FYI.

This story is over 5 years old.

News

Places, Not People, Are the Focus of This New Crime-Fighting Data Analysis Tool

Risk Terrain Modeling seeks to minimize crime by identifying places where it is likely to occur and encouraging officers to look at environmental features rather than people found in the area.
Photo by Lauren Pollet

Atlantic City, the once-booming casino resort town on New Jersey's coast that is now steeped in poverty, has the second-highest rate of violent crime in the state, though authorities have been working to improve the situation in recent years.

Beginning this month, the local police force and prosecutor's office will team with researchers at the School of Criminal Justice at Rutgers University-Newark to test an analysis tool that helps officers target high-crime areas by looking at environmental factors that might lead to illegal activities.

Advertisement

The forecasting tool, which is called Risk Terrain Modeling (RTM), offers police a more community-friendly data instrument than the New York Police Department's controversial CompStat program, according to Joel Caplan, who designed RTM at Rutgers along with Dr. Leslie Kennedy and then placed the tool and information online, for free, for police departments to experiment with.

"These days there are a lot of conversations about police and police interactions with the community," Caplan said. "RTM quantifies high-risk places but explains not just where police should go but also what to do there, which is not just focus on people, but focus on the attractive qualities of landscapes."

He cited the example of a dark alleyway as a way to think about environments that attract wrongdoing: poor lighting increases the risk of crime because a victim can't see a perpetrator, and being in alleyway increases risk because there are fewer ways for a victim to get away. The combination of these factors compounds the risk.

In Atlantic City, both the police department and the county prosecutor have expressed support for the experiment, and will take over the data collection and analysis after an initial trial period with Rutgers. It is the first time Caplan is launching an RTM experiment with an eye toward training a local police department to be able to implement their own analysis and strategy going forward.

Advertisement

Related: Predictive Policing Is Not Like 'Minority Report' — It's Worse

Atlantic City police will focus on problematic areas of the city in order to try and understand why crimes keep occurring there, rather than targeting specific individuals or groups in the city, said Police Chief Henry White Jr.

"Risk Terrain Modeling has all the benefits of CompStat without the reputation of an aggressive policing system causing concern within the community," he said in a statement. "This is primarily because, unlike CompStat, RTM is geographically-based, as opposed to being people-based. In other words, RTM focuses on the geographical characteristics that attract criminals to hotspots, rather than the people who happen to be inside a hotspot."

John Eterno, a retired New York City police captain who is now a criminal justice professor at Molloy College, has written multiple books on the effects of CompStat in police departments across the country. He said that CompStat's main issue is the pressure it puts on police to rack up numbers of arrests and summonses, which can lead to violations of individual rights, rather than develop strategies to eliminate crime and build ties with the community.

"The real basic problem is it becomes a numbers game," he remarked. "What went by the wayside are what I think are more important to policing, which are community partnerships, relations with community, and getting to know all the main players in your command."

Advertisement

Police officers on patrol along Atlantic City's boardwalk. (Photo by Micha? Ko?odziejski)

Caplan's background is in spatial dynamics, which he said prompted him to try and understand crime using that perspective rather than looking at individuals who might be suspected of committing crimes.

"Crime is not random," he said. "Whether activities are legal or illegal, human activity occurs somewhere on the earth's surface, in space, and the spatial dynamics of crime are very important to understanding crime."

CompStat's "hotspot" feature is also a spatial tool that helps identify places where crime is likely to occur. But too often, Caplan said, police respond to a place and then focus on the individuals who are there, treating them all as potential suspects, instead of focusing on the variables in that place that might attract illegal activity and trying to change or respond to them.

Related: How Some Police Chiefs Are Like LeBron James — and Why That's Bad for Cities

Eterno described how CompStat's influence spread from New York to cities around the country, which he said had led to many police departments — including those in Newark, Chicago, and Baltimore — being investigated for violating residents' rights.

"If you just focus on crime control, crime may go down, but if you're not respecting people's rights, it would be like a police state," he said. "That's the big downside to CompStat, when you don't balance it with due processes and community relations."

Caplan has worked with more than 10 departments since 2009 to help them experiment with reducing certain types of crime by focusing on the environments in which criminal acts regularly occur and working with locals to adjust features that potentiate lawbreaking.

Advertisement

In Glendale, Arizona, law enforcement was trying to reduce thefts in the city. By analyzing theft reports using the RTM model, they found that many of the thefts were of cellphones, and were occurring near convenience stores that had kiosks in them that allowed individuals to deposit cellphones and other devices for instant cash back. Police then worked with the stores to move the kiosks closer to the counter and to make the windows of the store more transparent so police could see if a suspect was using the kiosk after a report of a theft was made. The strategy reduced crime by 42 percent over the three months of the experiment without significantly increasing arrests, Caplan said.

Caplan has also worked with Newark's police department to identify ways to reduce gun violence. RTM analysis showed 11 features of the city's landscape that attracted particular types of crime — including clusters of drug markets, foreclosed properties, and liquor stores — and local law enforcement decided to focus on a handful of those features to test their response strategy.

Newark police looked closely at gas stations and fast-food takeout spots, stepping up interactions between officers and managers to better understand the factors that were abetting nearby gun violence and then addressing them. Newark saw a 35 percent reduction in gun violence in five months, Caplan said.

"The premise here is that most people are good people, most people are not offenders, and the response to crime by police and other stakeholders needs to be less of a focus on crime specifically than managing the risk presented by certain places, without assuming the people are the risk," he said.

The pilot program with Atlantic City is expected to last a year, beginning with the collection of crime reports starting this week and ending when Caplan hands over the reigns for future analysis and strategy to the department in 2016.

Follow Colleen Curry on Twitter: @currycolleen Photo via Flickr