Spatial Analysis of Probation in Cook County and Neighborhood-Level Implications

Background and Purpose: Probation is the leading form of correctional control in the United States however relatively little is known about the impact of probation on individual behavior, effectiveness in deterring crime, or its impact on communities and public safety. Existing research on probation outcomes narrowly focuses on individual-level characteristics associated with the likelihood of recidivating or evidence-based probation practices associated with lower rates of negative discharge. The purpose of this paper is twofold: first, to conduct the first spatial analysis of probationers in a large metropolitan area. Second, to examine an individual’s risk of being on probation based on their neighborhood context.

Methods: The current study provides a contextual analysis of probation trends through secondary data analysis examining all closed probation cases (both felony and misdemeanor) in the Adult Probation Department of Cook County, Illinois between 2006 – 2016. The dataset includes individual-level factors (race, gender, age), the addresses of probationers, risk-scores (Level of Service Inventory-Revised) in addition to conditions of probation and probation outcomes. Neighborhood-level factors are drawn from the American Community Survey and US Census dates falling in between 2006 – 2016. Geo-spatial analyses including spatial autocorrelation and Moran’s I were calculated using ArcGIS software. Second, autoregressive cross-lagged (ACL) models were generated using Mplus.

Results: Tests for spatial autocorrelation examine systematic patterns in the form of probation clusters or the dispersion of probation occurrences. Spatial autocorrelation assists in the identification of spatial patterns in the prevalence of probation. The ACL model is a form of structural equation modeling allows for an examination of the relationship between neighborhood characteristics and probation concentration over time, as well as the direction of causal flow between them. The first task will be to evaluate if patterns in probation concentration are stable over time. Subsequently, various ecological factors that could account for both the stability of spatial patterns will be explored as well as variability over time. For each one of these models, the core questions include: a) Does [neighborhood metric] predict concentration of probation over time? b) Does the concentration of probation predict [neighborhood metric] over time? c) Are spatial trends for [neighborhood metric] stable over time?

Conclusion/ Implications: This study is designed to fill critical gaps in knowledge regarding how community context relates to probationers’ recidivism risk, but how use of probation impacts community context, particularly in urban communities with high numbers of individuals serving sentences under probation.  The results of this study will advance knowledge and have important implications for policy and practice by examining the association between community context and probation outcomes over time.