Natasha Frost


Gregory Zimmerman, Glenn Pierce

Date of Award


Date Accepted


Degree Grantor

Northeastern University

Degree Level


Degree Name

Doctor of Philosophy

Department or Academic Unit

School of Criminology and Criminal Justice


criminology, mental health, corrections, inmates, mental illness, programming, risk, segregation


Criminology and Criminal Justice | Social Control, Law, Crime, and Deviance


Since the deinstitutionalization of psychiatric hospitals in the 1950s, American correctional institutions have become the primary sites for managing mental illness nationwide. Mentally ill inmates present many challenges to correctional administrators, including but not limited to the prevention of suicide, delivery of psychiatric medication, and implementation of key programming services. Moreover, each challenge must be met while ensuring the safety and security of staff, inmates, and the community. To this end, administrators use evidence-based practices that promote safe and efficient service delivery throughout detention and minimize the likelihood of re-offending after release. One of the most widely supported management strategies, known as the risk principle, assesses inmates' likelihood of recidivism and prioritizes the highest-risk inmates for the most intensive institutional programs. Unfortunately, there is very little research that has investigated the impact of risk-based correctional management strategies on mentally ill inmate populations.

With data on 866 inmates incarcerated within a Massachusetts house of correction (HOC), this dissertation analyzes how measures of criminogenic risk and mental illness separately and jointly predict key institutional outcomes. Specifically, I use path analyses and logistic regressions to predict mentally ill and non-mentally ill inmates' placements into intensive programming, administrative segregation units, and a pre-release community work program. Results indicate that, in adherence with the risk principle, the highest-risk inmate groups are most likely to be placed into intensive programs. While high-risk mentally ill inmates have at least equal access to intensive programming as their non-mentally ill counterparts, they are much more likely to be administratively segregated. Yet, only criminogenic risk, and not mental illness, achieves statistical significance in the prediction of program placement and segregation. I conclude by discussing the role of mental health needs in risk-based correctional management, ways to measure and define mental illness, and the implications of the findings for correctional theory and practice.

Document Type


Rights Information

copyright 2012

Rights Holder

Garrett A. R. Yursza Warfield

Click button above to open, or right-click to save.