Logo image
Who gets in? A mixed methods analysis of racial disparities and police discretion in a police-assisted diversion program for those with alcohol, drug, and/or mental health issues
Dissertation   Open access

Who gets in? A mixed methods analysis of racial disparities and police discretion in a police-assisted diversion program for those with alcohol, drug, and/or mental health issues

Adam Vose-O'Neal
Doctor of Philosophy (PhD), Brandeis University
2026
DOI:
https://doi.org/10.48617/etd.1497

Abstract

Deflection Police discretion Police-assisted diversion Racial disparities Substance use Mental Health Public Policy
In 2016, Albany, NY, launched a Law Enforcement Assisted Diversion (LEAD) programto increase public health and safety while improving racial equity at the intersection of addiction, mental health disorders, and criminal justice. The program provided police officers discretion to divert individuals suspected of low-level offenses—driven by addiction, mental health issues and/or poverty—away from the criminal-legal system into case management. However, from 2016 through 2021, Black adults comprised 64% of the city’s arrestee population but only 38% of the 289 individuals diverted into LEAD. This mixed-methods dissertation study analyzed data from primary and secondary sources. Semi-structured interviews and participant observation (e.g., police ride-alongs) were used to explore the influence of interpersonal, institutional, and structural racism on LEAD Albany design and implementation and stakeholder perspectives on the program (Aim 1). Using local and statewide administrative data, descriptive analysis and multi-level logistic regression modeling was used to examine the presence of interpersonal and institutional racism in LEAD Albany in the contexts of program eligibility and police discretion (Aim 2). Descriptive analysis and logistic and linear regression were used to conduct a preliminary evaluation to study the effectiveness of LEAD Albany as measured by legal outcomes for LEAD participants (Aim 3). vii The main finding was that from 2016 through 2021, missed opportunities for diversion due to officer discretion were prevalent: there were only 289 diversions out of 2,726 LEAD-eligible incidents. And, after controlling for a host of suspect, incident, officer, and neighborhood characteristics, Black male suspects were nearly 6 percentage points less likely to be diverted (95% CI –10.0 to –1.8, p < .005) than White suspects when comparing decisions made by the same officer. Referral was significantly more likely to be offered to suspects who were female, older, and had prior New York State arrests. Referral was also significantly more likely during the week and in the afternoon or evening. Two officer-level predictors were significant: both fewer years of experience and fewer arrests predicted a higher likelihood of referral. LEAD policymakers saw these missed opportunities as the result of barriers to diversion such as mission confusion, insufficient officer training, and overly restrictive access to the program. They believed the racial disparities to be driven largely by officer bias, but also involving a complex layering of institutional and structural racism. While almost all LEAD clients interviewed found the program life changing, a preliminary outcome evaluation showed LEAD had no effect on re-arrest or re-conviction; however, this evaluation did not adequately control for selection bias, an issue that should be addressed in future research. This study is the first, to the researcher’s knowledge, to examine racial disparities in a LEAD program in-depth, providing insight into the systemic barriers to racial equity in police-assisted diversion. This work offers actionable steps to Albany and other LEAD sites reflecting on racial equity, program eligibility, and referral policies and practices, as well as guidance to jurisdictions exploring or considering implementing a police-assisted diversion program.
pdf
AVO Dissertation_v24.45 MBDownloadView
Open Access

Metrics

1 Record Views

Details

Logo image