Post-Jail Mental Health Treatment and Recidivism Outcomes for Previously Unhouse Individuals

Background and Purpose: Unhoused individuals are significantly more likely to have interactions with law enforcement and increased risk for arrest and detention (Greenberg & Rosenheck, 2008). Oftentimes, offenses associated with being houseless (i.e. trespassing, etc.) are drivers of criminal/legal involvement (Burton et al., 2018). The unhoused population commonly have cumulative risk profiles (mental health and substance use needs, lack of economic resources, etc.) that complicate their stays in jail and community re-entry. This study asked if the post-jail mental health treatment and recidivism outcomes of previously unhoused individuals differed from housed individuals.

Methods: As part of a larger study between 2019 and 2020, 10 Midwestern County jails in one state collected a mental health screening instrument at jail booking. Then, treatment (Medicaid encounter data) and recidivism (jail records) data were merged with this study cohort (n=3,797). The sample were predominantly white (61.8%,n=2,318) and male (74.3%,n=2,820), with a mean age of 35 years (SD=11.6). A small proportion were unhoused (6.6%,n=198) in the 30 days prior to jail booking. Unhoused individuals were significantly more likely to have co-occurring mental health and substance use issues (34.2%,n=65), compared to the housed population (10.0%,n=200; x²(2)=109.70,p<.001).

To assess treatment outcomes in the year after jail release, two binary logistic regression models were used; The first examining any treatment engagement and the second assessed if the first treatment received after release was a high intensity service (i.e. crisis in-patient). To assess recidivism 1 year after release, 3 models assessed: 1) any recidivism (yes/no), 2) the number of jail stays, and 3) number of days spent in jail. Control variables in these models include race, age, sex, geographic location of the jail (rural/urban/metropolitan), co-occurring mental illness and substance use disorder, prior mental health treatment, prior year jail booking (yes/no), prior year jail stays, and prior year jail days.

Results: Across all models, after controlling for demographic, prior year treatment and recidivism factors, unhoused individuals were significantly more likely than housed individuals to: receive a high-intensity service as their first service post-release (AOR=2.05,p<.001; x²(7)=38.105,p<.001), to recidivate (AOR=1.61,p=.005; x²(6)=213.972,p<.001), have more jail stays (B=.084,p<.001; F(6/464.061)=64.844,p<.001), and spend more days in jail (B=.063,p=.002; F(6/217975)=20.274,p<.001). Post-year treatment engagement was trending significant (AOR=1.43,p=.052,95%CI[.996,2.04]; x²(7)=733.45,p<.001).

Conclusions and Implications: The unhoused population in these jails were significantly more likely to have co-occurring mental health and substance use issues and were more likely to utilize crisis treatment services in the post period. Furthermore, they had more frequent jail stays for longer periods of time, compared to their housed counterparts. The unhoused population finds themselves in a revolving door that leads to arrest and detention because of their housing status (Burton et al., 2018), further complicating their ability to access necessary economic resources. Policy makers and law enforcement agencies should consider evidence-based solutions to support this population with unique needs.