Just in Reach Pay for Success: Impact Evaluation and Cost Analysis of a Permanent Supportive Housing Program

Short abstract

This study examines a Los Angeles County initiative to provide long-term housing and supportive services as an alternative to jail for individuals with a history of homelessness and chronic behavioral or physical health conditions.

Abstract

Discharging individuals from jails and prisons who may be poorly equipped for independent living—such as those with a history of chronic health conditions, including serious mental illness—is likely to reinforce a pattern of homelessness and recidivism. Permanent supportive housing (PSH)—which combines a long-term housing subsidy with supportive services—has been proposed as a mechanism to intervene directly on this relationship between housing and health.

In Los Angeles County, jail has become a default housing and services provider to unhoused individuals with serious mental health issues. In 2017, the county initiated the Just in Reach Pay for Success (JIR PFS) project, which provided PSH as an alternative to jail for individuals with a history of homelessness and chronic behavioral or physical health conditions.

The authors of this study assessed whether the project led to changes in use of several county services, including justice, health, and homeless services. The authors examined changes in county service use, before and after incarceration, by JIR PFS participants and a comparison control group and found that use of jail services was significantly reduced after JIR PFS PSH placement, while the use of mental health and other services increased. The researchers assess that the net cost of the program is highly uncertain but that it may pay for itself in terms of reducing the use of other county services and therefore provide a cost-neutral means of addressing homelessness among individuals with chronic health conditions involved with the justice system in Los Angeles County.

Issue

In 2017, Los Angeles County initiated the Just in Reach Pay for Success (JIR PFS) project, which aimed to provide 300 permanent supportive housing (PSH) placements for individuals deemed eligible for clinical diversion from the county jail between October 2017 and September 2019. Individuals with a history of homelessness and with one or more diagnosable chronic behavioral or physical health conditions were targeted for the program, which was operated by the Los Angeles County Department of Health Services Office of Diversion and Reentry (ODR). The diversion program consisted of a court intervention that placed people in the community rather than jail. This evaluation sought to answer whether PSH led to changes in use of a broad array of county services, including justice, health, and homeless services, and associated costs.

Approach

Using program administrative data and county service data, including use and cost estimates, we conducted regression analysis with propensity score matching to examine changes in county service use and associated costs for 336 JIR PFS participants. In particular, we examined service use and costs during a baseline period (the 12 months prior to jail release) and a period after JIR PFS program enrollment (the 12 months following PSH placement), relative to a comparison group of 672 individuals. Comparison group members were drawn from a pool of individuals determined to meet JIR PFS program criteria (a history of jail time, homelessness, a diagnosable chronic condition or behavioral health condition) but who were not enrolled in JIR PFS and did not receive subsidized permanent housing during the same period. Comparison group members were matched to the treatment group based on a number of observable characteristics, such as demographics and county service use in the pre-release period. We report the JIR PFS program expenditures, in terms of the housing support and intensive case management service costs, in addition to the changes in other observed county service costs to provide an overall picture of the intervention's benefits and costs to Los Angeles County.

While sample sizes are too small to meaningfully analyze causal program effects by race/ethnicity, we do examine housing stability over time across racial/ethnic groups to provide a high-level view of how program experiences differed by race/ethnicity. We also examine the subgroup of participants successfully housed for 12 months to determine whether outcomes were substantially different among the group that achieved this milestone. Also, among participants who had up to two years of data post–housing placement, we examined their two-year service use and associated costs relative to the comparison group. Finally, we examine the influence of the coronavirus disease 2019 (COVID-19) pandemic on the study findings.

Key Findings

Relative to a comparison group, in the first year, the JIR PFS program was associated with

• a 24-day reduction in jail days, with an estimated cost savings of $16,891 per participant

• a 125-day increase in probation supervision, with an estimated cost increase of $672 per participant

• reduced use of homeless services, including a 22-percentage point decrease in the portion of individuals receiving temporary housing, with an estimated cost decrease of $1,643 per participant

• reduced inpatient service use as observed by

  • a 2.4-day decrease in care provided by the Department of Mental Health Services, for an estimated cost savings of $1,275 per participant

  • a 0.8-day decrease in care provided by the Department of Health Services, for an estimated cost savings of $3,308 per participant

  • a 0.3-visit decrease in emergency care provided by the Department of Health Services, resulting in an average savings of $691 per participant

• increased outpatient service use as observed by

  • a 21-visit increase in care provided by the Department of Mental Health Services, for an estimated cost increase of $2,512 per participant.

Next, we examined whether the JIR PFS PSH program costs—that is, the housing support and intensive case management services costs—were partially or fully offset by changes in other county service costs. We found that net program costs were an estimated $6,202 per participant per year, with a standard error of $4,966 and a 95 percent confidence interval of –$3,543, $15,947. This suggests that roughly 42 percent to 113 percent of PSH costs in the first year post-housing were offset by cost reductions in other areas. In particular, program costs associated with PSH and more frequent use of mental health outpatient and probation services post-housing were mostly (if not completely) offset by savings in other service areas—most notably, less county jail time, fewer inpatient days for mental health and physical health care, fewer emergency department visits, and less temporary housing use relative to the comparison group.

We also examined whether the results were different among the JIR PFS participants who were housed for at least 12 months (82 percent of the sample) and found that the results were relatively similar, except that the reductions in jail days increased to 32, with an estimated cost savings of $21,486. After taking into account the JIR PFS PSH housing costs, the annualized cost of services did not differ significantly between the treatment and comparison groups.

Among participants for whom we had two years of data post-housing (58 percent of the sample), we found similar patterns to the first year post-housing in terms of increases and reductions in service use and associated costs during the second year post-housing, although there was a lesser reduction in jail days that was not found to be statistically significant (12 days, with an estimated cost of $9,507). The overall net cost to the county, after taking into account the JIR PFS PSH housing costs, trended upward but was not statistically significant ($11,535 per participant) and does not rule out that the program was cost-neutral in year two.

Finally, we examined whether race/ethnicity was related to PSH retention rates among program participants. An analysis of time housed in the first year post-housing found a statistically significant difference between retention among Black and Hispanic participants, with Hispanic participants having on average 21 fewer days in housing due to program exits than Black participants. Separately, we also modeled the risk of program exit across the entire observed window; that analysis found no statistically significant difference in exit risk by race/ethnicity. We did find, however, that a higher proportion of White participants remained housed compared with the other racial/ethnic groups across the entire observed period.

Study Limitations

We acknowledge that there are certain limitations to our study. It is important to note that our results account only for county service use that is reflected in the aggregated county service dataset managed by the Los Angeles County's Chief Information Office (CIO) and shared with us for this project. We recognize that there may be other services that individuals access that are not accounted for in this approach, including, for example, court costs, law enforcement costs related to the field-based activities (e.g., arrests), and health and homeless services received from non-county-funded entities or not available in the dataset provided to us (e.g., pharmacy costs). Also, our study used an observational propensity matched comparison group, therefore it is possible that unmeasured, unobserved, or mismeasured characteristics may have influenced the study results. It is also important to note that, by design, the JIR PFS program starts at entry into PSH; the transitional period between jail release and entry into PSH falls outside of the scope of the JIR PFS initiative. This transitional period includes interim housing and related supports, which we describe in this study in order to provide a more holistic view of the broader ODR housing program in which the JIR PFS program operated, but we do not include this transition period in our primary outcome analysis as it technically falls outside the scope of the JIR PFS initiative. We also did not have information to evaluate how representative the JIR PFS program participants were to the broader ODR supportive housing population.

Study Implications, Recommendations, and Conclusions

We examined the association between JIR PFS PSH program enrollment and change in county services use. This is the first study using a quasi-experimental design with a propensity matched comparison group to examine the impact of the JIR PFS initiative. These findings shed light on the potential impact the JIR PFS program had on county service use and associated costs.

We found that use of the primary service type associated with the program, namely jail services, was significantly reduced after JIR PFS PSH placement, both in absolute terms and relative to the comparison group. Similar to other PFS PSH initiatives conducted elsewhere, the PSH intervention was associated with less reliance on crisis services, such as emergency shelter use and reductions in inpatient stays, demonstrating a broad impact of the JIR PFS PSH program on a wide range of services. Moreover, our study, like other PFS PSH studies, observed increases in the use of outpatient services, particularly mental health services, consistent with the goal of “Housing First” approaches to engage participants in ongoing, needed care that is not crisis-oriented or requires inpatient/residential stays. After taking into account JIR PFS PSH program costs, the net cost of the program is highly uncertain but may be cost-neutral: The program may pay for itself in regard to reducing the use of other county services and therefore provide a feasible alternative, from a cost perspective, for addressing homelessness among individuals with chronic health conditions involved with the justice system in Los Angeles County.

The goal of this evaluation was to understand changes in county service use and associated costs in relation to the JIR PFS PSH program, but the financial outcomes associated with the program should not be the only driver of policy decisions regarding the program's value. It is inappropriate to apply a standard that PSH should result in cost savings when other social and health interventions do not adhere to the same standard. Notably, PSH was primarily designed to increase housing stability, not derive cost savings. The one-year housing stability rate among the analytic sample was 82 percent, in line with rates reported elsewhere among similar programs.

While not all county service costs are likely accounted for in our analyses, policymakers may consider the significant savings associated with PSH, primarily to the Los Angeles County jail and secondarily to the inpatient and emergency health care systems. The study results also suggest the need for outpatient mental health services and probation services among the served population; if the JIR PFS program were to continue to expand, those services may also need to grow. However, the need for outpatient mental health care and probation services may decrease over time as participants remain stably housed in the community, as evidenced by the service rates in the second year after PSH entry. The program also may help reduce the strain on homeless services, including both the use of emergency shelter and temporary forms of housing.