Individuals – Saunders et al. (2016)
Saunders, J., Hunt, P., Hollywood, J. (2016). Predictions put into practice: a quasi-experimental evaluation of Chicago’s predictive policing pilot. Journal of Experimental Criminology, 12(3), 347-371.
Location in the Matrix; Methodological Rigor; Outcome:
Individuals; Focused; Highly proactive; Rigorous; No evidence of effect
What police practice or strategy was examined?
This study examined a predictive policing pilot program in the Chicago Police Department aimed at efficiently targeting limited resources towards specific individuals at high risk of participating in gun violence, either as perpetrators or victims. This program is based on concepts rooted in intelligence-led policing (ILP) and involves: 1) acquiring information; 2) analyzing intelligence/information; 3) reviewing and prioritizing; 4) acting on intelligence and tasking responsible parties with the plan; and 5) evaluating the impact of that action. The prediction model, which the program is centered on, utilizes social networks of previous homicide victims to predict the likelihood of an individual becoming a victim of homicides. Based on the model, 426 highest-risk individuals for gun violence were selected and placed on a Strategic Subjects List (SSL). The list of the individuals was then referred to local police commanders who managed the districts in which the SSL individuals lived. The commanders did not receive specific guidance on what interventions to apply to their SSL members, but were given authority to tailor interventions to these individuals.
How was the intervention evaluated?
At the individual level, the authors used propensity score techniques to match SSL-targeted individuals who were given the treatment with those who did not receive the treatment but who were also identified as high-risk individuals. These groups were then compared on the likelihood of being involved with gun violence (e.g., murder victim, shooting victim, arrest for murder, arrest for a shooting). At the city level, interrupted time-series analysis was used to estimate the effects of the predictive policing pilot program on homicide trends. Lastly, a mediation analysis was used to examine whether the effect of SSL membership on the likelihood of being involved with gun violence was mediated by additional contact with police.
What were the key findings?
Individuals on the SSL and receiving targeted attention were neither more nor less likely to become a victim of a shooting or homicide than individuals in the comparison group. However, individuals in the treatment group were more likely to be arrested for a shooting. The authors argue using mediation and other analysis, that this finding did not suggest the treatment increased shootings, but that the increased likelihood of arrest may be due to the treatment individuals being under greater surveillance. The interrupted time-series analysis also indicated that overall, the program did not have a significant city-level impact on homicide trends in Chicago.
What were the implications for law enforcement?
This study does not indicate that the predictive policing approach used here helped to reduce gun violence in Chicago.
Where can I find more information about this intervention, similar types of intervention, or related studies?