Neighborhood – Bilach et al. (2020)

Study Reference:

Bilach, T. J., Roche, S. P., & Wawro, G. J. (2020). The effects of the summer All Out Foot Patrol Initiative in New York City: A difference-in-differences approach. Journal of Experimental Criminology, 1-36.

Location in the Matrix; Methodological Rigor; Outcome:

Neighborhood; General; Proactive; Moderately Rigorous; Mixed effects

What police practice or strategy was examined?

This study evaluated the “Summer All Out” initiative, a 90-day saturation foot patrol program implemented in select precincts across New York City during the summer of both 2014 and 2015. The intervention was intended to augment the visible display of officers on foot within a subset of precincts that were selected specifically for their high count of total reported shooting incidents. The intervention utilized approximately 300 additional officers drawn from administrative positions to provide participating precincts with an average of 20 supplemental officers. This equated to 4-5 additional officers per square mile on visible foot patrol during deployment hours. These additional officers were instructed to maintain visibility but were not encouraged to initiate formal interactions with members of the public. The first iteration of the initiative ran from July 7, 2014, through October 4, 2014, while the second iteration ran from June 6, 2015, through September 5, 2015. Each iteration targeted 10 precincts, mostly in Brooklyn and the Bronx, and 6 precincts were selected for both the 2014 and 2015 iterations of the initiative. The goal was to assess the effectiveness of the SAO initiative in reducing crime and gun violence.

How was the intervention evaluated?

The authors compared the impacts of the “Summer All Out” initiative on monthly crime rates using a quasi-experimental difference-in-differences approach. These analyses compared changes in monthly crime rates and monthly shooting counts for precincts that received the SAO intervention to precincts that did not receive the intervention. For the 2014 iteration of the initiative, the authors compared 6 months of pre-treatment data with 3 months of post-treatment data. Evaluation of the 2015 iteration of the initiative compared 5 months of pre-treatment data with 3 months of post-treatment data. The authors also evaluated the full impacts of the initiative using a 5-year panel of crime data from 2012-2016. Across all analyses, the authors compared the pre-treatment and post-treatment periods using measures of “street-level” offenses (given that the foot patrols in visible public spaces would most likely impact these types of offenses) such as robbery, assault, auto larceny, weapon/drug offenses, misdemeanor/petty offenses, and burglary, as well as a separate measure for shooting incidents that were aggregated by month and converted into rates based on jurisdictional populations.

What were the key findings?

In the initial models comparing the pre-treatment crime data to post-treatment, the authors did not find any appreciable impact of the 2014 initiative on street-level crime rates aside from a significant, but small, reduction in per capita property crime. Results for the 2015 iteration were similarly weak and insignificant, with no significant impacts on any of the crime outcomes. Using the full 5-year panel data, the authors also report finding a significant but small reduction in property crime, but no significant impacts on violent crime or any additional forms of property crime.

What were the implications for law enforcement?

The authors suggest that a heightened foot patrol presence may cause residents to become resentful, and thus unwilling to share information with the police or even report victimization. Additionally, this formal presence increases the frequency of contact between citizens and officers, perhaps resulting in a net increase in crimes reported to the police. The authors suggest that future efforts should not focus solely on increasing the presence of officers, but that the nature of their behavior in specific places matters for crime prevention. For example, similar programs may be more effective if they specify certain crime prevention strategies for officers to engage in while deployed across different targeted areas. They also indicate that the sheer size of the precincts they were studying may have played a role in limiting the effects of the intervention, emphasizing that it may be more beneficial to focus on smaller geographic units, like hot spots, in the future.

Where can I find more information about this intervention, similar types of intervention, or related studies?