Community Policing: How Do You Know If It’s Working?

POLICE
The NYPD hopes community policing can build trust and reduce crime. Lucas Jackson/ Reuters

New York Police Department Commissioner Bill Bratton and Mayor Bill de Blasio recently described a community policing push to encourage trust between cops and civilians. The push would use meetings with residents and neighborhood forums to create “built-in time to address issues that go beyond the crunch of standard radio calls.”

This concept isn’t new—policing basically was community policing until the radio car came along. After a series of unarmed African-American men died at the hands of police across the country this year, however, law enforcement agencies and lawmakers hope renewed emphasis on these initiatives can ease deep tensions between cops and communities of color.

But how do we know if it’s working?

Community policing aims to foster trust and crime-reduction, with the goal being that the former leads to the latter says Lorie Fridell, co-editor of Policing: An International Journal of Police Strategies & Management and associate professor of criminology at the University of South Florida. One measure of success is how the community feels about the effort.

“That could be measured by a community survey, meaning that random segments of the broader community are selected and given before and after surveys on their attitudes toward police” says Fridell.

“Another type of survey is what I call a ‘consumer survey’—you survey people who have had recent contact with the police as opposed to people within the community generally.”

Data on how much communities interact with police can also indicate the state of relationships.

Departments using ShotSpotter, a gunshot detection system, can compare the number of detected shots with the number of 911 calls related to gunfire. This can provide “an inferential measure about people's willingness to call the police” explains Brian Jackson, director of The RAND Corporation’s Safety and Justice program.

Strict research standards should be used to assess community policing programs. That means using the program in some neighborhoods, and not in others, as a control, and comparing the results. Researchers must examine more than one area, otherwise they cannot accurately assess the program.

Since May, the NYPD’s has “piloted” its neighborhood policing plan in four precincts. The Department and Mayor say the pilots “have demonstrated some promising initial developments,” such as year-over-year crime declines.

The NYPD will compare these areas to several other precincts and communities and make general assessments, but evaluation will mostly be based on community reaction, a department official tells Newsweek.

Figuring out whether any community policing program works, of course, depends on another big variable: officers' participation.

“One of the challenges with community policing interventions: You can do foot patrols but it also matters whether people on the foot patrols are stopping and talking to people,” Jackson says. “You can do community policing in name only, or you could do community policing. For understanding the effects, you need to drill into the details about what's being done and how it's being done to really make sense of the data and get good conclusions out of it.”

Also, departments largely measure their success with data-driven crime analysis, such as the NYPD’s CompStat, not so much hearts-and-minds initiatives.

“That is a big challenge overall—to measure anything else, particularly crime prevention activities, systematically within an agency,” says Susan Shah, Chief of Staff of Vera Institute of Justice.  “Performance evaluations of officers don't often focus on whether or not the officers are skilled and making partnerships and engaging in community-oriented problem solving.”

It’s not impossible, though. Researchers have determined CompStat and similar programs can measure community policing in addition to crime, such as by measuring complaints against officers, use of force, and fear of crime, Shah says.