Predictive Policing: The Next Wave of Crimefighting Technology Is Being Tested in NYC
"Predictive policing used to be the future,” NYPD commissioner William Bratton, told me, “and now it is the present.” The NYPD launched a predictive policing pilot program last summer that Bratton says is currently being “field tested.”
So, how does this approach actually work? Is it effective — and sufficiently sensitive to civil rights?
“The evidence suggests that crime is predictable, at least statistically, because criminals follow patterns. They tend to go back to work in the same area... because they’re familiar with the setup — and they got away with it the first time.”
UCLA anthropology professor Jeff Brantingham and George Mohler of Santa Clara University, pioneers in the field, are the founders of a company called PredPol.
PredPol divides a precinct into 500-by-500-foot “boxes.” A computer analyzes past crimes there, what is happening in real time and other inputs. A box thought likely to be a locus of crime becomes a “hot box.”
The police then visit these locations multiple times during a tour, making their presence felt and, by doing so, deterring crime. The cops look at an “intersection of risk factors,” said Megan Yerxa, a crime analyst at the Tacoma, Washington, police department, which fully deployed PredPol in October 2013.
The evidence suggests that crime is predictable, at least statistically, because criminals follow patterns. They tend to go back to work in the same area, the same house or business, because they’re familiar with the setup — and they got away with it the first time. Criminal justice experts are impressed by the potential.
“I got hooked on Moneyball and sabermetrics (Bill James’ revolutionary analytic tool for judging baseball players’ performance),” says criminologist Craig Uchida. At the U.S. Department of Justice, Uchida oversaw grants to develop predictive-policing methods.
Cops have always depended on intuition, Uchida says. Now they can “back up intuition with information — have cops talk with people on the streets to find out what is really happening. Write it down. Turn it into statistics, and then — using new technology that we now have — crunch the shared experience.”
Cops need good data to make it work. But when the data is there, so are the results.
“We can clearly show that with the use of new algorithms within a certain geographic area within a certain time frame we can predict that, absent intercession by the police, a crime is likely to occur,” said LAPD Commander Sean Malinowski, who was implementing PredPol for Bratton when Bratton was chief in Los Angeles.
Other jurisdictions concur.
“In some cases, we were 60% predicting where or when,” says Santa Cruz deputy police chief Steve Clark. “It blew my mind.” Other jurisdictions were seeing increases in crime; Santa Cruz saw reductions.
“In our first six months, we saw a drop of 28% in burglary compared with the same locations and same time in the previous year,” Clark added. They started knocking out “the low-hanging fruit.”
Future reductions flow from already reduced numbers. “If you catch one burglar, he can’t knock over a bunch of houses” later, as Clark put it. And crime continues to drop.
“The PredPol police departments, though — which are the ones that inspired Bratton’s pilot — work differently. PredPol focuses on geography — "place-based policing.”
At what point, however, does protecting society run the danger of infringing excessively on individual rights? And are there reasonable alternatives?
There, the jury is out.
“Predictive policing has limited use,” argues civil rights activist Constance Rice. She finds the name “misleading. It’s not like they have a crystal ball, and they can predict where crime can be.” She’s skeptical that the time and resources devoted to predictive policing are worth it.
According to Chris Calabrese of the American Civil Liberties Union, the main problem with predictive policing is, “If you put bad data in, you get bad results. By bad data, I mean historical inequities.”
The worry is that police will create, as one headline suggested, a “real Minority Report,” referring to the dystopian story and film in which individuals are arrested based on offenses that they would commit in the future.
A number of predictive-policing software and analytics systems have been developed that do tend to focus on individuals. Their use could lead to constitutional or civil rights challenges in the future.
The PredPol police departments, though — which are the ones that inspired Bratton’s pilot — work differently. PredPol focuses on geography — “place-based policing,” as Tacoma assistant police chief Peter Crimmins describes it.
Bratton is sensitive to the need to balance security with freedom. He says “it’s also important to make sure to have good people elected and appointed to office.”
Most important, the use of predictive policing technologies must be transparent — and carefully overseen by vigilant citizens themselves.
This piece originally appeared in New York Daily News, adapted from the Winter 2016 Issue of City Journal
This piece originally appeared in New York Daily News