Skip Repetitive Navigation Links

California Department of Corrections and Rehabilitation
It Has Poorly Administered the Integrated Services for Mentally Ill Parolees Program, and With Current Funding Cuts, It Must Find Ways to Transition Parolees to County Services

Report Number: 2020-103

Audit Results

Corrections Has Not Provided Adequate Oversight of Critical Aspects of the Integrated Services Program

Corrections has limited its ability to ensure that its providers are operating the program correctly by not conducting all required provider reviews. Its contracts with its providers spell out a multi‑layered oversight program that includes meeting with its providers, performing on-site reviews, and receiving and reviewing periodic program reports. However, Corrections has performed little of this oversight. Further, it has poorly stewarded the program’s public funds because it has not verified providers’ claims for housing reimbursement for program participants or ensured that claims that exceed the allowed monthly rents are valid before paying them. In addition, we estimate that Corrections could have saved up to $3.7 million each year had all of its providers billed Medi-Cal for mental health care services for program participants. Moreover, in reviewing case files for participants, we found that the files did not always corroborate why individuals qualified for the program.

Corrections Has Performed Little Oversight of Its Providers or the Housing Payments It Makes to Them

Corrections has failed to perform oversight in two categories: frequent and infrequent. The frequent oversight was to consist of quarterly meetings with providers as well as weekly, monthly, quarterly, and annual reports from those providers. However, Corrections has conducted few of the quarterly meetings. In fact, the program manager told us that she has held meetings with staff at only one of the eight program locations with any frequency; these were quarterly conference calls with Telecare–Los Angeles. With regard to the periodic reports, although Corrections has received them from its providers, it has done little with the information. For example, the program manager confirmed that no one has systematically reviewed the quarterly reports from the providers or followed up on their contents. This type of frequent oversight could have provided Corrections with insights into program participation, including what services participants received and their status. For example, Corrections could have identified how many participants had completed or dropped out of the program. Through a systematic review of these reports, Corrections could also have learned about program outcomes, such as how many participants had transitioned to county mental health services and permanent housing.

The infrequent type of program oversight is related to site visits, and here again we found that Corrections’ record for conducting these visits is poor. Beginning in 2009, Corrections notified its providers that it might conduct optional annual reviews of each provider to assess how well the provider is adhering to laws and contract provisions. In sum, Corrections conducted only three sets of reviews in the nine years from 2009 through 2017. These reviews occurred in 2012, 2013, and 2017, and Corrections could not provide evidence that it had reviewed all of the program locations. It reviewed four of eight locations in 2012, two of eight locations in 2013, and five of eight locations in 2017. During the 2012 and 2017 reviews, Corrections generally reviewed records and documents as well as staff qualifications, inspected facilities and housing, and interviewed the providers’ program directors. However, the 2013 reviews were of facilities only; Corrections did not assess case files or personnel files. Across these reviews Corrections found that providers were generally compliant with contract provisions. Beginning with the 2018 contract with its private providers, Corrections is required to perform site visits, and it did so at all locations in 2019. It also performed site visits at Santa Clara and San Francisco counties in 2019, although it was not required to do so.

Corrections also has not exercised enough oversight of its housing reimbursements to providers. It has not verified that the amounts the providers claimed for reimbursement reconcile with the rents the providers paid to landlords. Corrections’ contracts allow the providers to spend up to $1,000 per month per participant for housing. To exceed this amount, the providers have to seek preapproval by submitting a form to Corrections. The providers’ contracts with Corrections also require them to maintain all rental agreements and provide copies on request. The program manager stated that she looks at the providers’ monthly housing invoices for red flags, such as if a provider claims an amount over $1,000 without approval, yet she confirmed that she sometimes pays invoices without ensuring that the providers have submitted requests to pay rents higher than $1,000 per month. She also confirmed that she does not compare the providers’ housing invoices to the rental agreements between the participants and landlords to verify that the providers are invoicing Corrections the correct amounts. Neither the program administrator nor the program manager could show that Corrections had ever conducted this sort of verification. Because housing subsidies account for a significant portion of the providers’ contracts, such verification is important. For example, the contract with Quality Group Homes for the three-year period from July 2018 through June 2021 totals $8.4 million, and housing makes up $3 million, or 36 percent, of the total contract. Corrections’ program administrator offered several reasons that Corrections had not engaged in more program oversight, the most significant being a lack of staff. We analyzed program staffing and found that since 2013, the program has had one full-time position: the program manager. However, as we discuss in the following section, the program likely had funds to justify requesting additional staff, but Corrections did not do that.

Further, had the private program providers billed Medi-Cal for eligible services for program participants, Corrections could have saved millions per year. To bill Medi-Cal, the program providers have to be county-approved Medi-Cal providers. Corrections’ goal has been to help the providers enter agreements with their respective counties to bill Medi-Cal for eligible program services. Doing so would save the State money, since counties can receive reimbursements from the federal government of up to 95 percent of the cost of providing services to Medi-Cal clients. However, the private providers have not billed Medi-Cal. Only one of the providers, Santa Clara County, bills Medi-Cal for services. As such, Santa Clara County saves the program about $6,400 per participant slot each year. Based on that savings, we estimate Corrections may have been able to save the program as much as $3.7 million per year had it successfully helped its private providers bill Medi-Cal for program services.

Corrections has made efforts to assist its providers with billing Medi-Cal, but it has not succeeded at overcoming the barriers to enlisting the counties in its efforts. For example, Corrections tried in 2013 to contract directly with Los Angeles County for the program, but it was unsuccessful because Los Angeles planned to subcontract the program provider role, and that raised concerns over state contracting requirements for subcontracting and competitive bidding. In addition, the program administrator stated that Corrections was willing to cover the county’s administrative expenditures for billing Medi-Cal, but Corrections and Los Angeles County could not agree on which expenditures not covered by Medi-Cal were attributable to the integrated services program. Several of the providers reported to Corrections in 2016 that they were in various stages of applying to become Medi-Cal providers in their respective counties. One of those providers—Turning Point–Kern—told us that negotiations with the county fell through when Corrections and Kern County could not agree on administrative costs. Corrections tried again to contract with Los Angeles County in 2016 but was unsuccessful in negotiating a contract.

Corrections Did Not Use All of the Program’s Funds on Program Services, and It Understaffed the Program

Corrections has used contracted providers to deliver program services to participants, and its contracts with those providers are the program’s largest expenditure. From fiscal years 2014–15 through 2019–20, the program’s budget was between $12.3 million and $15.8 million per year, and as Table 1 shows, Corrections’ tracked contract expenditures ranged from $8.3 million to $12 million. Although these figures demonstrate that Corrections did not spend all of its allotted program funds on these contracted services each year, according to a budget manager, Corrections could not further isolate integrated services program expenditures for our analysis. Thus, neither Corrections nor we could determine whether Corrections had spent the budgeted funds on the program.

Table 1
Program Contract Costs Were Less Than Its Budget
(Dollars in Billions)

BUDGET, CONTRACT COSTS, AND ANALYSIS 2014–15 2015–16 2016–17 2017–18 2018–19 2019–20*
Program budget $13,605 $12,320 $13,135 $12,986 $14,320 $15,805
Contract costs, by service category
Program services NA NA $8,275 $8,513 $8,936 $6,896
Housing NA NA 1,674 2,096 2,786 2,472
Inpatient care NA NA 387 500 334 184
Totals $8,312 $8,339 $10,336 $11,109 $12,056 $9,552
Contract costs as a percentage of the program’s budget 61% 68% 79% 86% 84% 60%
Difference between the program’s budget and contract costs $5,293 $3,981 $2,799 $1,877 $2,264 $6,253

Source: Analysis of enacted budgets, Corrections’ budget calculations, and program expenditure data.

NA = Not available.

* Contract costs for fiscal year 2019–20 include data that Corrections entered into its accounting system as of April 22, 2020.

 Corrections’ data do not break out contract expenditures by category for fiscal years 2014–15 and 2015–16.

We were able to ascertain that Corrections has understaffed the program for a number of years. In 2009, Corrections had four staff positions to complete oversight and administrative tasks for the program. This number agreed with a 2007 Corrections workload analysis for the program. However, in 2013 Corrections reduced the program’s staff to one analyst as part of a reduction in staffing levels across the agency when it implemented realignment—changes to California’s prison system to reduce costs and overcrowding. In a 2018 budget change proposal, Corrections requested and received a number of analyst positions for the parole division, the parent division of the integrated services program. Corrections’ internal analysis reflects that one of the 23 positions requested for the division was specifically for the integrated services program; this would have doubled the number of staff working on the program. However, Corrections ultimately determined another unit had a greater need and allocated the position elsewhere within the division. Therefore, since 2013 Corrections has administered the program with a single analyst. At its start, the program contracted with five providers to serve 300 participants at one time, but by 2019 the providers were serving 615 participants at one time. In addition, Corrections was obligated to conduct oversight, including periodic meetings and site visits. Despite this growth in the program and the increased responsibility for oversight, Corrections did not increase the number of staff administering the program. Corrections’ failure to staff the program properly has contributed to the oversight and management weaknesses we discuss earlier.

Some of the funds Corrections did not spend on contracted services could have supported a request to increase program staffing levels. According to the program administrator, Corrections did not prioritize allocating additional staff to the program because realignment constrained the parole division’s budget. As Table 1 shows, Corrections spent between 61 percent and 86 percent of available funds on contracted services from fiscal years 2014–15 through 2018–19, leaving between $2 million and $5 million in funds each year that were potentially available for other expenditures, such as analysts to perform oversight. The program funding Corrections did not use for contracted services would likely have been sufficient to pay for additional staff. For example, according to Corrections’ cost estimates, one program analyst costs the program about $115,000 each year. Costs for three additional analysts would equal about $345,000 annually, which is still only a small fraction of the program funds Corrections did not use for contracted services. In addition, Corrections can only demonstrate its expenditures specifically on the program’s contracts; it cannot demonstrate how it spent the rest of the program’s funds. Given the amount of funds budgeted to the program beyond those spent on contracts, and the very small size of the program’s staffing, the program likely had resources available to justify a request for additional staff beyond the request it made in 2018.

Corrections Cannot Demonstrate That All Participants Met the Eligibility Requirements

During the program referral process, Corrections is responsible for ensuring that program participants meet all eligibility criteria related to severe mental illness and homelessness. The contracted program providers are responsible during enrollment for ensuring that participants agree to participate in the program voluntarily. Corrections documents its referral decisions on referral forms and maintains case files with supporting information for the individuals it assesses for program eligibility. However, for the 10 referrals we reviewed, which Corrections processed in 2019, the case files did not always corroborate how the individuals had qualified for the program. Because Corrections has failed to adequately document that participants meet eligibility requirements, it is not possible for it or for us to know whether Corrections has accepted into the program only those individuals whom the program is intended to serve. In contrast, we found that the providers obtained written consent from the referred individuals to participate voluntarily in the program.

Corrections’ referral forms and case files did not always clearly document that the individuals Corrections referred to the program met the requirements for a serious mental health disorder. Under state law, parolees or inmates must have a serious mental disorder and exhibit substantial functional impairments or symptoms in order to qualify for the program, among other requirements. We expected to find that the referral form made these criteria clear and reflected how the individual was impaired as justification for how he or she qualified for the program, yet that was not the case. In addition, the supporting information in the case files did not always make clear how the disorder impaired the individual’s day‑to‑day functioning. For example, the referral form for an individual in Santa Clara listed the individual as having depressive disorder and anxiety but stated that the individual had been stable for the previous six months. Neither the referral form nor the information in the case file explained how the stable depression and anxiety was impairing the individual’s functioning.

Corrections also did not always clearly document on the referral form that an individual was homeless or at risk of homelessness. According to state law, inmates are eligible for the program if they are at risk of homelessness and parolees are eligible for the program if they are homeless. State law largely relies on the definitions established in federal law to determine this status. Federal law describes five patterns for risk of homelessness and for homelessness, which range from imminent eviction with no subsequent residence identified to lacking a fixed, regular, and adequate nighttime residence. As with the mental health eligibility criteria, we expected that Corrections would have made clear on the referral form or in the case file documents how each individual met these homelessness requirements. However, eight of the 10 case files we reviewed did not make clear that the individuals met those criteria. For example, five referral forms listed addresses in the Residence boxes on the forms—indicating the individuals had places to live—and three other files contained addresses on some forms and the word “transient” on other forms, making it unclear whether those individuals had housing and were not therefore homeless.

A combination of factors contributed to the concerns we describe above about individuals meeting the mental illness and homelessness requirements to participate in the program, including a reliance on institution-based classifications and the lack of a standard for documenting referral decisions. Two of the five senior psychologists working on the program said that, for those individuals with mental illness they review for qualifying for the program, they rely upon the individuals having designations of EOP (Enhanced Outpatient Program) or Correctional Clinical Case Management Systems (CCCMS). Corrections assigns these designations to individuals in prison depending on their mental health. Case files from the other senior psychologists reflected the same. We believe that the EOP and CCCMS designations should have been only the starting point for determining eligibility for the program; the classifications alone were not sufficient to indicate that an individual met the mental health requirement in state law. Inmates with EOP designations require increased mental health care and are typically in segregated housing to receive this care, whereas those designated as CCCMS require a lower level of mental health care and are typically housed with a prison’s general population. Although individuals with EOP designations could qualify for the program, their mental disorders would have to be severe and persistent, and the case files we reviewed did not make those conclusions clear. Individuals with a CCCMS designation in the case files we reviewed had stable behavior by Corrections’ definition, meaning their symptoms were largely controlled and they did not require a structured, clinical environment; based on that designation alone, they would not qualify for the program.

Corrections also has no standard for the information that it expects its senior psychologists to place in each case file as support for their referral decisions. The program administrator stated that completing the referral form requires the senior psychologists to review at least four Corrections databases to complete the fields on the referral form, such as those concerning disabilities and accommodations, and identifying other programs the parolee may be enrolled in, but he indicated that Corrections does not provide specific instruction on this. We found that the referral case files did not show that each senior psychologist had reviewed all of these databases. Although the senior psychologists stated that they reviewed some clinical case notes on a parolee or inmate to assist with a referral, they did not clearly document which sources they evaluated or how their review of case notes informed their referral decision.

Finally, Corrections does not offer eligibility training for those involved in the referral process, including parole agents and mental health clinicians in parole outpatient clinics who identify program candidates, as well as senior psychologists who are responsible for ensuring that those candidates meet all program eligibility criteria. It is reasonable to assume that a lack of familiarity and understanding of the eligibility requirements contributed to the poor or omitted justifications for how individuals were deemed eligible for the program. Because of the lack of training and the poor documentation in case files, it is not possible for Corrections or for us to determine if Corrections adequately considered whether the individuals it referred to the program met all necessary eligibility requirements.

All individuals referred to the program have to agree voluntarily to participate, and the providers are tasked with obtaining their consent. As the providers enroll individuals in their programs, each has the individuals sign enrollment forms that represent their consent to participate. In our review of case files, we found that they contained signed forms of consent, except when the individuals chose not to enroll.

Corrections Has Not Collected Data to Demonstrate That the Program Has Met the Legislature’s Intended Outcomes

Corrections has also not maintained comprehensive data on the program, which has prevented it from monitoring and evaluating program outcomes. Corrections does collect some information on program participation from program providers and from its various internal data sources. From the providers, Corrections receives regular reports containing the participants’ names, the dates they received services, and the specific services they received. The providers also submit information on participants who have received housing. However, the information the providers submit to Corrections is in a mix of formats instead of a uniform format that Corrections could use to aggregate the data to view trends or to easily identify the participants who received services over a specific period of time. For example, Corrections could not easily identify common reasons why some participants exited the program early. Corrections also does not have aggregated information it could use to analyze the types of housing that program participants have received or how long it took them to receive housing. As a result, Corrections could not determine whether the program has accomplished its goals.

Corrections’ internal data were also inconsistent, hampering its ability to identify whether the program benefited participants. Corrections only recently began tracking program participation in a useful way. In August 2019, Corrections updated its Parolee Automated Tracking System database to accept the dates its senior psychologists referred individuals to the program. Before 2019 that information was stored in different places. According to a computing resources manager at Corrections, senior psychologists noted referral decisions and the dates in free-form case notes on staff interactions with participants. Also, Corrections’ mental health staff entered program referrals in yet another database, and each of the senior psychologists maintained independent tracking spreadsheets of individuals they referred to the program. As we explain later, without centralized, consistent data, Corrections has not been able to analyze important information such as recidivism rates.

Because Corrections lacks comprehensive, consistent data on program participation and services, we also could not reach conclusions about program effectiveness and outcomes. The Joint Legislative Audit Committee (Audit Committee) asked us to identify information about program outcomes, such as the number of participants who received interim interventions or housing and the reasons why some participants dropped out of the program. The Audit Committee also asked us to analyze participants’ Medi‑Cal eligibility and recidivism. A key to performing these analyses was identifying participants by name and by Corrections’ assigned identifier in order to search for participants in the California Department of Justice’s database of arrests and convictions. However, Corrections does not maintain a master participant list; rather, that information is reflected in individual invoices Corrections receives from its providers.

In addition, creating our own dataset identifying participants’ services and housing received and program completion rates would have required manual compilation of this information from each of the providers’ monthly invoices for the past five years. This same problem confronted a 2017 University of California, Los Angeles, research team that undertook a study of the program’s effects on recidivism, albeit to a lesser extent. We believe that the research team manually built its dataset because, although its report states that the team used data from a program‑specific treatment and housing database, according to the Corrections program administrator, Corrections does not maintain such a database. In addition, the research team studied the rate of recidivism only for the one year immediately following the parolees’ release from prison, rather than the standard three-year period following release. The research team’s analysis would have required a much smaller dataset than one we would have needed to create. Because of these data limitations and others we describe in the Scope and Methodology in Appendix A, we did not build a dataset manually.

Corrections also does not gather enough information about housing to understand certain program outcomes, and the providers record housing information inconsistently. As we mention in the Introduction, in 2012 the Legislature made housing services a key part of the program. State law requires the providers to offer services to participants that would help them obtain and maintain housing stability during their parole, services that include locating housing, assisting with move-in costs, and subsidizing rent. Providers also need to help participants develop a plan for remaining housed in the future, both after they complete the program and after they complete their parole. Because housing is a critical program element, we expected Corrections to have gathered a detailed understanding of participants’ housing needs. We also expected it to have defined key housing terms and to collect basic information on housing services. Instead, we found that Corrections collects minimal housing information and does not provide enough guidance to its providers on what and how to report on housing, so even the basic information the providers report to Corrections is inconsistent.

That inconsistent information has hampered Corrections’ ability to determine the effectiveness of the program’s housing services. Varying terms to describe program housing appear in state law or Corrections’ contracts with the providers, including supportive housing, transitional housing, permanent housing, and independent housing. With so many housing categories, it is essential that Corrections give the providers guidance on what the terms mean and how to categorize housing for reporting purposes, but Corrections has not done so. Every month, each provider sends Corrections an invoice for housing costs that lists each participant, the cost of his or her housing, and the housing type secured for the participant. However, because Corrections has not provided guidance on the terminology, the providers use different housing categories for the same housing type. For example, Telecare–Los Angeles categorized a sober-living home as “permanent” housing and Turning Point categorized this same type of housing placement as “transitional.” Therefore, Corrections could not be certain in what type of housing the providers had placed participants, and because it did not have consistent classifications, it lacked the necessary, accurate information to evaluate the type of housing where program participants thrived, and thus the program’s effectiveness in reducing and resolving homelessness.

Similarly, other undefined program measures have hampered Corrections’ ability to demonstrate participant progress and the program’s overall effectiveness. Corrections requires the providers to determine each participant’s goals at the beginning of the program and to categorize those goals among a list of what it terms presenting needs, including food, clothing, income, medical and dental services, and shelter, among others. Each month, the providers submit a register of program participants that indicates whether participants have met their identified needs. Corrections could use the “shelter” need to measure whether participants’ housing goals have been met. However, Corrections has not defined the criteria for meeting this need, and consequently, the providers are again submitting inconsistent information. Telecare–San Diego and Turning Point, for instance, both stated that they check that the need has been met when they help fulfill either a short-term or long-term housing need for a participant. Telecare–Los Angeles, though, marks it only if it has fulfilled a participant’s long-term need. This inconsistency among providers’ reports has made it impossible for Corrections to measure the program’s effectiveness related to housing. Although program managers have communicated verbally and through email with the providers over the years on how to report housing information, these communications were in response to questions the providers asked and the program managers did not then issue clarifying guidance to all providers. When combined with turnover in program management at Corrections, inconsistencies in the information persisted.

Corrections Must Take Steps to Mitigate the Risks for Parolees From the Cancellation of the Program

Because its program funding is set to expire, Corrections needs to augment its existing efforts—called prerelease—to transition inmates with severe mental illness and who risk homelessness to available county services as they get ready to start parole. As we discuss in the Introduction, funding for the program will no longer be available after December 2020. Even so, Corrections retains the responsibility to assist people on parole to safely transition back into communities. As Table 2 details, Corrections and county services may be able to replace some of the key services the program provides. However, there are weaknesses and risks associated with these options. For example, Corrections has not conducted any internal or external reviews of how well parole agents perform their functions of identifying parolees’ key needs and attempting to connect them to available services, as specified in Corrections policy. With the program ending, Corrections will now be relying on parole agents in the eight counties that have the program to perform these functions effectively. Additionally, the officials in those counties that we spoke with reported that their existing programs are already at full capacity and they expect to face upcoming budget cuts. Consequently, even if parole agents successfully connect parolees who have mental illness to county programs, the possibility exists that county-based mental health and housing services will not be available in time to safely transition these individuals into the community as their parole terms begin. Corrections will need to mitigate the risks associated with the loss of the program, and we have identified ways for it to do so.

Table 2
How Corrections May Be Able to Mitigate the Impact of Losing Key Services Provided by the Integrated Services Program

Intensive case management
  • Corrections’ prerelease programs identify inmates’ needs and help enroll them in Medi-Cal.
  • Prerelease efforts do not go beyond Medi-Cal enrollment to directly connect inmates to county case management services they will need once paroled.
  • Corrections’ parole agents set reintegration goals with parolees, refer them to parole outpatient clinics (POC), and follow up with providers of services.
  • Corrections has not reviewed these efforts.
  • Parole agents and POC staff may help identify resources for parolees, including county case management services.
  • Limited availability of county services.*
Mental health and medication support services
  • POC staff provide mental health services to parolees: evaluation, medication management, therapy, crisis intervention, and case management.
  • Unlike the program, POCs do not operate 24/7.
  • Corrections’ parole agents and POC staff may help identify resources for parolees, including county outpatient mental health treatment.
  • Corrections has not reviewed parole agent and POC efforts to connect parolees to county services.
  • Limited availability of county services.*
  • Parole agents and POCs may refer parolees to Corrections’ transitional housing programs for specific populations, such as individuals who abuse substances.
  • Corrections did not design its transitional housing programs for those with a serious mental illness.
  • Corrections’ parole agents and POC staff may help identify resources for parolees, such as county‑operated sober living homes or other supportive housing.
  • Counties may not have relationships with housing providers that will accept certain parolees.
  • Limited availability of county services.*

Source: Analysis of Corrections’ policies and procedures as well as reviews of integrated services program documentation and interviews with officials from selected counties.

* Some counties indicated their programs may not be an exact match for the integrated services program, are already at full capacity, and are expected to face upcoming budget cuts.

Corrections Must Successfully Mobilize Existing Resources to Meet Its Responsibilities to Reintegrate Parolees Into Their Communities

Eliminating the integrated services program will not remove Corrections’ responsibility for helping parolees with mental illness and homelessness risks re-enter their communities safely, so Corrections must leverage its parole agents to connect these individuals to the services they need. Although Corrections has operated the program in only eight of California’s 58 counties, four of those counties—Los Angeles, Sacramento, San Bernardino, and San Diego—account for 50 percent of Corrections’ population of parolees. Los Angeles alone accounted for 31 percent of individuals released to parole in 2018 and had the highest number of program slots, 220. Because many parolees have mental illness, the loss of the program represents a significant challenge for Corrections of safely re-integrating those parolees into their communities. Corrections’ parole agents are in a key position to help connect individuals who would typically participate in the integrated services program to necessary services upon release from prison. As outlined in Corrections’ parole operations manual, agents are already required to meet with parolees in order to set goals for addressing their criminal risk factors, including housing stability, employment skills, substance use treatment, and reliable income. The parole operations manual also directs parole agents to follow up with providers of mental health services, which may include county-administered services, to ensure that the parolees are receiving the appropriate services.

Key partners for the parole agents in safely re-integrating parolees with mental illness are the clinical staff of Corrections’ parole outpatient clinics. Corrections maintains these clinics throughout the State. State law requires parole agents to refer individuals to a parole outpatient clinic if they were in a mental health treatment program while incarcerated or if they begin exhibiting symptoms of mental illness after release. Clinical staff—psychiatrists, psychologists, and social workers—at the clinics address parolees’ mental health needs. The clinical staff evaluate mental illness and provide medication management, therapy, crisis intervention, and case management services. They may also assist parolees in connecting to county services as necessary.

Counties operate a number of programs that parolees with mental illness can access. Counties fund their mental health programs through a mix of federal, state, and local funds, including funding from the Mental Health Services Act and Medi-Cal. Each county has contracted with the California Department of Health Care Services (Health Care Services) to provide mental health services through county mental health plans, which under state law must provide care to Medi-Cal-eligible beneficiaries within the county who require mental health services, including parolees. Using these various funding streams, counties are to offer support, such as case management services, mental health services, and medication support services. They also offer housing programs, which include sober‑living homes and other types of housing that provide supportive services to the residents. With the integrated services program ending, Corrections’ parole agents and outpatient clinical staff will need to coordinate closely with the counties to ensure that the needs of the individuals previously served by the program are met through county services. As we discuss next, although existing county systems serve the mental health needs of individuals—including parolees—there are significant risks and challenges associated in transitioning participants from the integrated services program to county programs.

Replacing the Program’s Case Management Services Will Require Additional Focus on Parole Agents’ Ability to Connect Parolees to Community-Based Services

Providing its parole agents with additional support in finding adequate replacements for services the program previously provided is one way for Corrections to mitigate the impact of the loss of the program. The contracted providers have assisted participants with meeting immediate needs, such as food and clothing, as well as in meeting longer-term needs, such as establishing steady income and benefits. The providers have acted as full-time case managers, assisting participants with whatever needs arise. As we discuss previously, because of Corrections’ poor management of the program, sufficient data are not available to demonstrate how effective these providers have been at helping participants meet those needs. However, it is reasonable to conclude that the contracted providers possess the requisite focus and skills to provide these intensive case management services because they are licensed social workers and therapists and the work is in line with their training. Parole agents, on the other hand, must perform numerous other duties, including apprehending individuals who violate their parole, preparing required reports, and testifying in court proceedings.

Corrections’ parole outpatient clinics may also not be complete replacements for the program’s mental health services. For example, the integrated services program providers are required to provide 24-hour crisis care, but the parole outpatient clinics are not available outside of regular business hours; thus, the parole outpatient clinics’ crisis services do not fully replace those of the program. In addition, although Corrections maintains various transitional housing programs throughout the State, none of them exist specifically to serve individuals with serious mental illness who are at risk of homelessness. Therefore, it is unclear whether Corrections’ existing housing programs will have the capacity and expertise necessary to accommodate all of the relatively low‑functioning parolees previously housed through the integrated services program.

As the integrated services program is set to expire in December 2020, Corrections needs to ensure that its parole agents can successfully connect individuals on parole to county services. To its credit, Corrections has increased its efforts in recent years to enroll individuals in Medi-Cal and to screen them for mental illness before their release from prison. Through its prerelease benefit application assistance program, Corrections reported that it had enrolled 86 percent—27,000—of the more than 30,000 inmates it screened in fiscal year 2018–19. Corrections also maintains records of parolees who received a mental illness diagnosis while in prison. However, these necessary efforts do not actually connect inmates nearing release to county services. Although parole agents receive training and resources that cover the subject of connecting parolees to community services, Corrections needs to identify particular parole agents to specialize in serving the program’s target population. It then needs to supplement the training of these agents with tools or resources that make it easier for them to identify and leverage community-based services. Finally, as it has not conducted any reviews—internal or external—of how well parole agents currently connect parolees to these services, Corrections is faced with relying on what can be considered an untested resource. A review of how effective its parole agents are at connecting parolees with mental illness to county services will help ensure that the at‑risk population the program has been serving receives services for their needs.

Corrections also needs to ensure that its parole outpatient clinic staff have the tools they need to connect parolees to county services. Staff in the parole outpatient clinics do not receive specific training on connecting parolees to county services. According to the program administrator, who also oversees the parole outpatient clinics, clinic staff learn about available community services on the job through education and networking with parole agents and providers. As we discuss above for parole agents, Corrections has not conducted internal or external studies of how effective its parole outpatient clinic staff are in connecting individuals to necessary county services. As a result, it is unclear how well staff in the parole outpatient clinics will be able to connect individuals who would have previously participated in the program to county services.

Counties May Not Have Adequate Mental Health and Housing Programs to Serve the Population Associated With the Program

Even when parolees get connected to county services, those services may not be a complete replacement for the integrated services program or have the capacity to accept new clients. We contacted officials in six counties—Fresno, Kern, Los Angeles, Sacramento, San Bernardino, and San Diego—in which the integrated services program operated to get their perspective on the elimination of the program and how county services could replace it. Several officials raised concerns about finding space in their existing services. Kern County, for example, operates a program called the Adult Transition Team, which includes case management services and focuses on reducing or eliminating individuals’ risks of re-entry into jail or prison while providing specialty mental health treatment. It currently serves around 500 adults each year. According to the deputy director of Kern County’s Behavioral Health and Recovery Services (Kern County deputy director), this program is at full capacity and does not offer as intensive a level of services as the integrated services program.

Counties are also likely to face upcoming budget cuts that will affect their behavioral health departments at the same time that their client population expands. Whereas the providers of the integrated services program serve only parolees, individuals paroled after December 2020 who formerly could have been program participants will now join a growing population that needs services through county programs. The County Behavioral Health Directors Association (directors association), a nonprofit advocacy association, sent a letter to the Legislature opposing the elimination of the integrated services program and noted that funding for county mental health programs may drop by up to 19 percent over the next three years from the impacts of COVID-19. The letter noted that counties may also experience an increase in the number of people needing services. The directors association anticipates that Medi-Cal caseloads will grow by 1.2 to 2 million beneficiaries over the next year.

Counties may also not be able to replace the housing services the integrated services program has provided. Participants in the program had access to, at minimum, nine months of housing subsidies, which they did not need to repay. The Kern County deputy director noted that the county provides some short-term housing subsidies, typically three months. However, recipients must reimburse the county once they have an income. A different concern is that some parolees, such as registered sex offenders, must abide by additional housing restrictions or may be excluded from programs. Behavioral health department staff we spoke to in other counties also noted the challenges of finding housing for this population. The integrated services program providers have built relationships with landlords who would accept program participants. The director of Sacramento County’s Behavioral Health Services referred to these relationships as “priceless.” These relationships now may be lost, and as a result counties may struggle to meet the unique needs of parolees.

Counties in which the integrated services program has not operated offer services that possess many of the same risks and weaknesses. We contacted officials in four counties—Orange, Riverside, Shasta, and San Joaquin—to get a better understanding of the services available in those counties to parolees who have a mental illness. Orange and Riverside are among the counties that receive the most parolees in the State, and Shasta and San Joaquin receive high numbers of parolees relative to their respective populations. The counties operate some programs that are designed specifically for parolees or others involved in the justice system. Orange County, for instance, operates a program that provides services to adults with severe mental illness who have recent involvement in the criminal justice system. According to the director of operations of Orange County’s Behavioral Health Services, the program has a capacity of approximately 140 slots. However, because there are many ways to have involvement in the criminal justice system, including being on parole, the program serves a much larger population than one focused on parolees only and parolees might face more competition in accessing these services. We also learned that other county programs may not always accept parolees. For example, a social worker in Shasta County explained that the crisis residential and recovery center accepts registered sex offenders only under certain circumstances and rarely accepts anyone with a history of violence, in order to protect the safety of staff and other clients. Similarly, officials from Riverside and Orange counties noted that it was challenging to find housing for registered sex offenders. An official with San Joaquin County did not respond to our inquiry.

Corrections recognizes that transitioning parolees from the program to county services will take effort, and it has begun this transition. The program administrator has acknowledged that the services Corrections’ parole agents and outpatient clinics can offer after the program ends will not fully replace the program’s services. As such, he met in May 2020 with representatives from four counties in which Corrections operates the program, and with representatives from other counties, to discuss how to transition program participants and future parolees to county services. The program administrator told us that he plans to continue meeting with county representatives to discuss the transition and to establish an ongoing referral process for parolees. He also has affirmed that Corrections’ senior psychologists will act as liaisons to facilitate communication between Corrections and the counties. Further, Corrections maintains 24 parole agent specialists throughout the State who have localized expertise and training in connecting parolees to available programs. A parole administrator from Corrections stated that those agents can help transition parolees with mental illness and homelessness risks to county services. Finally, the program administrator agreed that conducting a review of Corrections’ effectiveness at connecting individuals to county services would be useful for ensuring that parolees are receiving the mental health and housing services they need, as well as for identifying any barriers to accessing services.


To increase public safety and reduce the likelihood of recidivism, Corrections should take the following actions:

To determine whether parolees with mental illness who have housing needs are receiving necessary services and support during their parole terms, Corrections should review its processes for connecting these individuals to county services by:

Corrections should develop its plan by July 2021 and include at least the eight counties formerly served by the integrated services program. Corrections should complete its first review by December 2021.

We conducted this performance audit in accordance with generally accepted government auditing standards and under the authority vested in the California State Auditor by Government Code 8543 et seq. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on the audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives.

Respectfully submitted,

California State Auditor

August 20, 2020

Back to top