”

Priorities and Best PracticesAmong CompStat’s calling cards is an aggressive focus on reducing seven major index crimes — murder, rape, robbery, felony assault, burglary, grand larceny and grand larceny auto. Commanding officers are held strictly accountable for results within their precincts at famously intense CompStat meetings, where they must answer to their superiors.

“You’re speaking in front of a room in front of all your peers and you’re being questioned directly by the highest-ranking people in the shop,” Malin said. “It’s upper management talking to middle management in front of a room loaded with peers assessing the quality of investigations and noting deficiencies and telling you to improve.”

Though the experience can at times be unpleasant for commanders with unsatisfactory results or responses, CompStat meetings engender a clear sense throughout the department of priorities and best practices.

“The reason we’re all in the room is so everyone knows what the top brass expects, and then you take that back to your command and apply it to your own issues,” Malin said.

A Crime Prevention ToolThough some of its innovations may now seem like common sense, when it was introduced 25 years ago CompStat represented a paradigm shift within the context of a broader effort to reorient policing toward crime prevention.

“The idea of calling the precinct commander and saying, ‘This is your precinct, these are the crimes that are happening, what are you doing about it?’ — it sounds obvious today, but it really hadn’t been done for at least 30 years and perhaps even longer,” said Peter Moskos, an associate professor at John Jay College of Criminal Justice, whose current research is focused on the New York City crime drop of the 1990s.

Before CompStat, crime reports were compiled on a quarterly basis and statistics were already months out-of-date by the time they reached commanders’ desks. In contrast, he information gathered under CompStat could be put to use immediately.

Crucially, CompStat was not itself intrinsically linked to any particular policing philosophy or tactic, and gave commanders latitude to address problems and shoulder responsibility for the results.

“CompStat did not provide solutions,” Moskos said. “It just said, ‘You’re smart people. You’re executives here. Figure it out.’ And those that did rose in the ranks and survived and those that didn’t got pushed back.”

Negative PressuresBut according to some observers, CompStat’s relentless focus on producing ever lower crime figures has also produced unintended side-effects.

“There’s so much pressure on commanders to show up at CompStat meetings and say crime is going down,” Eterno said. “The unsaid message is that otherwise your career is going to be stunted.”

Eterno’s research with Professor Eli Silverman of John Jay College has focused on crime data manipulation within the NYPD — “juking the stats,” in police parlance — which he says became commonplace as the directive to post consistent crime drops became a political imperative.

“When CompStat started in 1994, a lot of the pressures were absolutely necessary because the department needed a culture shift,” Eterno said. “There was a lack of accountability and a true need to really clamp down on things. But as time went on, the pressure of CompStat continued to get ratcheted up. It never eased.”

“We’re finding that the pressures are too much and that the officers are gaming the system to make themselves look better,” Eterno continued, noting that the phenomenon is present not just in New York but in other CompStat departments throughout the U. S. and the world.

“What’s happening is that the words on the reports are being changed to downgrade what actually is occurring so that it’s not classified as an index crime,” Eterno said. “Instead of taking a report as a burglary, in many precincts they’re taking them as criminal trespass and larceny. In this way, the crime is broken down into its constituent parts and it’s no longer an index crime.”

CompStat also contributed to the rise of stop-and-frisk in the 2000s.

“If crime was up in your precinct they would ask ‘What are you doing about it?’ And for a good five to ten years the answer was ‘stops are up,’” Moskos said.

The very fact that stops were being counted led to their being utilized as an indicator of productivity — a case of the tail wagging the dog that led to excessive use of the tactic and blowback from the public that culminated in a class action civil rights lawsuit and the appointment of an independent monitor in 2013 to oversee reform efforts.

A Case Study ApproachIn recent years, Malin said, CompStat has moved away from its former focus on arrest and stop totals and toward a more case-driven approach.

“It’s different now,” he said. “They really focus on the quality of investigations, as opposed to the amount of enforcement being done.”

CompStat meetings now roughly resemble the case-study learning method used in top business schools. NYPD executives identify important individual cases to highlight — usually (but not always) related to violent crime — then call on precinct commanders to review their actions and thought process to ensure all relevant actors are sharing information and working together toward solutions.

“CompStat is definitely much smarter now than it’s ever been,” Malin said.

The program’s longevity is a testament to its adaptability in the face of evolving departmental philosophies and strategies. “It all comes down to good leadership,” Moskos said. “If you have good leadership it works, and if you don’t it kind of goes haywire. Other cities have tried to do this and CompStat alone isn’t enough, just to say that we’re going to have a weekly meeting and keep track of crime statistics. If you don’t have leadership, it all fails.”