GIS to target crime "hot spots"

By ELLEN GOLDBAUM
News Services Editor

Where does crime pay the most?

Researchers at UB are conducting a project for the National Institutes of Justice that attempts to answer that question by determining which neighborhoods are more likely to be targeted by criminals.

For society at large, of course, crime doesn't pay, but the researchers say criminals not only see a payoff from crime, they expect a certain level of benefit depending on specific factors present in the neighborhoods in which they operate.

The main goal of the UB project is to use geographic-information-science techniques to quickly identify "hot spots" of increasing criminal activity, as well as areas where such activity is on the decline. The ability to develop such tools for the mapping and analysis of crimes-a recent emphasis of researchers at the UB site of the National Center for Geographic Information and Analysis (NCGIA)-will allow law-enforcement agencies and policymakers to target more strategically their crime-fighting efforts.

The NCGIA research team was one of only six groups from institutions around the U.S. chosen to present highlights of its work Tuesday at a briefing for legislators in Washington, D.C. The forum was sponsored by the University Consortium for Geographic Information Science, a national organization dedicated to advancing the understanding of geographic processes. David Mark, professor of geography and director of NCGIA's UB site, is immediate past president of the organization.

Rajan Batta, professor and chair of the Department of Industrial Engineering, and Christopher Rump, assistant professor in the department, presented their work on developing a socioeconomic model that can predict how crime patterns shift in response to changes in police enforcement within a particular jurisdiction. Co-investigator on the project is Peter Rogerson, professor of geography.

The objective is to help determine how best to allocate police resources among adjacent neighborhoods in order to most effectively reduce crime, not simply cause a shift in such activity from one area to another.

According to the researchers, factors that determine whether or not a criminal will target a particular area include a neighborhood's level of wealth, how much competition there is from other criminals, the level of police enforcement and the likelihood of arrest.

"Many crimes, such as burglary, robbery and auto theft, are committed by mobile criminals with economic motivations," said Rump. "Using qualitative observations of this type of criminal behavior, we have developed a socioeconomic quantitative model that attempts to predict the number of crime incidents within a police jurisdiction."

According to this model, the criminals are assumed to compare an expected reward for committing a crime in any specific neighborhood against a minimal acceptable award, which may be seen as what would be the likely benefit for some alternative activity, such as gainful employment.

The model will be used to most effectively allocate police resources in order to minimize the difference in the number of crimes between neighborhoods and the total number of crimes among several neighborhoods.

The researchers base their model on data on auto thefts, burglaries and narcotics crimes collected through collaborations with the Buffalo Police Department and other local law-enforcement agencies.