By Patrick F. Hibbard, Michael R. McCart, Jason E. Chapman, & Ashli J. Sheidow

Introduction. This report describes results from a randomized controlled trial (RCT) of a short-term housing (STH) program for individuals released from prison who are on supervision and without an approved place of residence. Developed by Community Corrections leaders in Oregon, STH aims to provide a brief period of housing to individuals in need following exit from prison, with the goal of easing their reintegration back into the community and reducing their likelihood of recidivism. To control costs, STH emphasizes the provision of housing only and without the other services that are commonly included in more comprehensive reentry programs. STH consists of basic housing (i.e., a hotel room, shared room at a county’s office, or in a home owned by the county) not to exceed thirty (30) days. Housing can be provided to those with felony prison release, as well as those with local control release. STH is a one-time benefit. STH recipients also are provided a one-month voucher for public transportation. Excluded from STH are high-risk individuals (e.g., sex offenders), individuals whose needs would be better met in a recovery housing program (e.g., Oxford House), and those with an active threat to self or others. Recognizing the potential promise of STH, Oregon’s Criminal Justice Commission contracted with the authors of this report to evaluate the model’s impact in a multi-site RCT. The primary target outcome in the RCT was reduced recidivism, defined in accordance with SB 366 Section 1 (2015) (codified in Oregon Revised Statutes [ORS] 423.557). As used in that section, recidivism refers to any arrest, conviction, or incarceration for a new crime within three years of prison release. Study Design. Across three participating counties (“County A,” “County B,” “County C”) in the RCT, the study compared STH to usual community housing (UCH), with randomization at the participant level. Archival arrest, conviction, and incarceration records for approximately 3 years post-randomization were obtained from state databases for 114 RCT participants (County A n = 55; County B n = 8; County C n = 51). Participants were aged 21-65 years (M = 38.7), and 88.6% were male. Race/ethnicity was 84.9% White, 8.7% Latinx, 3.3% Black/African American, and 3.0% Native American. Secondary observational data were obtained for three additional counties (County “D,” County “E,” County “F”) who provided housing to individuals releasing from prison during the same timeframe as the RCT. After pre-processing data for comparability to STH, these data (County D n = 29; County E n = 1; County F n = 54) included individuals aged 18-82 years (M = 36.2), and 86.9% were male. Race/ethnicity for secondary observations was 82.1% White, 7.1% Latinx, 7.1% Black/African American, and 3.6% Native American. STH was expected to yield reduced arrest, conviction, and incarceration recidivism relative to UCH. Data Analysis. For all models, the primary analysis looked at data generated through the RCT. A small sample size, however, may not have produced sufficient power to detect small effect sizes. Thus, supplementary analysis included matched secondary observational data. Arrest, conviction, and incarceration recidivism are compared in 3 ways: The first model tested for a difference in the likelihood of each recidivism outcome. The second tested for a difference in the count of conviction and incarceration. The third tested for a difference in the time to arrest, since count was not available. Each model also included several control variables. Finally, the models included fixed effects for each participating county, as well as the month and year for each individual’s release from prison. When secondary data were added for supplementary analyses, a dichotomous covariate indicated if an observation was from secondary or RCT data sources. Results and Conclusion. STH does not appear to be an effective approach to lowering recidivism. The primary and supplementary analyses returned no evidence of an effect of STH on arrest, conviction, or incarceration recidivism. The failure of this approach might be due to its (a) strict 30-day housing limitation, (b) lack of other services, and/or (c) higher levels of supervision scrutiny. Nevertheless, we caution against concluding from this report that housing is not needed or useful for some individuals releasing from prison. Clearly, housing is a major challenge for a large number of parolees (HUD, 2022), and when counties have the resources to provide a shortterm stay at a hotel or other housing location, doing so seems both compassionate and logical. At the same time, 30 days of housing appears insufficient to meaningfully impact an individual’s reintegration trajectory. Instead, we recommend that Oregon provide its counties (particularly those in lower-resourced areas) with more funding to develop comprehensive reentry programming. Such programming would ideally have the capacity to address multiple reentry challenges (in addition to housing) for longer periods, and in a way that individual needs are addressed along customizable timelines. This recommendation is consistent with prior research showing that programs providing comprehensive services inclusive of housing to individuals releasing from prison tend to be effective at reducing recidivism when implemented with fidelity (Lutze et al., 2014; Miller & Ngugi, 2009).

Salem: Oregon Social Learning Center, 2024. 21p.

By Melissa Serafin, Julie Atella, Piere Washington, Sophak Mom

Reentry and community context Within three years of release from state prisons, federal prisons, and local jails, about half of people are reincarcerated and two-thirds are rearrested (Office of the Assistance Secretary for Planning and Evaluation, n.d.). Existing research highlights how challenges to reentry are compounded in rural communities because they are often underserved and under-resourced. Employment, housing, transportation, and limited social and health services are all common barriers for individuals reentering rural communities, and access to reentry programming is often limited (Benavides et al., 2023; Ward et al., 2016; Ward, 2017; Zajac et al., 2014). Additionally, the opioid crisis has significantly impacted rural communities, including those in Maine (Maine Governor’s Office of Policy Innovation and the Future, 2023). The rate of overdose fatalities in Maine increased from 0.3 per 1,000 population in January 2017 to 0.6 in January 2022, and the rate in Waldo County also increased from 0.2 in 2017 (8 total) to 0.5 in 2022 (20 total; Maine Drug Data Hub, 2024). Overdose fatalities most notably increased during the COVID-19 pandemic: the fatality rate in Maine increased 33% between 2019 and 2020, with 83% of deaths in 2020 due to opioids. Without opportunities, resources, and support, people reentering rural communities may experience hopelessness and criminogenic needs contributing to recidivism (Ward, 2017; Pettus-Davis & Kennedy, 2019). The findings from the current project also identified many of these challenges. Limited access to and availability of resources and services like internet, public transportation, health care and insurance, employment options, and affordable housing across counties and the state were stressed as barriers hindering the success of the program. Housing was identified as a particularly significant challenge, and this is compounded by the lack of viable employment options that provide a livable wage. The median cost for a studio efficiency or one-bedroom apartment in Waldo County, Maine in 2023 was $916-$922, and a two bedroom was $1,174 per month (RentData, 2024). According to the Maine Department of Labor (2024), the minimum wage in 2023 was $13.80 per hour in the state. Individuals employed full time, earning minimum wage, and who rent at least a studio efficiency would then be considered “cost-burdened”1 or “severely cost-burdened”2 (Office of Housing and Urban Development, 2014). Rent burden contributes to lacking the means to afford other necessities such as food, clothing, child care and more. Furthermore, the National Low Income Housing Coalition (NLIHC, 2024) notes the extreme shortage of rentals that are affordable to households whose incomes are at or below the poverty guideline or 30% of their area median income. In addition to the shortage of affordable housing and within the vicinity of the VOANNE network—within a four-city radius—there were no homeless shelters, nor mental health facilities with detox beds, and only one sober house. While VOANNE provides tents and camp supplies in more dire cases, and connects some participants with temporary transitional housing (including motels), they have access to very few permanent housing supports. Experiencing housing insecurity and homelessness can also negatively impact substance use. For example, efforts to stay awake to protect belongings and avoid violence is associated with developing and worsening substance use disorders (Mehtani et al., 2023). Existing research and findings from the current evaluation indicate that transportation is often a significant challenge for individuals reentering rural communities (Benavides et al., 2023; Ward, 2017; Zajac et al., 2014). While Maine does have some public transportation, rural areas have limited public transportation. Some communities have fixed route bus services that do not operate 24 hours a day and seven days a week, but these may only be in some counties (Eichacker, 2021). Limited public transportation means limited independence upon reentry to get to and from places like the DMV, a job interview, or to work. Additionally, some participants in the current evaluation did not possess driver’s licenses, further limiting their transportation options. This evaluation also identified how lack of phone service and high-speed internet make communication difficult and pose a barrier to building independence and integrating into society. Previous research has also identified these as significant reentry barriers (PettusDavis et al., 2019). Correctional system initiatives In response to the challenges that individuals face while reentering their communities, VOANNE and their partners have implemented several initiatives. Specifically, VOANNE has prioritized building trusting and collaborative partnerships with law enforcement agencies; this evaluation identified the strength of these partnerships. These partnerships facilitate opportunities to strengthen impact, such as the ability to provide services and programming within county jails while individuals are incarcerated. Additionally, postrelease offices and community spaces provide critical support to individuals who are reentering their communities after incarceration, and VOA’s diversion programming allows VOANNE to serve anyone who could benefit from services (e.g., not just individuals who are or have been incarcerated). VOANNE has also been able to expand to providing services in additional counties and correctional facilities over time.

St. Paul, MN: Wilder Research, 2024. 43p.