Campus & Community

Tracking the trends of criminal activity

2 min read

Study questions long-range effectiveness of short-term crime strategies

Short-term downturns in criminal activity do not necessarily result in sustained crime reductions. That is a primary finding in a new research report co-authored by Kennedy School Assistant Professor Brian Jacob.

The report, titled “The Dynamics of Criminal Behavior: Evidence From Weather Shocks,” draws upon weekly crime data from the FBI’s National Incident-Based Reporting System. Jacob and co-authors Lars Lefgren and Enrico Moretti found that short-term reductions in crime in one week were often followed by crime increases in subsequent weeks. The result of these reversals was that the total reduction in crime was often substantially mitigated.


Read the study as a working paper


This study indicates consistent patterns of displacement, in which criminals respond to the relative cost or benefits of their actions at a certain time.

“While our study did not examine law enforcement programs in particular, these findings may have important practical policy implications for local police and other law enforcement agencies that design crimefighting strategies,” Jacob said. “In effect this study shows that crime delayed is not necessarily crime prevented.”

The authors examined suspected correlations between crime trends and weather, relying on the knowledge that heat waves often lead to an increase in crime in the short run while severe rain or snowstorms generally decrease crime. They found that crime was systematically lower than normal in weeks following periods with hotter temperatures (during which crime typically increased).

Conversely, they found that crime was systematically higher in weeks following periods with considerable precipitation (during which crime typically decreased).

Specifically, the authors found that a 10 percent decrease in violent crime due to a weather shock increases violent criminal activity by about 2.6 percent the following week. Accordingly, about 25 percent of the “benefit” of the crime reduction is eliminated within one week. The evidence indicates even larger displacement over longer time frames.

“These issues are particularly relevant today given the tendency by some departments to target neighborhood ‘hot spots’ for short periods of time. The problem with this approach is that the crime is simply displaced to a different time or location,” Jacob concluded.

Jacob is assistant professor of public policy at Harvard’s Kennedy School of Government. His research interests include labor economics, program evaluation, and the economics of education.