Implementation and Outcomes of Forensic Housing First Programs

Background and Purpose: Housing First (HF) program models are increasingly being considered for individuals with serious mental illness and criminal justice histories. Providing HF for these individuals, especially when linked to jail diversion strategies, challenges the core HF principles of harm reduction, consumer-driven services, and client self-determination. The degree to which mandated treatment and reporting mandates affects the implementation and outcomes of a HF model is an important consideration that has largely gone unaddressed in the literature. Methods: This mixed-method study used administrative data from 93 supportive housing programs (n=4,780) and qualitative site visit data from a subset of four forensic programs to (a) compare fidelity to the Housing First model and residential client outcomes between forensic and nonforensic programs and (b) investigate whether and how providers working in forensic programs can navigate competing Housing First principles and criminal justice mandates. HF fidelity scores of forensic and nonforensic programs were compared using chi-square tests. A series of generalized estimating equations (GEEs) were used to compare changes in days in residential settings between forensic and nonforensic programs. Qualitative analysis of site visit transcripts was used to understand the context and challenges of providing housing services to forensic clients. Results: Forensic programs were more likely to require participants with substance use disorders to participate in substance use treatment (p<.01) and less likely to follow a harm reduction approach to substance use (p<.05) or to provide lease or occupancy agreements that specify rights and responsibilities without provisions regarding treatment participation, sobriety, or adherence to program rules (p=.01), as is prescribed by the HF model. Days living in congregate settings increased by 68 among forensic clients as compared to 39 days among nonforensic clients (p < .01). Qualitative findings suggested that an interplay of court involvement, limited resources, and risk environments influenced staff decisions regarding housing and treatment. Conclusions and Implications: Study findings indicate that forensic involvement can compromise recovery-oriented care. Forensic programs struggled to implement the principles of Housing First while accommodating the expectations of the criminal justice system. Existing mental health and criminal justice collaborations necessitate adaptation to the Housing First model to accommodate client needs.