Following a Monday announcement that crime in New York City is at a "historic low" (down 5.9% for the month of October and 3.1% this year to date) NYPD Commissioner Bill Bratton told reporters that beat cops will soon be equipped with an updated version of CompStat—that data tool, introduced in the 1990s, that digitally maps crime statistics for precinct-by-precinct review.
Using CompStat, NYPD precincts determine where to concentrate their resources, with the aim of sending officers where crime is, statistically, most prevalent.
An updated version of the technology aims to provide this information in real time. "It's really pushing CompStat down to the beat cop," Bratton said.
The NYPD's information technology deputy commissioner Jessica Tisch told the Post that CompStat 2.0 will allow beat cops to see a breakdown of crimes in different sections of their assigned precincts, and show which crimes are "trending" in order to "better focus their patrols." Until now, this information has been recored in notebooks and made available to beat cops on a weekly basis.
CompStat data, informative when considered in full, has also inflamed the tabloid tendency toward scaremongering generalizations, as seen in yesterday's NY Post article that drew sweeping conclusions about overall crime based on just one week's worth of stats.
The numbers this month are, at first glance, promising. Murders are down 12% compared to October 2014, and rape is down 16.8%. Robbery decreased .8%, assaults 5.1%, burglary 14.2%, and shootings are down 11.8%.
But look at the year-to-date statistics, and the picture is less rosy: murder up 7.7%, and rape up 4.2%. Homicides by gun are up 25%.
Squint past these numbers, and there's more information to take into account. Last fall, the John Jay College of Criminal Justice released a report that found a 190.5% increase in misdemeanor arrests—turnstile jumping, public urination, other Broken Windows offenses—between 1980 and 2013. The majority of these arrests were concentrated in low-income and minority neighborhoods which, thanks to CompStat analysis, tend to have more heavily concentrated foot patrol.
"We go where the 911 calls send us, we go where the 311 calls send us, we go where the politicians ask us to go," Bratton said at a press conference last summer, responding to the racial disparity in low-level arrests. "I'm not gonna apologize for serving the needs of New Yorkers who need to be served."
Robert Gangi, Director of the Police Reform Organizing Project, picked at the prospect of a more immediately responsive CompStat. "CompStat, if the use is not abuse, is common sense," he told us this afternoon. "It helps the NYPD allocate resources. If they are using it to focus on real crime, then that's a positive."
"But if they are using it to... become even more aggressive with Broken Windows, that's negative, and basically enables them to be more sophisticated in carrying out a blatantly racist police practice," he added.
The Vera Institute of Justice and the Police Foundation recently announced a new initiative that, they hope, will help police departments respond to community concerns about policing in their neighborhoods, not just the crimes documented. Also dubbed, confusingly, CompStat 2.0, it is currently in a two-year pilot phase.
The goal of the program, according to an October press release from the Institute, is to incorporate data on community members' satisfaction with the police into standard data on crime, using neighborhood surveys.
The leaders of the initiative wrote an op-ed for the Marshall Project describing what they perceive to be the possible benefits of such an approach:
By integrating data from community satisfaction surveys into a Compstat system, precinct commanders can get a real sense of their communities’ public safety priorities and trust in law enforcement, and compare it with data on neighborhood crime patterns. If a neighborhood has both a low rate of reported crime and a low level of citizen satisfaction, commanders may deduce that the crime stats are not reflecting reality on the ground so much as a reluctance to report.
The NYPD says that their own version of CompStat 2.0, the "trending" crimes version, should be available to beat cops in the next three months.