United States General Accounting Office GAO Report to Congressional Committees June 1999 HOMELESSNESS State and Local Efforts to Integrate and Evaluate Homeless Assistance Programs GAO/RCED-99-178 United States GAO General Accounting Office Washington, D.C. 20548 Resources, Community, and Economic Development Division B-281482 June 29, 1999 Congressional Committees In 1987, the Congress enacted the Stewart B. McKinney Homeless Assistance Act, recognizing that state, local, and private efforts alone were not adequate to address the growing problem of homelessness in America. Since the McKinney Act was passed, federal resources for alleviating homelessness have increased significantly, and a number of new federal programs have been created specifically to serve homeless people. Yet despite these increased federal efforts, homelessness in America has persisted. The most widely accepted research indicates that up to 600,000 people may be homeless at any given time, and most experts on homelessness agree that programs targeted specifically to people who are homeless do not have sufficient resources to meet the needs of this population. To provide more assistance for homeless people and to meet their multiple and complex needs, states and localities are seeking to link and integrate homeless assistance programs with mainstream social service systems.1 In addition, some states and localities are beginning to use outcome measures to better manage their programs and to ensure that their limited resources are being used for those programs that achieve the best possible results. Using outcome measures shifts the focus from counting outputs, such as the types and numbers of services provided by a program, to measuring outcomes, such as the results achieved by the program. Interested in these developments, you asked us to describe some notable examples of efforts by states or localities to (1) link and integrate their homeless assistance programs with mainstream systems and (2) measure and evaluate outcomes for their homeless assistance programs. This is the second in a series of reports that you asked us to prepare on homelessness.2 To identify notable examples of state or local efforts to link and integrate, and to measure and evaluate outcomes for, their homeless assistance 1 For this report, we used the term “link” for efforts that seek to improve homeless people’s access to mainstream resources, and we used the term “integrate” to refer to more fundamental changes in the ways that agencies or systems of care share or consolidate their resources, planning efforts, and clients to improve the services they provide to the homeless. 2 Homelessness: Coordination and Evaluation of Programs Are Essential (GAO/RCED-99-49, Feb. 26, 1999) was our first report responding to your request. In addition, we recently issued a report on homeless assistance programs provided by the Department of Veterans Affairs, Homeless Veterans: VA Expands Partnerships, but Homeless Program Effectiveness is Unclear (GAO/HEHS-99-53, Apr. 1, 1999). Page 1 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 programs, we interviewed experts on homelessness, including government officials, academics, advocates for homeless people, providers of services to homeless people, and others. As a result of their recommendations, we focused our review on the efforts of two counties—Franklin County, Ohio, and King County, Washington—and two states—Massachusetts and Minnesota. Because these efforts were identified by experts as particularly effective or innovative in serving homeless people, they are not necessarily representative of efforts being made throughout the country. Among the sites we visited, there were several notable examples of state Results in Brief and local efforts to link and integrate homeless assistance programs with mainstream systems. In some cases, these linkages are designed to improve homeless people’s access to mainstream services. For example, to increase the number of eligible homeless people enrolled in Medicaid, the Massachusetts Department of Medical Assistance is conducting outreach at homeless shelters and streamlining the Medicaid application process for this population. In other cases, efforts are being made to integrate entire systems of care. For instance, King County, Washington is seeking to integrate its mental health and substance abuse treatment systems. As part of this effort, King County has created the Crisis Triage Unit—a single place where people, many of them homeless, undergoing mental health or substance-abuse-related crises, can receive treatment and referral through an integrated set of services. In addition, in some communities, mainstream systems are developing policies and programs designed to prevent homelessness, particularly by addressing the discharge practices of institutions that may “feed” homelessness by releasing people who have no place to go. For example, to reduce the number of people who become homeless after leaving correctional facilities, Massachusetts is making efforts to improve its discharge planning for prison inmates and is allocating recovery beds for soon-to-be-released inmates with substance abuse problems who are at risk of becoming homeless. Despite these initiatives, many state and local officials were concerned about the lack of coordination and integration of homeless assistance programs at the federal level, which, they said, adversely affects their efforts at the state and local levels. Nationwide, communities are increasingly using outcome measures to manage their homeless assistance programs, thereby focusing less on the types and numbers of activities performed and more on the results achieved. In Minnesota, for example, the state-funded Family Homeless Prevention and Assistance Program is an outcome-based program that Page 2 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 provides agencies with flexible grants but holds them accountable for achieving certain measurable outcomes related to preventing homelessness among families. One outcome measure used by the program is the number of at-risk families who maintain stable housing. A growing number of communities across the country are also using management information systems to collect uniform data on the use of homeless assistance services as a tool for measuring outcomes and better managing their resources. For example, the Community Shelter Board in Franklin County, Ohio, has developed a comprehensive management information system that collects uniform data from all of the emergency shelters in the county. This system helps the Community Shelter Board track and measure the outcomes of homeless assistance programs countywide and hold service providers accountable for achieving the desired outcomes. This system also helps the community develop strategies for improving policies and programs to serve homeless people. In general, homeless assistance providers told us that they often lack the resources to conduct comprehensive evaluations of their homeless assistance programs, but they hope that their increased use of data systems and outcome measures will enable them to better evaluate their programs in the future. Homelessness in the United States is a widespread and complex problem. Background While the exact number of homeless people is unknown, research by the Urban Institute, which was conducted in 1987 but is still widely cited today, estimated that over a 1-week period, approximately 500,000 to 600,000 people lived on the streets or in emergency shelters.3 About one-half of homeless single adults are believed to have a problem with alcohol abuse and about one-third with drug abuse, according to estimates from a series of studies funded by the National Institute of Mental Health in the mid-1980s. In addition, these studies estimated, about 20 to 25 percent of homeless single adults have a lifetime history of serious mental illness, and about half of those with a serious mental illness also have an alcohol or a drug abuse problem.4 The U.S. Conference of Mayors estimated, in a survey of 30 major cities, that families with children made up about 38 percent of the homeless population in 1998, compared with 3 Martha R. Burt and Barbara E. Cohen, America’s Homeless: Numbers, Characteristics, and Programs that Serve Them (The Urban Institute Press, July 1989). 4 The results of these studies are described in a paper by Robert Rosenheck, Ellen Bassuk, and Amy Salomon entitled Special Populations of Homeless Americans. This paper was presented at the National Symposium on Homelessness Research: What Works, which was cosponsored by the Department of Housing and Urban Development and the Department of Health and Human Services in Oct. 1998. Page 3 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 about 27 percent in 1985.5 Moreover, the needs of people who are homeless vary greatly, as does the nature of the assistance they require. While homelessness is an episodic event for many people who rely temporarily on emergency shelters to help them get through a difficult situation, it is often a chronic condition for others, particularly for those who have a serious substance abuse disorder or a serious physical or mental disability. Consequently, in addition to housing, these individuals may require intensive and ongoing supportive services, such as mental health care or substance abuse treatment, to keep them out of homelessness. A wide range of local, state, and federal agencies, as well as nonprofit organizations, provide shelter and services to homeless people in America. The Stewart B. McKinney Homeless Assistance Act (P.L. 100-77), passed by the Congress in 1987, is the principal federal legislation designed to assist homeless people. The McKinney Act’s programs award grants to communities for activities that provide homeless individuals and families with emergency food and shelter, transitional housing, and supportive services. In fiscal year 1997, the federal government obligated over $1.2 billion for federal programs that are specifically targeted to people who are homeless. Most of the federal government’s funding for programs targeted to homeless people is administered by the U.S. Department of Housing and Urban Development (HUD).6 HUD’s strategy for addressing the problem of homelessness is known as the Continuum of Care. Under this strategy, communities that apply for McKinney Act funds undertake a community-based planning process to help identify the needs of homeless people and develop a comprehensive system, or “continuum of care,” to meet those needs. The Continuum of Care strategy is intended to incorporate a wide array of resources and activities—including homelessness prevention, outreach and assessment, emergency shelter, transitional and permanent housing, and supportive services such as job training, substance abuse treatment, and mental health services—into the system that serves homeless people. 5 A Status Report on Hunger and Homelessness in American Cities – 1998, U.S. Conference of Mayors (Dec. 1998). 6 Other federal agencies that administer programs targeted to the homeless are the departments of Agriculture, Education, Health and Human Services, Labor, and Veterans Affairs and the Federal Emergency Management Agency. Page 4 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 At the locations we visited, we found various examples of state and local Efforts to Link and efforts to link and integrate services for homeless people with mainstream Integrate Homeless systems. In some communities, these linkages seek to improve homeless Assistance Programs people’s access to mainstream services. In other communities, efforts are under way to integrate entire systems of care so as to improve the With Mainstream coordination and quality of services provided to homeless people. Finally, Systems in some communities mainstream systems are developing policies and programs designed to prevent homelessness among people being discharged from institutions such as correctional facilities and psychiatric hospitals. At the same time, many state and local officials noted, a lack of coordination and integration of homeless assistance programs at the federal level adversely affects their efforts at the state and local levels. Efforts to Improve Experts on homelessness, including academics, government officials, and Homeless People’s Access providers of services for homeless people, differ in their opinions as to to Mainstream Programs whether the needs of homeless people are better served by mainstream programs or by programs that are specifically targeted to homeless people.7 While some experts believe that homeless people may be better served by a single coordinated service system specifically targeted to them, others believe that having a separate service system for homeless people “institutionalizes” homelessness and diminishes the will and capacity of the mainstream systems to help the homeless. However, most experts take a middle position on this issue and maintain that although some targeted programs are necessary to address the special needs of homeless people, the major emphasis needs to be on facilitating homeless people’s access to benefits and services provided through mainstream programs. This approach was recognized as the preferred strategy in the federal government’s long-term plan for addressing homelessness published by the Interagency Council on the Homeless in 1994.8 This plan states that mainstream programs must be adapted to ensure that they meet the special needs of homeless people. Moreover, according to the plan, creating a service system specifically for homeless people that is separate from the mainstream system is both inefficient and ineffective. 7 Examples of federal programs targeted specifically to the homeless are Emergency Shelter Grants, Health Care for the Homeless, and the Homeless Children Nutrition Program. Examples of federal programs available to low-income people in general are Public and Indian Housing, Medicaid, and the Food Stamp Program. Across the country, states and localities also offer a wide range of programs, including some targeted to the homeless and others intended for low-income people generally. 8 Priority Home: The Federal Plan to Break the Cycle of Homelessness, Interagency Council on the Homeless (1994). Page 5 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 In the communities we reviewed, we found several examples of state and local efforts to link homeless people with mainstream resources, thereby improving their access to these resources. For example, in Massachusetts, the Division of Medical Assistance has a pilot project designed to improve homeless people’s access to Medicaid.9 The state has combined the eligibility and enrollment process for applicants and has streamlined this process so that it is easier for homeless people to apply for Medicaid. In addition, the state has strengthened its outreach efforts to increase the number of eligible homeless people who are enrolled in Medicaid and has trained staff at emergency shelters so that they can better assist homeless people in completing Medicaid application forms. Massachusetts is also linking its management information system for homeless assistance programs with an automated benefits eligibility system. This effort will automatically link data entered into a homeless shelter’s database to a system that will provide homeless clients with individualized information on which federal, state, and local programs they may be eligible for. Linking the two systems should facilitate homeless people’s access to mainstream programs and services, according to state planning documents. (See app. I for more detailed information on Massachusetts’ efforts in these areas.) Efforts to improve homeless people’s access to mainstream services are also taking place through Seattle-King County’s Health Care for the Homeless Network.10 This model for implementing the Health Care for the Homeless program combines direct services provided by the staff of the Seattle-King County Department of Public Health with contracted services provided by mainstream health service providers. Dedicating public health staff specifically to providing health care services to homeless people helps ensure that adequate outreach is conducted to meet the special needs of this population. At the same time, contract agreements with hospitals and other community providers help ensure that existing mainstream health care resources are used to serve homeless people and that these mainstream systems are held accountable for providing care to the homeless population. (See app. II for more detailed information on the Seattle-King County program.) 9 Medicaid finances health care for certain poor and disabled individuals nationwide. It is jointly funded by the federal government and the states and is administered by the states with broad federal guidance. 10 Seattle-King County’s Health Care for the Homeless Network is funded, in part, by the U.S. Department of Health and Human Services’ Health Care for the Homeless program, which provided grants to 128 projects nationwide in fiscal year 1998, with the goal of making high-quality health care accessible to homeless people. Page 6 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 Efforts to Integrate Experts on homelessness widely agree that integrated social service Systems to Improve systems are needed to meet the numerous and complex needs of homeless Services for Homeless people. Many of these experts believe that the social services required by homeless people—such as mental health, substance abuse treatment, and People job training services—already exist. However, these services tend to be fragmented and uncoordinated and, as a result, are not well suited to serving homeless people, who may have multiple problems and often face many barriers to receiving assistance. To address this issue, many communities are attempting to integrate the systems of care that are provided to homeless people by different agencies. For most communities, “systems integration” requires fundamental changes in the ways that agencies share information, resources, and clients. Systems can be integrated, for example, through the development of cross-agency strategic plans and interagency management information systems, the consolidation of programs or agencies, and the pooling of funds.11 In particular, community officials and service providers told us that people who are homeless would benefit from better integration of the mental health and substance abuse treatment systems. Traditionally, institutional and philosophical differences have divided these two service systems, creating problems in providing services to people who have co-occurring mental health and substance abuse disorders—a condition common among homeless people. Because people with co-occurring disorders, including homeless people, frequently receive treatment from two different systems, their care is often not coordinated, and neither the mental health nor the substance abuse system is willing to take full responsibility for their care. Furthermore, experts say, effectively treating people with co-occurring disorders often requires a “holistic” approach to effectively address all of their needs. King County, Washington, has taken several steps to integrate its mental health and substance abuse systems. The county is currently merging the two divisions that provide mental health and substance abuse services and has a full-time “systems integration administrator” who is responsible for facilitating the integration of the two systems and creating links with other county systems, such as corrections, housing, and welfare. King County’s systems integration efforts operate on a “no wrong doors” philosophy, under which people with mental illness or substance abuse problems are offered the services they need whether they seek assistance through the hospitals, detoxification centers, emergency shelters, mental health 11 The concept of systems integration is discussed more fully in a paper by Deborah L. Dennis, Joseph J. Cocozza, and Harry J. Steadman entitled What Do We Know About Systems Integration and Homelessness?, presented at the National Symposium on Homelessness Research (Oct. 1998). Page 7 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 treatment facilities, or correctional facilities. As part of this effort, in July 1998, the county implemented a pilot project, the Crisis Triage Unit, which serves a single place where people undergoing mental health, substance abuse, or other behavioral health crises can receive services and referrals. About half of those brought to the unit are homeless, and many more are at risk of becoming homeless. In addition, the county has established the Chronic Public Inebriates Systems Solutions Workgroup to help address problems related to the street homeless who are chronic abusers of alcohol and often have secondary drug abuse or mental illness disorders as well. The workgroup has implemented a series of measures, including a sobering sleep-off center and a housing plan for this population. (See app. II for more detailed information on King County’s systems integration efforts.) Another example of an effort to create a coordinated system for homeless assistance is in Franklin County, Ohio, where the Community Shelter Board, a nonprofit agency, coordinates and plans all emergency shelter services for the county. According to Franklin County officials, service providers, and state officials, the Community Shelter Board’s role as a single coordinating body allows the emergency shelters in Franklin County to work as a system rather than as a fragmented set of resources, improving linkages between the emergency shelter system and mainstream resources within the community. The Community Shelter Board provides a single conduit for funding the shelters in the county, organizes the county’s Continuum of Care plan, and serves as a bridge between and among the public, private, and nonprofit sectors on issues and planning efforts related to homelessness and emergency shelters. (See app. III for more detailed information on Franklin County’s Community Shelter Board.) Initiatives by Mainstream In some communities, mainstream social service systems are increasingly Systems to Prevent developing policies and programs designed to prevent homelessness. In Homelessness the past, efforts to prevent homelessness consisted mainly of activities such as preventing evictions by providing short-term rental assistance to families. However, there is a growing recognition that it may be possible to prevent homelessness by modifying the discharge practices of institutions such as correctional facilities, hospitals, and psychiatric institutions. These systems may “feed” homelessness because people released from these systems often have no place to go. Experts believe that collaboration between these mainstream systems and the homeless assistance system can facilitate the development of measures for preventing homelessness. Page 8 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 In Massachusetts, efforts are being made to reduce the number of people who become homeless after leaving correctional facilities. Both the state’s Department of Correction and county correctional agencies have devoted more resources to planning for the discharge of inmates who will soon be released. In addition, the state’s Department of Public Health has implemented a criminal justice initiative, which allocates a number of recovery beds for those who are being released from the corrections system, have a substance abuse problem, and are at risk of becoming homeless. Moreover, the Massachusetts Department of Mental Health has in place a number of policies and procedures that are designed to prevent patients who are being discharged from psychiatric hospitals from becoming homeless. For example, the Department’s Homeless Services Unit works with formerly homeless mental health clients to help them find adequate housing before they are discharged from mental health facilities. Similarly, the Massachusetts Division of Medical Assistance requires the private contractor that provides mental health services for many of the state’s Medicaid recipients to identify strategies and resources to help prevent clients who are being discharged from inpatient psychiatric facilities from becoming homeless. (See app. I for detailed information on Massachusetts’ homeless prevention efforts.) King County, Washington, recently started the Mental Health Court, a pilot effort designed, in part, to prevent individuals with mental illness from cycling between homelessness and the correctional system. Under this effort, mentally ill people who have been charged with misdemeanors will typically have the option of receiving court-ordered treatment as an alternative to prosecution or sentencing. Unlike the regular court system, the Mental Health Court provides a number of individual treatment and supportive services, as well as a limited amount of temporary housing. County officials estimate that about one-third of those who will use the Mental Health Court will be homeless and many more will be at risk of becoming homeless. (See app. II for detailed information on the King County Mental Health Court.) State and Local Several federal initiatives encourage states and localities to link and Perceptions That Federal integrate their homeless assistance programs with mainstream service Efforts to Integrate systems. For example, HUD’s Continuum of Care strategy encourages communities to create linkages between services for the homeless and Services for Homeless mainstream services such as job training, child care, substance abuse People Could Be Improved treatment, and mental health services. A 1996 HUD-contracted evaluation of the Continuum of Care strategy found that it had generally been successful Page 9 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 in helping communities develop a more focused and structured process for bringing together a wide range of stakeholders and encouraging collaboration among service systems at the state and local levels.12 Efforts by the U.S. Department of Health and Human Services (HHS) also encourage linkages and program integration at the state and local levels. For example, HHS’ Health Care for the Homeless program emphasizes a multidisciplinary approach to delivering health care to the homeless, combining outreach with integrated systems of primary care, mental health and substance abuse services, and case management. Similarly, HHS’ Access to Community Care and Effective Services and Supports (ACCESS), a 5-year demonstration project, has been evaluating the effectiveness of integrated systems of care for homeless people with mental illness. In addition, as we stated in our February 1999 report,13 efforts to assist homeless people at the federal level are coordinated in several ways. Coordination occurs through (1) the Interagency Council on the Homeless,14 which brings together representatives of federal agencies that administer programs or resources that can be used to alleviate homelessness; (2) jointly administered programs and policies adopted by some agencies to encourage coordination; and (3) compliance with the requirements of the Government Performance and Results Act of 1993, which requires federal agencies to identify crosscutting responsibilities, specify in their strategic plans how they will work together to avoid unnecessary duplication of effort, and develop appropriate performance measures for evaluating their programs’ results. However, the consensus of the state and local government officials, advocates for homeless people, and homeless assistance providers with whom we spoke was that the federal government has not done a good job of coordinating its programs, and this lack of coordination adversely affects the ability of states and localities to integrate their programs. Although HUD and HHS have stated that they have a number of activities to promote coordination between the two departments, state and local 12 Ester Fuchs and William McAllister, The Continuum of Care: A Report on the New Federal Policy to Address Homelessness (Dec. 1996). 13 Homelessness: Coordination and Evaluation of Programs Are Essential (GAO/RCED-99-49, Feb. 26, 1999). 14 The McKinney Act established the Interagency Council on the Homeless, an independent council with its own funding and staff, to promote the coordination of homeless assistance programs across federal agencies. In 1994, because of concerns that the Council was not effectively coordinating a federal approach to homelessness, the Congress stopped appropriating funds for the Council, and it became a voluntary working group under the President’s Domestic Policy Council. According to HUD, the discontinuation of funding has significantly changed the role of the Council, and its activities are now limited mostly to facilitating the exchange of information and managing limited special projects. Page 10 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 officials and service providers told us that they were particularly concerned about what they perceive as insufficient communication and coordination between the two departments. Many were particularly critical of what they felt was HHS’ lack of involvement in addressing homelessness. As a result, in their opinion, HUD has funded and administered most of the non-housing-related supportive services for the homeless through its McKinney Act programs. Some state and local officials also felt that HHS should do more to integrate mental health and substance abuse programs at the federal level. Such integration, they said, is necessary to effectively treat homeless individuals with co-occurring disorders. These officials also said that even though various federal grants to states and localities have similar goals, they often have differing eligibility criteria, funding cycles, and reporting requirements, which make it difficult to incorporate these programs into an integrated system of care at the local level.15 In commenting on a draft of this report, while HHS agreed that more could be done at the federal level to better serve the homeless population, it did not agree with state and local officials’ perceptions that the department was not adequately involved in addressing homelessness or integrating mental health and substance abuse programs to effectively treat homeless people with co-occurring disorders. According to HHS, it has undertaken several initiatives in conjunction with HUD and other agencies to better address the needs of homeless people in general, as well as serve people with co-occurring disorders. In its comments, HHS restated its commitment to exploring additional opportunities to improve coordination with HUD and other federal agencies as they continue to address homelessness and develop and implement approaches to improve services for those with co-occurring disorders. Moreover, HHS emphasized that the coordination of resources received from federal agencies must fundamentally occur at the state and local levels, and that state and local entities must work together to appropriately address and balance the needs of homeless people with the needs of a multitude of other groups. (See app. V for the full text of HHS’ comments on this report.) 15 We will explore these issues in greater detail as part of our planned review of the barriers faced by homeless people in gaining access to federal programs. Page 11 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 Many communities across the country are increasingly using outcome Use of Outcome measures to manage their homeless assistance programs, and we found Measures, Data several examples of the use of the measures at the sites we visited. Using Systems, and Program outcome measures to manage programs is becoming increasingly popular with federal, state, and local governments as they wrestle with ways to Evaluations for improve the effectiveness and quality of government-provided services Homeless Assistance while limiting the costs to deliver these services. The use of outcome measures shifts the focus from outputs, such as the types and numbers of Programs activities performed, to the outcomes, or results achieved. For homeless assistance programs, this means a shift in focus from tracking outputs, such as the number of people sheltered, to measuring outcomes realized, such as the number of people who move out of homelessness and into a stable housing situation.16 In addition to using outcome measures, more communities are using management information systems to collect uniform data on their homeless population and on the resources used by them so they can improve the management and coordination of these resources. Providers of services to the homeless and state and local officials said that they generally lacked the resources to conduct comprehensive evaluations of their homeless assistance programs but hoped that the increased use of data systems and outcome measures would improve their ability to evaluate these programs in the future. Communities’ Increasing Communities nationwide are increasingly setting and using outcome Use of Outcome Measures measures to evaluate their homeless assistance programs, according to for Homeless Assistance researchers and homeless assistance providers. Several reasons may account for this increased emphasis by states and localities on measuring Programs outcomes. First, there is a growing recognition among state and local governments that they need to spend their limited resources on programs that “work.” Consequently, agencies that provide services to the homeless are being required to focus on achieving results–such as moving people out of homelessness–rather than on just providing units of service. Second, an increasing number of management information systems for homeless assistance programs have been developed and implemented in recent years. The availability of these systems makes it easier for state and local officials to collect and use standardized outcome data to manage their homeless assistance programs. Third, states and localities have been 16 While stable housing is generally the ultimate outcome goal of homeless assistance programs, many programs also have important intermediate outcome goals for the homeless people they serve, such as involvement in mental health or substance abuse treatment, improved level of functioning, or improved health status. These can represent important intermediate steps on the path to stable housing for some homeless people, particularly those suffering from mental illness, a substance abuse disorder, or a chronic health problem. Page 12 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 influenced by the federal government’s move towards the use of outcome measures under the Government Performance and Results Act of 1993, which requires federal agencies to set specific performance goals and to measure outcomes for federal programs. Finally, some private foundations are requiring greater accountability for the funds they provide to agencies that serve the homeless. For example, in Minnesota, the Family Housing Fund, which provides funds for two single-room-occupancy projects that largely serve formerly homeless individuals, requires the managers of the projects to track several performance measures, such as tenants’ stability in housing and employment. Similarly, the United Way of King County, Washington, outlines in its contract with the YWCA of Seattle several specific outcome goals, such as increased housing stability for those served by the program. At the sites we visited, we found several examples of how states and localities are using outcome measures to manage and improve their homeless assistance programs, including the following: • Minnesota’s state-funded Family Homeless Prevention and Assistance Program is an outcome-based program that focuses on three specific goals—preventing homelessness, reducing the length of stay in emergency shelters, and eliminating repeat episodes of homelessness. The program provides local government and nonprofit agencies with flexible grants that can usually be used however an agency decides as long as the agency sets specific outcome goals, develops a method for tracking these outcomes, and achieves and reports on these outcomes. (See app. IV for more detailed information on Minnesota’s program.) • In Massachusetts, the state’s Division of Medical Assistance has set certain performance standards related to homeless people in its contract with the company that provides behavioral health services for many of the state’s Medicaid recipients. One performance standard requires the company to implement measures that will reduce the inappropriate discharge of people into homelessness from psychiatric facilities. The second performance standard provides incentives to the company for increasing the number of eligible homeless individuals enrolled in Medicaid. The company receives financial bonuses or penalties on the basis of its success in meeting these performance standards. (See app. I for more detailed information on Massachusetts’ programs.) • The Ohio Department of Development has started to implement the use of outcome measures for some of its housing programs that serve homeless people. Agencies that receive state funds for supportive housing programs are required to develop outcome-based performance targets that the state Page 13 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 will hold them accountable for achieving. For example, a general outcome measure for a transitional housing program might be the percentage of clients that were moved to some kind of permanent housing. State officials told us that they hope to improve the quality of the projects they fund by focusing on the outcomes achieved and hope that these requirements will encourage agencies with poorly performing programs to improve, while highlighting the “best practices” of those agencies that have successful programs. At the county level, the Community Shelter Board in Franklin County, Ohio, has been working with the state to establish outcome measures for service providers in the county. Contracts with service providers that receive funds from the Community Shelter Board include specific outcome measures, such as the percentage of clients moved out of shelters into transitional housing within a given period of time. (See app. III for more detailed information on Ohio’s efforts to use outcome measures.) States’ and Localities’ A growing number of states and localities are using various data systems Efforts to Develop Data to manage their homeless assistance programs. Both individual homeless Systems and Evaluate assistance providers and entire service systems are using these management information systems to collect, track, and analyze Homeless Assistance information on their clients and the services they use. As many as 50 cities Programs are using or are in the process of implementing an estimated 15 to 18 different software applications designed to automate the collection and management of data on the use of homeless assistance services, according to a researcher who has worked with several of these cities. This information can be collected at various points in the system, such as emergency shelters, transitional housing programs, or programs that provide supportive services to homeless people. Communities and service providers can use the data collected by these systems in a variety of ways, from tracking a client’s movement through the system, to assisting in a client’s case management, to gathering general demographic data on the homeless population, to developing policies and plans. Massachusetts, for example, is expanding its use of a computerized record-keeping system for the homeless, called the Automated National Client-specific Homeless services Recording (ANCHoR) system, and is implementing the system statewide.17 This system allows service providers to collect uniform information on their homeless clients over time. It is designed to help service providers assess the needs of their homeless 17 The ANCHoR system was developed with funding from HUD, HHS, and others. At present, approximately 30 cities across the nation are either using the system or are in the process of implementing it. Page 14 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 population, manage their emergency shelter resources, and provide better case management services, including referral and follow-up. By implementing the system in emergency shelters, transitional housing programs, and other homeless assistance programs across the entire state, a Massachusetts official told us, they hope to better coordinate resources for homeless people statewide and better evaluate programs’ effectiveness. (See app. I for more detailed information about Massachusetts’ use of the ANCHoR system.) Similarly, the Community Shelter Board in Franklin County, Ohio, has developed a comprehensive management information system to collect uniform data from all of the county’s emergency shelters. This management information system includes both client- and provider-specific data and can provide information on various outcomes, such as the average length of stay in a shelter for homeless men in the county and the percentage of homeless people who move to permanent housing within a given time period. A Community Shelter Board official said that the management information system helps them track and measure the outcomes of homeless assistance programs countywide and hold service providers accountable for achieving agreed-upon outcomes. In addition, the system helps the community develop strategies for improving policies and programs for homeless people. (See app. III for more detailed information on Franklin County’s use of management information systems.) State and local homeless assistance providers and officials told us that they typically have not had sufficient resources to conduct comprehensive evaluations of their homeless assistance programs. However, they hope that the increased use of data systems and outcome measures will improve their ability to evaluate these programs in the future. Experts on homelessness whom we spoke to cited Minnesota as a state that has been unusually active in evaluating homeless assistance programs and collecting comprehensive data on its homeless population. Every 3 years, Minnesota conducts a comprehensive statewide census and survey of homeless people. According to state and local officials, these surveys help policymakers and planners gauge trends in, and assess the needs of, the homeless population and plan and lobby for the resources required to address these needs. State, county, and nonprofit agencies in Minnesota also perform a relatively large number of evaluations to determine the effectiveness of specific programs for homeless people. According to government officials and service providers, these evaluations have helped them determine which programs and activities are most effective in aiding Page 15 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 homeless people and in preventing homelessness. (See app. IV for more detailed information on Minnesota’s data collection and evaluation efforts.) We provided a draft of this report to HHS and HUD for review and comment. Agency Comments Both departments provided us with comments that appear in appendixes V and VI of the report, along with our detailed responses. HHS stated that it appreciated the timeliness of this report and our earlier February 1999 report on homelessness because federal, state, and local agencies continue to struggle with the persistent problem of homelessness in the United States. However, HHS also made several points to clarify issues raised in this report. HHS’ primary concern related to our reporting of state and local officials’ perceptions that the Department is not adequately involved in addressing homelessness in general or in integrating federal programs to meet the needs of people with co-occurring disorders. HHS disputed this characterization and cited several initiatives—such as ACCESS, a national survey of homeless assistance providers and clients, a symposium on homelessness research, and various forms of technical assistance that it has provided to the states—as examples of its involvement in addressing homelessness. HHS also described several efforts it has initiated to integrate mental health and substance abuse programs to better serve individuals with co-occurring disorders. While HHS agreed that more could be done to coordinate the efforts of various federal agencies to address homelessness, it also described several joint initiatives that it has undertaken with HUD and other federal agencies to improve federal programs that serve the homeless. HHS also emphasized that the coordination of resources received from federal agencies must fundamentally occur at the state and local levels and that state and local entities must work together to appropriately address and balance the needs of homeless people with those of a multitude of other groups. In its comments, HHS also restated its continuing commitment to developing better solutions for serving homeless people in general, as well as those with co-occurring disorders, and to improving coordination with other agencies. Although we agree that HHS is engaged in several initiatives concerning homelessness, our study raises some issues about how the Department’s efforts are perceived by states and localities. The observations we have reported are based on interviews we conducted with more than 50 state and local officials in four different locations across the country and clearly suggest that many at the state and local level believe that the Department can do more to address Page 16 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 the issue of homelessness. HHS also provided us with technical comments that have been incorporated in the report as appropriate. HUD was pleased that the report highlighted the good work of several communities to integrate the housing and services needed by homeless people. However, HUD stated that the report did not fully reflect the significantly changed role of the Interagency Council on the Homeless. We have revised the report to include information that describes the current role of the Council. To identify notable examples of efforts by states and localities to (1) link Scope and and integrate their homeless assistance programs with mainstream Methodology systems and (2) measure and evaluate outcomes for their programs that serve homeless people, we interviewed national experts on homelessness. These experts included HUD and HHS officials that administer programs for homeless people; representatives of national advocacy groups for homeless people, including the National Coalition for the Homeless and the National Alliance to End Homelessness; and researchers and others with expertise in this area. Of all of the sites suggested by these experts, we selected four from among those most often identified as being particularly effective or innovative in linking or integrating homeless assistance programs with mainstream systems or using program evaluations and outcome measures to manage their homeless assistance programs. As a result of this process, we selected two counties—Franklin County, Ohio, and King County, Washington—and two states—Massachusetts and Minnesota. Because these counties and states were chosen for having programs or initiatives that experts considered particularly effective or innovative, they are not necessarily representative of all states and localities throughout the country. We visited each of the four sites we selected and interviewed state and local officials, providers of services to homeless people, advocacy groups for homeless people, private foundation employees, community-based researchers, and others to obtain information and documents on their efforts to integrate or evaluate their homeless assistance programs. We also collected information on federal initiatives to promote the coordination and evaluation of homeless assistance programs at the federal, state, and local levels from officials at HHS and HUD. We conducted our work between July 1998 and May 1999 in accordance with generally accepted government auditing standards. Page 17 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 We are sending copies of this report to the appropriate congressional committees, the Honorable Donna Shalala, the Secretary of Health and Human Services, and the Honorable Andrew Cuomo, the Secretary of Housing and Urban Development, and other interested parties. Copies will be made available to others on request. If you have any questions about this report, please call me or Anu Mittal at (202) 512-7631. Key contributors to this assignment were Jason Bromberg and Myrna Pérez. Judy A. England-Joseph Director, Housing and Community Development Issues Page 18 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs B-281482 List of Congressional Committees The Honorable Phil Gramm Chairman, Committee on Banking, Housing and Urban Affairs U.S. Senate The Honorable Pete V. Domenici Chairman, Committee on Budget U.S. Senate The Honorable James M. Jeffords Chairman, Committee on Health, Education, Labor and Pensions U.S. Senate The Honorable Arlen Specter Chairman, Committee on Veterans’ Affairs U.S. Senate The Honorable Christopher S. Bond Chairman, Subcommittee on VA, HUD, and Independent Agencies Committee on Appropriations U.S. Senate The Honorable Wayne Allard Chairman, Subcommittee on Housing and Transportation Committee on Banking, Housing and Urban Affairs U.S. Senate The Honorable Bill Frist Chairman, Subcommittee on Public Health Committee on Health, Education, Labor and Pensions U.S. Senate Page 19 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Contents Letter 1 Appendix I 22 Background 22 Massachusetts Improving Access to Medicaid and Setting Performance 23 Standards for Managed Care Services Efforts to Prevent Homelessness for Those Released From 24 Correctional Facilities Massachusetts’ Use of Management Information Systems 27 Appendix II 30 Background 30 King County, Systems Integration in King County 30 Washington Seattle-King County’s Health Care for the Homeless Network 33 Appendix III 37 Background 37 Franklin County, Ohio Coordination of Emergency Shelter and Other Services Through 38 the Community Shelter Board Data Collection and Program Evaluation Efforts in Franklin 40 County State’s and County’s Use of Outcome Measures to Improve 41 Programs for Homeless People Appendix IV 43 Background 43 Minnesota Minnesota’s Family Homeless Prevention and Assistance 44 Program Minnesota’s Statewide Survey of Homeless People 45 Minnesota’s Evaluations of Programs That Serve the Homeless 47 Appendix V 49 GAO’s Comments 53 Comments From the Department of Health and Human Services Page 20 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Contents Appendix VI 54 GAO’s Comments 56 Comments From the Department of Housing and Urban Development Abbreviations ACCESS Access to Community Care and Effective Services and Supports ANCHoR Automated National Client-specific Homeless services Recording System CSB Community Shelter Board DMA Division of Medical Assistance FHPAP Family Homeless Prevention and Assistance Program GAO General Accounting Office HCHN Health Care for the Homeless Network HHS Department of Health and Human Services HUD Department of Housing and Urban Development ROOF Rebuilding Our Own Futures YWCA Young Women’s Christian Association Page 21 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts This appendix describes some of the initiatives taken in Massachusetts to improve homeless people’s access to mainstream resources, prevent homelessness, and use management information systems to provide better services for homeless people. A number of national experts on homelessness identified the state of Massachusetts and the city of Boston as particularly innovative in linking programs for homeless people with mainstream programs and adopting policies within their mainstream systems to try to prevent homelessness. The state has several efforts under way to improve homeless people’s access to Medicaid and ensure that the program’s mental health services adequately serve the needs of homeless people. In addition, various state agencies are implementing initiatives to help reduce the number of people who become homeless after being released from correctional or psychiatric facilities. Finally, Massachusetts is expanding its use of a computerized record-keeping system for homeless assistance services and is implementing the system statewide. It is also linking this system to a benefits eligibility system. Massachusetts had a population of about 6.1 million in 1998, according to Background a U.S. Census Bureau estimate. Although the state has the fourth highest per-capita income in the nation, its cost of living is also among the highest. Housing costs in Massachusetts are considerably higher than the national average, particularly in the Boston metropolitan area. About two-thirds of the state’s homeless population is located in Boston. In December 1998, a one-night census of the homeless conducted by the city counted 5,272 homeless people. Of this population, 44 percent were living in adult shelters, 23 percent were in family shelter programs, 4 percent were living on the street, and the remainder were in transitional housing programs, hospitals, and other settings. The Massachusetts Department of Transitional Assistance funds the majority of the state’s emergency shelters. Various state agencies are responsible for most of the supportive services provided to homeless people, including mental health and substance abuse treatment. The state’s Interagency Task Force for Housing and Homelessness coordinates planning activities and services for homeless people and also develops programs that serve homeless people. In Boston, the city’s Emergency Shelter Commission coordinates policy development, advocacy, and public education on homelessness, while the Department of Neighborhood Development manages, oversees, and distributes most of the grants Page 22 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts received by the city for services for homeless people. The city also funds two emergency shelters. In Massachusetts, the Division of Medical Assistance (DMA) administers the Improving Access to state’s Medicaid program, known as MassHealth. DMA has initiated a pilot Medicaid and Setting project to increase the enrollment of homeless people in MassHealth by Performance streamlining the eligibility and enrollment process for this group. DMA also uses performance outcomes to manage the Medicaid contractor that Standards for provides mental health and substance abuse services for most Medicaid Managed Care clients in the state, and two of the performance standards that it uses are related specifically to the issue of homelessness. Services State Initiatives to Improve DMA has established a pilot project to increase the enrollment of homeless Homeless People’s Access people in MassHealth, the state’s Medicaid program. One goal of the pilot to Medicaid project is to make it easier for homeless people to enroll in the program by allowing the state to determine their eligibility and enroll them at the same time. Normal enrollment procedures require people to go through a two-step process. For the pilot project, DMA has streamlined the process to suit the special circumstances faced by homeless people. For example, under normal enrollment procedures, forms are sent to an applicant’s permanent mailing address, but under the pilot project, these forms can be sent to a staff member at an emergency shelter who serves as the homeless applicant’s “contact person.” In addition, DMA has increased its outreach efforts to educate community organizations, advocates for homeless people, and others about MassHealth, its eligibility requirements, and the enrollment process. As part of these outreach efforts, DMA is providing special training to staff at the four homeless shelters participating in the pilot project. Shelter staff have been trained to assist homeless clients in completing the forms to determine their eligibility for MassHealth and to provide information on how the enrollment process works. Shelter staff have been given special access to certain client-specific eligibility information that allows them to call DMA to learn whether a homeless client is eligible for MassHealth. Performance Standards for About half of the Medicaid recipients in Massachusetts receive mental Serving the Homeless health and substance abuse treatment through the Massachusetts Included in Medicaid Behavioral Health Partnership, a private company that provides mental health and substance abuse services under a contractual arrangement with Service Provider’s Contract Page 23 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts DMA. DMA monitors the Partnership’s performance against 18 performance standards that were included in its fiscal year 1999 contract. If these standards are met, the Partnership receives financial bonuses and if they are not met, penalties are assessed. Two of the 18 performance standards specifically address issues relating to homeless people. The first performance standard included in the contract expects the Partnership to collaborate with advocates for homeless people to find ways to ensure that patients in psychiatric facilities are not discharged inappropriately to shelters. It also expects the Partnership to educate its providers of inpatient mental health care and monitor their performance to ensure that homeless patients are appropriately discharged from their facilities. To meet this standard, officials from the Partnership told us that they now require a senior manager to approve a patient’s discharge plan before the patient can be discharged from a hospital to a homeless shelter. They will approve a patient’s discharge to a shelter only after all other alternatives and resources have been considered. The Partnership has also created a Homeless Task Force that, among other things, works with mental health care providers to promote appropriate psychiatric discharge policies and practices. In addition, the Partnership has contributed funding for the establishment of a toll-free telephone system that is being set up by the Massachusetts Housing and Shelter Alliance. This system will provide discharge planners and case managers with access to current information on housing options and services available for homeless individuals. The Partnership is giving its providers special training on how to use the information that is provided by the telephone system to avoid the inappropriate discharge of patients into homelessness. The second performance standard included in the contract provides a financial incentive through the Partnership to certain homeless shelters and detoxification programs that enroll new members in MassHealth. To help meet this standard, the Partnership has provided training to staff at these facilities on MassHealth’s enrollment procedures and has helped DMA in its efforts to streamline the eligibility and enrollment process for homeless people applying for MassHealth. There has long been concern about ex-offenders who become homeless Efforts to Prevent after they complete their sentences and are discharged from correctional Homelessness for facilities. In Massachusetts, the Department of Correction estimates that Those Released From 15 percent of those released from state correctional facilities have nowhere to go. Using a representative sample, the Massachusetts Housing Correctional Facilities Page 24 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts and Shelter Alliance estimated that 1,259 ex-offenders went directly from state and county prisons into emergency shelters in 1998. To prevent people who are leaving correctional facilities from becoming homeless, Massachusetts is making efforts to improve discharge planning and is targeting substance abuse recovery home beds for released inmates who are at risk of becoming homeless. Criminal Justice System’s Over the past few years, concerns about the corrections system Efforts to Improve discharging people into the shelter system has led to increased Discharge Planning for communications between the corrections system and advocacy groups for homeless people in Massachusetts. This has provided a stimulus for the Those Leaving Department of Correction to seek improvements in discharge planning for Correctional Facilities soon-to-be-released inmates, according to a department official. One of the purposes behind the move for improved discharge planning is to prevent former inmates from cycling through the “revolving door” between the shelter system and the corrections system. In 1998, the Department of Correction revised its Release and Lower Security Preparation Policy, which sought to improve discharge planning and services for all soon-to-be-released inmates from the state corrections system. Under this policy, when inmates in the state corrections system have 1 year before their release, they attend transition workshops. A personalized transition plan is developed for each inmate that addresses postrelease issues such as employment and housing. The corrections system has contracted with a community-based agency that makes appropriate referrals for needed services and housing for each individual who is to be released. The county corrections systems, which are adminstered separately from the state system, have hired full-time discharge planners to perform similar discharge planning functions for the counties’ houses of corrections. Criminal Justice Initiative Massachusetts has a criminal justice initiative whose goal is to provide Designed to Provide beds in recovery homes for persons with substance abuse problems who Recovery Homes for have been released from correctional facilities and are at risk of becoming homeless. This initiative stemmed from discussions that began in 1996 Ex-Offenders With between the Massachusetts Housing and Shelter Alliance, the Department Substance Abuse Problems of Correction, and the state’s Executive Office of Public Safety on ways to prevent ex-offenders from becoming homeless. Because the Department of Correction is not legally responsible for individuals after they have completed their sentences, these groups determined that partnerships Page 25 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts with other agencies were required to address this issue. Since an estimated 80 percent of those entering the shelter system from prisons have substance abuse problems, the Department of Public Health, which funds the state’s substance abuse services, became involved in these discussions. The criminal justice initiative began in 1996, and approximately $2.1 million was allocated for this initiative in fiscal year 1997. These funds support about 135 recovery home beds specifically targeted for persons released from correctional facilities who have substance abuse problems and are at risk of becoming homeless, according to a Department of Public Health official. The Department of Public Health contracted with the Massachusetts Housing and Shelter Alliance to coordinate the initiative. Beginning in September 1997, monthly meetings were held with representatives from a variety of agencies, including the state departments of Correction and Public Health; the Parole Board; county corrections facilities; and recovery home providers. The primary purpose of these meetings was to coordinate the allocation and use of the 135 recovery home beds. For example, a subcommittee was established to survey inmates and determine what information the inmates needed to have about each recovery home so that they could choose the facility that best met their needs. Similarly, another subcommittee developed a standard application form so that inmates could use one application to apply to different recovery homes throughout the state. Participating agencies also addressed a wide variety of other issues, including the need for transitional housing for soon-to-be-released inmates for whom recovery home beds are not yet available. To help gauge the impact of the program, the Massachusetts Housing and Shelter Alliance will be tracking data on the number of people entering shelters for the homeless after being discharged from correctional facilities. Department of Mental The Massachusetts Department of Mental Health, which serves individuals Health’s Efforts to Prevent with severe and persistent mental illness, estimates that about one-third of Discharge From State its clients who are released from the corrections system become homeless. In April 1998, the department instituted the Forensic Transition Facilities Into Team, whose goal is to assist mentally ill individuals who are making the Homelessness transition from correctional facilities back into society. A department official said that preventing homelessness is one goal of the program and helping clients find housing is one task of the Forensic Transition Team. Page 26 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts The initiative to prevent homelessness for mentally ill ex-inmates is part of the Department of Mental Health’s general policy of preventing homelessness among mentally ill clients being discharged from state facilities. The department’s Homeless Services Unit is notified whenever a homeless client enters an inpatient mental health facility, and the unit works to secure housing and other services for the client as part of the discharge planning process. Department of Mental Health staff are prohibited from discharging a client into an emergency shelter unless all other housing options have been considered and the client refuses the housing that is offered. Massachusetts is implementing a computerized management information Massachusetts’ Use of system statewide that will allow providers of services for homeless people Management to collect and access uniform information about their homeless clients and Information Systems the services they use. In addition, Massachusetts is linking its management information system with an automated benefits eligibility system, which will allow homeless individuals to more easily identify the mainstream programs and services that may be available to them. Statewide Implementation The Automated National Client-specific Homeless services Recording of a Computerized (ANCHoR) system is a computerized record-keeping system designed to Management Information allow service providers to collect uniform information on their homeless clients. The ANCHoR system was developed with funding from the U.S. System Department of Housing and Urban Development (HUD), the U.S. Department of Health and Human Services (HHS), and other sources, and is currently being used or is in the process of being implemented by approximately 30 cities nationwide.1 The ANCHoR system is designed to help service providers assess clients’ needs, manage shelter stays, and provide overall case management, including referral and follow-up. When a homeless individual enters an agency and requests services, the staff will first conduct an intake survey and use the ANCHoR system to enter information about the homeless client, such as the client’s name, age, race, residential history, health status, and employment. Various steps have been taken to try to ensure the client’s privacy. Boston was one of 16 pilot sites that began using ANCHoR in 1996. The system is currently being used by 73 programs throughout the state, of 1 In addition to the approximately 30 cities using or in the process of implementing ANCHoR, as many as 20 other cities are using or are in the process of implementing an estimated 15-18 other similar homeless information systems, according to data provided by a researcher who has worked with several of these cities. Page 27 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts which 45 are in Boston. It is also in the planning stage for an additional 15 programs throughout the state. In 1998, the state Executive Office of Health and Human Services decided to implement ANCHoR throughout Massachusetts as a coordinated statewide system. The statewide expansion of ANCHoR will be supervised and coordinated by the ANCHoR Steering Committee, which was created and appointed by Boston’s Homeless Planning Committee in 1997. The agencies that will use ANCHoR under the statewide expansion include those that provide emergency shelter, transitional housing, referrals, and supportive services to homeless people. Implementing the ANCHoR system statewide is intended to benefit homeless people, agencies that provide services to homeless people, public policymakers, community planners, and researchers, according to the director of the project and state planning documents. Homeless people may benefit by receiving improved assessments of their needs, more coordinated services, and better case management, while the agencies that serve homeless people may benefit by gaining capacity to plan and manage their resources, since they will have better information about patterns of use and resources available to serve homeless people in other parts of the state. According to state planning documents, public policymakers and community planners may also benefit because the system should provide them with information that will improve their ability to coordinate resources communitywide, gauge programs’ effectiveness, assess the overall needs of the community, and, if necessary, request more resources. By implementing the system statewide, Massachusetts hopes to better coordinate care for homeless people, particularly through improving services and case management for individuals who may travel to providers in different locations across the state. According to a state official, the statewide implementation of ANCHoR could be particularly beneficial to Massachusetts because, unlike most states, the state government—rather than municipal or county governments—operates the majority of homeless shelters and the system will give the state more comprehensive data for managing all of these facilities. Linking ANCHoR With an Massachusetts is also the first state that is linking ANCHoR to an automated Automated Benefits benefits eligibility system. When a service provider enters information Eligibility System about a homeless client into ANCHoR, the information is automatically linked to a software program called MicroMax, which has a database of information and eligibility requirements for over 80 federal, state, and local benefit programs, including many specific to Boston and Page 28 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix I Massachusetts Massachusetts. Using the information about the homeless client that has already been entered into the ANCHoR system, MicroMax can develop a report of the public benefit programs and services for which the client may be eligible and calculate the benefits the client would likely receive from each program. Clients can receive individualized documents that include a list of the programs for which they may be eligible, information on where to apply for benefits, and applications for some of these programs that have some of the personal information already filled out. According to state planning documents, several benefits are anticipated from linking the ANCHoR and MicroMax systems. First, case managers using ANCHoR should be better able to identify homeless clients’ eligibility for a variety of programs, including income assistance, medical services, and job training. This information should help link homeless persons more quickly with the mainstream public resources available to them, thereby helping them move more quickly out of homelessness. Second, the ANCHoR-MicroMax link should make the process of applying for mainstream programs easier for homeless people, in part because the system automatically prints out partially completed applications. Finally, the aggregate data obtained from reports generated by the ANCHoR-MicroMax link should provide useful information for planning and policy purposes. For example, the reports will allow the state to track the public resources used by homeless individuals, the number of homeless clients assisted by these resources, and the types and values of the benefits that homeless people received from various programs. Page 29 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington This appendix describes the systems integration initiatives and the Health Care for the Homeless Network of King County, Washington. The communities of Seattle and King County, Washington, were identified by a number of national experts on homelessness as particularly effective in integrating programs that serve homeless people with mainstream programs. The county’s systems integration initiative creates connections between the mental health, substance abuse, and criminal justice systems in an effort to address the multiple and complex needs of many of the county’s homeless in a more coordinated and effective manner. The Health Care for the Homeless Network, as implemented in King County, illustrates how programs can be targeted specifically to the homeless while tapping into existing mainstream resources. About 1.7 million people lived in King County, Washington, in 1998, Background including about 525,000 in the city of Seattle in 1996, according to U.S. Census Bureau estimates. Although personal income in King County is significantly higher than the national average, about 9 percent of the population lived in poverty in 1995, according to the U.S. Census Bureau. King County has a tight housing market—rents are high compared with income, rents have been rising, and the vacancy rate is low. On any given night, about 5,500 people are homeless in King County, according to the Seattle-King County Homelessness Advisory Group. Roughly 54 percent of those that are homeless are single adults, and 46 percent are families or youth. At any given time, an estimated 1,360 homeless people are believed to be living on the street, while most of the remainder are housed in emergency shelters or transitional housing. King County’s homeless population is heavily concentrated in Seattle. Seattle and King County collaborate in developing the Continuum of Care plan for the community and jointly submit a single application to HUD for funding through its McKinney Act programs. The King County government, under contract with the state of Washington, provides most of the county’s supportive services, such as mental health and substance abuse treatment. Within Seattle, the city government provides funding for most of the emergency shelter and transitional housing programs. King County has undertaken a series of initiatives to integrate various Systems Integration in social service systems that serve homeless people. These include efforts to King County integrate the mental health and substance abuse systems, address the Page 30 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington problem of chronic public inebriates, and provide alternatives to county jails for those with mental illness or substance abuse disorders. In addition, Seattle’s participation in HHS’ Access to Community Care and Effective Services and Supports (ACCESS) program has been an important aid to the county’s systems integration efforts. The county defines “systems integration” as the sharing of information, planning, clients, and resources by different social service systems. At the operational level, this means getting different systems, such as the mental health, substance abuse, corrections, and housing systems, to work together in an integrated fashion to provide a continuum of services to their clients. Integration of the Mental The primary focus of King County’s systems integration efforts has been Health and Substance on unifying the county’s mental health and substance abuse systems. Part Abuse Systems of the impetus for this integration is the recognition that many homeless people in the community are dually diagnosed with both mental health and substance abuse disorders. In 1998, the county created the Bureau of Unified Services to stimulate the integration of systems and services for individuals and families suffering from mental illness and/or substance abuse. The county also proposed combining the Division of Mental Health and the Division of Alcoholism and Substance Abuse Treatment Services into a single Mental Health, Chemical Abuse and Dependency Services Division so that the county government’s organizational structure would be better aligned with the integrated systems approach. The county is currently waiting for the County Council to approve this proposed restructuring. As part of its systems integration strategy, King County developed a “no wrong doors” philosophy. This means that persons with mental or addictive illness are offered the services they need whether they seek assistance through a local hospital, detoxification center, emergency shelter, mental health treatment program, or correctional facility. In July 1998, as a pilot project, the county opened the Crisis Triage Unit at Seattle’s Harborview Medical Center. The triage unit is designed to serve as a single place where someone experiencing a behavioral health crisis, particularly related to mental health and/or substance abuse issues, can receive immediate care and referral to other longer-term services. According to county officials, about half of the people who are brought to the triage unit are homeless and more are at risk of becoming homeless. The triage unit is staffed with personnel qualified to assess medical, mental health, and substance abuse conditions, as well as with a housing coordinator, who assists clients in gaining access to short-term housing or Page 31 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington in maintaining existing housing. The triage unit is designed, in part, to divert people from local jails or hospitals, where they might otherwise be taken, to more appropriate housing and treatment situations. Efforts to Address the In September 1997, King County began searching for solutions to the issue Problem of Homeless of chronic public inebriates. These individuals are usually homeless Public Inebriates chronic abusers of alcohol who often have secondary problems with drug abuse or mental illness. The county convened a Chronic Public Inebriates Systems Solutions Workgroup, which included representatives from the city and county governments, the business community, homeless assistance service agencies, and other affected parties. This effort stemmed, in part, from a recognition in the community that many of the severely distressed individuals in this population were repeatedly entering certain parts of the county’s systems, such as hospital emergency rooms and the courts, where their conditions could not be appropriately addressed. In December 1997, the workgroup developed a housing plan that recommended a series of policy changes and housing actions to help address the needs of chronic public inebriates living on the streets, as well as reduce the negative effects of this population on the community. The actions taken thus far have included opening a sobering sleep-off center, reaching agreement with downtown merchants not to sell certain alcoholic products favored by street inebriates, improving outreach services, and taking steps to develop more supportive housing units for this population. Alternatives to Jail for Beginning in 1985, in response to concerns that the county’s jails Offenders With Mental contained large numbers of mentally ill inmates whose needs would be Illness and Substance better addressed through treatment, King County developed several jail diversion projects. These projects sought to prevent recidivism among Abuse Disorders mentally ill offenders–a large percentage of whom were homeless–by providing them with increased services and intensive case management as an alternative to incarceration. In 1997, these projects were redesigned, resources for treatment were increased, a housing component was added, and for the first time, persons whose primary disorder was substance abuse were included in the project. These projects were jointly funded by the county agencies overseeing criminal justice, detention, mental health, and substance abuse services, as well as by the city of Seattle. Page 32 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington In December 1998, the county replaced these jail diversion projects with the Mental Health Court, a pilot project that incorporates many of the elements of the prior projects. Defendants with mental illness who have been charged with misdemeanors can now choose to have their cases heard in a special court, where they typically receive court-ordered treatment as an alternative to prosecution or sentencing. On the basis of past experience, King County officials expect that about one-third of those using the Mental Health Court will be homeless and many more will be at risk of becoming homeless. Integration Efforts According to King County officials, an important aid to their systems Stimulated by Participation integration efforts has been Seattle’s participation in the HHS’ ACCESS in ACCESS Program program. ACCESS is a 5-year demonstration program that began in 1994 and will end in 1999. The goal of ACCESS is to evaluate the impact of systems integration on the provision of services for homeless people who are severely mentally ill. Eighteen sites—nine control sites and nine experimental sites—in nine states across the country were selected to participate in the ACCESS program. Seattle is home to both a control site and an experimental site, located in different parts of the city. Both Seattle sites received resources to fund services for homeless people who are mentally ill, and the experimental site received additional resources to fund activities designed to enhance systems integration. This included the hiring of a full-time systems integration administrator within the King County Department of Community and Human Services and the creation of working groups designed to improve collaboration and communication between provider agencies and the community. Although the ACCESS program will end this year, a county official told us that the county is “institutionalizing” the lessons learned from the program through the creation of a new Homeless Outreach, Stabilization and Transition Program, which will incorporate many of the systems integration activities that were provided under ACCESS. The goal of HHS’ Health Care for the Homeless program is to make Seattle-King County’s high-quality health care accessible to homeless people nationwide. The Health Care for the program awards grants to local public or private nonprofit organizations Homeless Network to provide health care services to the homeless. In fiscal year 1998, the Health Care for the Homeless program funded 128 projects nationwide that were administered by local public health departments, community Page 33 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington and migrant health centers, hospitals, and local community coalitions. According to HHS, the program encourages an interdisciplinary approach that incorporates health, mental health, substance abuse, and social services to build a coordinated network of services for homeless people within a community. Health Care for the Homeless projects throughout the country are implemented through a variety of different organizational models. About half of the projects are housed in community health centers, about 25 percent in public health departments, and the remainder in other organizations, such as nonprofit agencies, hospitals, and shelter coalitions. The Health Care for the Homeless Network (HCHN) model in Seattle-King County combines services provided directly by the county’s public health staff with contracted services provided by mainstream health care providers. Several national experts on homelessness told us that Seattle-King County’s HCHN was particularly effective. However, Seattle-King County’s model is one of many that have been successful and experts say that the most appropriate model for implementing Health Care for the Homeless in any given location will depend on the specific needs and characteristics of the particular community. County and Mainstream Seattle-King County’s HCHN is administered by the Seattle-King County Services Linked Through Department of Public Health, which provides certain services directly to HCHN homeless people and contracts with mainstream health care providers for other services. Services provided directly by Department of Public Health staff include immunizations, family planning, dental screening, tuberculosis outreach, communicable disease control, and health education. Most of these services are provided at sites operated by the department. The Department of Public Health also has a full-time public health nurse available to provide technical assistance on health and safety issues to agencies that serve homeless people. For example, the public health nurse provides training to staff in emergency shelters on first aid and disease prevention. The Department of Public Health also provides emergency shelters with certain supplies, like soap and liquid soap dispensers, to help improve the general hygiene of their homeless clients. The Department of Public Health contracts with 10 community-based health care providers, including hospitals, community health centers, and social service agencies, to provide most of the network’s services. These services include street outreach, primary care, substance abuse and mental health services, medical respite, and assistance with enrollment Page 34 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington and the use of Medicaid managed care. To be more accessible to the homeless population they serve, most of the health care professionals working for the community-based health care providers are located at emergency shelters. Benefits of County’s HCHN The Health Care for the Homeless model implemented in Seattle-King Model County has a number of benefits, according to local officials. These include better access to city-, county-, and community-based resources; more continuity in the provision of services to homeless people; and improved data collection capabilities that can help city and county governments better plan services for homeless people. According to a Seattle official, placing the Seattle-King County HCHN within a major government agency like the Department of Public Health, rather than in a community-based nonprofit service agency, improves its access to the community’s major health care resources. At the same time, by contracting with community providers for health care services, HCHN is able to tap into existing mainstream resources, such as hospitals and community health centers, without having to create a separate system of care for homeless people. The requirements in HCHN’s contracts with providers in mainstream systems also allow HCHN to hold these systems more accountable for serving homeless people, who are traditionally a more difficult and expensive population to serve. These requirements also ensure that mainstream systems provide the special outreach and support that the homeless population requires. Moreover, components of Seattle-King County’s HCHN help to ensure continuity of care for homeless people as they move from location to location, and even after they move out of homelessness. Under the Pathways Home program, a team of health care professionals track and monitor homeless families–whether they are living on the street, in an emergency shelter, or in temporary housing–and continue to provide them with the range of health care services that they need, from screening and case management to comprehensive mental health treatment. The team provides health care to these clients for up to a year after they have been placed in permanent housing. Finally, the Seattle-King County HCHN has in place a data system that provides important information on homeless people and the services they are receiving. Each provider that contracts with HCHN records every encounter with a homeless client on a standardized intake form. All of the Page 35 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix II King County, Washington data are maintained on a central database, and each homeless client is given a unique identification number that allows HCHN to track the client throughout the system. According to a program official, the Seattle-King County HCHN database has recorded about 60,000 encounters with about 20,000 individuals in the past year. This information aids city and county governments in identifying the major health problems affecting homeless people, as well as in monitoring general health and demographic trends among this population. Page 36 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio This appendix describes the efforts of Franklin County, Ohio, to integrate its emergency shelter programs and related homeless assistance services into a coordinated and unified system, primarily through its Community Shelter Board. Franklin County, which includes the city of Columbus, was identified by a number of national experts on homelessness as particularly successful in getting communitywide support for its homeless assistance programs, coordinating its emergency shelter resources, and reducing the administrative burden on providers. The county also has a management information system that allows it to collect client-specific data in a uniform fashion across the entire emergency shelter system. The Community Shelter Board, with guidance from the state, is using these data to develop and measure programs’ outcomes so that it can better manage homeless assistance programs and services. The population of Franklin County, in central Ohio, was just over 1 million Background in 1998, according to a U.S. Census Bureau estimate. The majority of the county’s population resides in Columbus, which in 1996 had a population of about 660,000. Franklin County has a fairly strong economy and relatively low unemployment. While housing costs are lower than those of many other metropolitan areas nationwide, the county has a shortage of affordable housing for low-income residents and a substantial waiting list for subsidized housing. During 1998, 840 shelter beds served 8,911 homeless individuals in Franklin County. In addition, there were 1,042 transitional housing beds in the county. The number of families needing emergency shelter has risen significantly in the past several years. Currently, about half of the people that use the county’s emergency shelters are families with children and half are single adults, whereas in the past most of the homeless were single adult men. The county’s homeless population is heavily concentrated in Columbus. The Community Shelter Board (CSB) is a nonprofit organization that coordinates and administers most of the government and private funding for Franklin County’s emergency shelters and certain related services for homeless people. In its fiscal year ending March 1999, CSB budgeted about $4.8 million to help fund 11 agencies. About two-thirds of this funding was used to support adult and family shelter programs, and most of the remaining funds were used for homeless prevention programs, housing resource programs, technical assistance, research, and special services. CSB receives funds from both government and private sources, including Page 37 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio the city of Columbus, Franklin County, the state of Ohio, HUD, the United Way, and private donations. CSB also coordinates the Continuum of Care planning process for Franklin County. The Franklin County Department of Human Services, which is supervised by the Ohio Department of Human Services, provides certain supportive services that benefit low-income people in the county, such as income support programs and Medicaid. The Alcohol, Drug and Mental Health Board of Franklin County, which is funded and overseen by the Ohio Department of Mental Health and the Ohio Department of Alcohol and Drug Addiction Services, contracts with 52 agencies to provide the county’s mental health and substance abuse services. In addition, the Ohio Department of Development administers a variety of state- and federally-funded programs that benefit homeless people statewide, including the Emergency Shelter Grants and Supportive Housing for the Homeless programs. The Coalition on Homelessness and Housing in Ohio, a nonprofit agency, coordinates the statewide Continuum of Care planning process and provides advocacy, technical assistance, training, and some direct assistance to state agencies and homeless service providers. The Community Shelter Board serves as an intermediary between funding Coordination of sources and the nonprofit agencies that provide emergency shelter and Emergency Shelter related services to homeless people in Franklin County. Many of the and Other Services government officials, advocates, and providers of services for homeless people that we spoke with–at the county, state, and national Through the levels–described CSB as a highly effective organization. They noted that its Community Shelter distinctive role allows it to plan countywide shelter services and foster successful collaborations between the various players and systems that Board serve the homeless in Franklin County. Benefits of Intermediary CSB is neither a government agency nor a direct provider of services to Role homeless people; instead, it functions as an intermediary between the sources that fund shelter services and the agencies that provide these services. As a result, CSB benefits from the community’s perception that it is a neutral body that is not unduly influenced by either local government politics or service providers’ agendas. For example, CSB receives most of its funding from government sources; however, because it is a private nonprofit agency, it is perceived as somewhat immune to local politics when making funding and planning decisions. Moreover, because CSB itself Page 38 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio does not provide direct services to homeless people and because the agencies that it funds do not have positions on its board of trustees, it is able to represent the interests of homeless people and yet avoid the conflicts that might occur if funding and planning decisions were seen as based solely on the interests of the agencies it funds. In addition, because CSB was founded by a group of local businesspeople and has a number of business and civic leaders on its board of trustees, it has been able to attract a high level of support and participation from the local business community. Benefits of Centralized CSB serves as the single organization that coordinates and plans all shelter Structure services in Franklin County and coordinates the county’s Continuum of Care planning process. In this role, CSB can ensure that all of these services and programs are considered as part of a whole “system” that works together rather than as a fragmented set of independent resources. According to community officials, CSB has provided a centralized structure for what was previously a decentralized set of community-based services and programs. An example of the benefit of this centralized structure is CSB’s work on a plan to address the needs of homeless men who live in an area of Columbus called the Scioto Peninsula. Half of the city’s single men’s shelter beds are located in this area, and many of the city’s street homeless people reside there. In 1997, the city asked CSB to develop a plan to address the needs of the large number of homeless men who would be affected by development planned for the area. CSB coordinated the Scioto Peninsula Relocation Task Force, which used the Scioto Peninsula issue as an opportunity to conduct a more comprehensive review of the needs of all single adult homeless men in Columbus and Franklin County. The task force’s resulting report serves as a strategic plan that incorporates all of the various systems and resources required to address the needs of this population, including emergency shelters, permanent housing, and supportive services. Benefits of a Single CSB serves as a single conduit for funding from a variety of different Conduit for Funding sources, thus reducing the administrative burden for the community-based service providers who receive these funds. CSB receives funds from a number of sources, including city and county general tax funds, the federal Emergency Shelter Grants and Community Development Block Grant programs, the Ohio Housing Trust Fund, the United Way and other public Page 39 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio and private sources. Service providers apply directly to CSB for these funds rather than to the funding sources. CSB determines, on the basis of program evaluations, eligibility requirements, and other considerations, how much and which funds each provider will receive from each source. Each provider signs a contract with CSB ensuring that it will comply with any program requirements associated with the funds it receives. This “one stop” blended funding process lessens the administrative burden placed on service providers in several ways. First, it reduces the number of funding applications they have to complete. Second, it reduces the need for them to keep track of the differing reporting and fiscal year requirements used by different funding sources. Finally, it can help ease cash flow problems that service providers may face. For example, as a financial intermediary, CSB is in a position to advance money to providers who have been awarded grants but have not yet received the money. CSB collects both client-specific and systemwide data from Franklin Data Collection and County’s emergency shelter system. These data are used in a variety of Program Evaluation ways for planning, policy analysis, evaluation, and needs assessment for Efforts in Franklin homeless assistance programs. County Uniform, Systemwide Data CSB has implemented a management information system to collect Collected comprehensive, uniform data from the entire emergency shelter system in Franklin County. CSB stipulates in its contract with each of the county’s emergency shelters what types of data must be collected on homeless clients. A standardized intake form is used by each shelter and includes questions about basic client demographics, as well as income and benefits and the reasons for homelessness. The information is collected and entered into CSB’s centrally located management information system. CSB officials said that although the computer system and software itself are somewhat dated (there are plans to move to a more modern Windows-based system in the near future), the information management system has allowed them to develop a uniform historical database that includes information from all of the county’s shelters on the clients they have served since 1991. Data Used for Managing, The data collected by CSB from emergency shelters in Franklin County are Planning, and Evaluating used in a variety of ways to better manage the resources available in the Services community to serve homeless people. For example, CSB’s management Page 40 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio information system can provide information on the use of shelter beds over time, both at individual shelters and systemwide, allowing CSB to track the use of these scarce resources. The system can also provide information on the demographics of shelter clients and the patterns of shelter use over time. This information has been used by the community for policy development, needs assessment, and planning. In addition, because each shelter client is given a unique identifying number, individual clients can be tracked across time as they move through the system and are referred to different service agencies. CSB can thus develop an unduplicated count of how many people are using shelters and analyze the movement of clients from program to program. The Scioto Peninsula Task Force used CSB’s historical database to analyze patterns of use of the men’s shelter system. They found that 15 percent of the city’s homeless men used 56 percent of the shelter system’s resources, while the remaining 85 percent of the men entered the system transitionally for relatively short stays. In addition, CSB found that the long-term users of the shelter system often needed other services, such as mental and physical health services or substance abuse treatment. To meet these needs, the task force’s final plan recommended that the city and county develop service-enriched supportive housing for long-term users of the system, thereby freeing shelter resources for those requiring shelter for only a short period of time. The state of Ohio has started to develop performance standards that are State’s and County’s intended to measure programs’ outcomes and improve the provision of Use of Outcome services to homeless people. In Franklin County, CSB has been working Measures to Improve with the state to establish outcome measures for the service providers it funds. Programs for Homeless People Like some other state housing agencies nationwide, the Ohio Department of Development has recently started to use outcome measures for its housing programs that serve homeless people. State officials told us that their intent is to improve the quality of the programs they fund by focusing more on results—such as moving people out of homelessness–rather than on outputs–such as the number of units of service delivered. Like many other private and government organizations that provide funding for homeless programs, the state wants to ensure that it is getting the best results for its dollars. State officials believe that the use of outcome measures will encourage poorly performing agencies to improve their programs, as well as identify the “best practices” of providers who are Page 41 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix III Franklin County, Ohio meeting their outcome goals and can provide replicable models for other agencies to use. In 1998, the Ohio Department of Development began a pilot project under which agencies that receive state supportive housing grants were required to develop outcome-based performance targets and were to be held accountable for meeting their outcomes. All 53 of the department’s supportive housing grantees have attended special training seminars that were intended to clarify and provide guidance on how outcome measures and goals should be developed.1 As their efforts progress, state officials told us, they hope to refine their benchmarks and set individualized outcome measures that better reflect the nature of each grantee’s work and the population the grantee serves. For example, the general outcome measure for a transitional housing program might be the percentage of clients who move into some kind of permanent housing after a certain period of time. However, an agency that serves a more difficult population, such as the mentally ill, would not be expected to have the same success rate as an agency that serves a population with fewer barriers to becoming self-sufficient. In Franklin County, CSB has been working with the Ohio Department of Development to establish outcome measures for the service providers it funds. For emergency shelters, these outcomes include success in moving clients out of shelters and into more appropriate housing, such as transitional housing. For a transitional housing program, the outcomes measured include occupancy rates (to ensure that resources are being fully used), length of stay (to ensure that clients are not staying too long without moving forward), and the percentage of clients that move to permanent housing. CSB’s management information system is able to provide the data needed to measure many of these outcomes. It does not, however, follow up on clients after they leave the homeless service system altogether. 1 The training session was provided by the Rensselaerville Institute, a not-for-profit institute that provides consultation services to government and nonprofit organizations on performance and outcome management. Page 42 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota This appendix describes Minnesota’s use of outcome measures, data collection, and program evaluation to address the problem of homelessness in the state. National experts on homelessness with whom we spoke consistently identified Minnesota as especially active and innovative in evaluating its programs for homeless people and using outcome measures to manage these programs. In particular, Minnesota’s Family Homeless Prevention and Assistance Program provides communities with flexible grants but uses outcome measures to hold providers accountable for achieving results. Minnesota also conducts a comprehensive statewide survey of its homeless population, which is used to assess the needs of, and plan programs for, homeless people. In addition, Minnesota conducts a relatively large number of evaluations to measure the effectiveness of specific homeless assistance programs. Minnesota had a population of about 4.7 million in 1998, of whom about Background 2.8 million lived in the Minneapolis-St. Paul metropolitan area in 1996, according to U.S. Census Bureau estimates. The state has expressed concerns about a shortage of affordable housing, particularly in the metropolitan area, where the economy is relatively strong but the housing market is tight, with a rental vacancy rate of about 2 percent. There are also concerns about a lack of affordable housing in smaller communities outside the metropolitan areas where employment is growing. A statewide survey in October 1997 found that about 5,590 persons were homeless in Minnesota on a given night. More than three-quarters of the homeless individuals in temporary housing were women and children. The number of homeless families in Minnesota has increased significantly since 1991. About 82 percent of the homeless individuals live in the Minneapolis-St. Paul metropolitan area, while the remaining individuals live in other parts of the state, known as Greater Minnesota. The Minnesota Housing Finance Agency funds and administers several state homeless service and prevention programs, coordinates the Continuum of Care plan for Greater Minnesota, and convenes the state’s Interagency Task Force on Homelessness. The task force is composed of representatives from a variety of state agencies and helps coordinate and administer state programs specifically targeted for homeless people. The state’s Department of Children, Families, and Learning administers the state’s federally funded Emergency Shelter Grant program, as well as other programs that serve homeless people. Individual county governments—especially Hennepin County, which includes Minneapolis, Page 43 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota and Ramsey County, which includes St. Paul—also provide housing and services to homeless people. Municipal governments in Minnesota play a limited role in providing or funding services for homeless people. Minnesota uses outcome measures to manage its Family Homeless Minnesota’s Family Prevention and Assistance Program (FHPAP). The state expects agencies to Homeless Prevention meet the outcomes set for their programs and, in return, gives the agencies and Assistance considerable flexibility in using program funds. Program FHPAP is a state-funded program whose goals are to (1) prevent homelessness, (2) reduce the length of time people stay in emergency shelters, and (3) eliminate repeat episodes of homelessness. The program is targeted primarily to homeless families and provides funding for such things as short-term rental assistance, security deposits needed to secure housing, and housing search services. FHPAP is administered by the Minnesota Housing Finance Agency in conjunction with the state’s Interagency Task Force on Homelessness. The state legislature provided $6.05 million for the program for the 1997-99 biennium, according to a state official, during which time it awarded 16 grants. In the Minneapolis-St. Paul metropolitan area, FHPAP made grants to county agencies, which generally distributed the money to the community-based nonprofit service providers that were the subgrantees. In Greater Minnesota, FHPAP has usually provided grants directly to nonprofit organizations. FHPAP grants are very flexible, and grantees have considerable leeway in spending the funds. However, grantees are required to (1) set specific performance goals and outcome measures that are consistent with each program’s objectives, (2) develop a method for tracking these outcomes, and (3) achieve and report on the outcomes they have set. Each of these requirements is described below. Setting Goals and Measures. When applying for program funds, grantees must state specific, measurable outcome goals for their projects that relate to FHPAP’s three overall goals. The agencies must include the time frames within which these goals will be achieved. For example, a program for preventing homelessness might state that 90 percent of the families and youth that participate in the program will be in stable housing 6 months after they leave the program. According to a program official, the program allows outcome goals to be set by grantees rather than by the state, partly because conditions vary so greatly in different parts of the state. Page 44 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota Tracking Outcomes. FHPAP grantees are required to develop methods of tracking and measuring their programs’ outcomes. For example, a grantee may choose to conduct follow-up phone surveys with families that have received assistance through a program or review clients’ records at an emergency shelter to measure how long the clients stay in the shelter. Hennepin County, which had 28 FHPAP subgrantees in fiscal year 1998, has developed its own data system for tracking purposes. This system provides all of the subgrantees with software that allows them to collect basic demographic and outcome information on clients. These data are later entered into a centralized data management system administered by the county. The system assigns each client a unique identifier, which allows the county to evaluate programs’ outcomes by determining, for example, how many of the clients who are enrolled in a homeless prevention program are staying at an emergency shelter. Achieving and Reporting Outcomes. Each FHPAP grantee is required to submit a quarterly and an annual report to the state that provides programs’ overall results and outcome data for individual clients. As long as providers successfully achieve the outcome goals they have set for their programs, the state does not specify how they must spend their FHPAP funds. A state official told us that this flexibility benefits service providers because it reduces their administrative burden, allows them to tailor their programs to local needs and situations, and gives providers the freedom to try new ways of preventing homelessness. In addition, the results reported by the service providers have helped the state revise the program on the basis of what has proved to be effective or ineffective in addressing homelessness. For example, a state official told us that service providers no longer use FHPAP funds for long-term rental assistance because outcome information from past programs showed that this was not a cost-effective way of serving a large number of people. Minnesota has been conducting a statewide survey of its homeless Minnesota’s Statewide population since 1991. Although other states count and survey their Survey of Homeless homeless populations, Minnesota’s survey is notable because it is People comprehensive and has been conducted every 3 years. Minnesota conducted comprehensive surveys of the state’s homeless population in 1991, 1994, and 1997, and plans another survey in 2000. These surveys were commissioned by Minnesota’s Interagency Task Force on Homelessness and were conducted, under contract, by the Wilder Page 45 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota Research Center.1 The surveys were funded jointly by state agencies, including the Minnesota Housing Finance Agency and the Minnesota Department of Human Services, as well as by nonprofit service providers and private foundations. The cost of the most recent survey was about $100,000. For the 1997 survey, more than 440 trained volunteers surveyed homeless individuals at 150 different agencies that serve homeless people in 48 cities, as well as 18 street locations in 8 cities. The survey identified 5,590 people as homeless on one particular night, including people in emergency shelters, transitional housing, and battered women’s shelters, as well as living on the street and in other nonshelter locations. Separate surveys were conducted for adults (including families) and for unaccompanied youth. The surveys not only produced a statewide count of the homeless but also provided comprehensive data on the characteristics of the homeless population. Adults and youth in shelters and transitional housing, as well as those living on the street, were asked a detailed set of questions covering demographics, income, shelter use, housing, employment, substance abuse, and mental and physical health. State and local officials have used the results of these surveys for a variety of purposes in planning their programs for homeless people. For example, because the surveys have been conducted at regular intervals, state policymakers and others have been able to use the results to gauge trends in the homeless population over time. One trend that the surveys have shown is a significant and steady increase in the number of homeless families and in the proportion of the overall homeless population that families represent. The surveys have also documented a rise in the percentage of homeless people who are employed. According to an official at Wilder Research Center, this suggests that homelessness in Minnesota may be increasing more because of a shortage of affordable housing than because of a lack of income sources. Officials from the Wilder Research Center and two of the organizations that funded the survey told us that two of the primary uses of the survey results are to help persuade lawmakers and others of the need for more resources and to help prepare grant applications. For example, city planners often use the data from the survey when they write grant proposals, and state agencies and providers use the information to support their requests for more resources. One official stated that the results of the 1 The Wilder Research Center is the research arm of the Wilder Foundation, a private nonprofit foundation that focuses on social welfare issues in the St. Paul metropolitan area. Page 46 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota surveys were a factor in convincing the state legislature of the need to create the Family Homeless Prevention and Assistance Program. The information collected through the surveys is also useful in assessing the needs of and in planning programs for homeless people, according to a state official. For example, when survey data indicated an increase in the number of unaccompanied homeless youth (i.e., children who are not with their parents), communities increased their efforts to address the needs of this population in their Continuum of Care plans. Minnesota has also conducted a number of evaluations to determine the Minnesota’s effectiveness of some of its programs for homeless people. Some of these Evaluations of are described below. Programs That Serve Evaluation of the Supportive Housing Demonstration Program. The the Homeless Minnesota Supportive Housing Demonstration Program provided $2.2 million in state funding for 180 supportive housing units for people with mental illness, substance abuse disorders, or HIV/AIDS who were either homeless or at risk of becoming homeless. The project used a portion of the funds that would normally have been used to provide institutional care (such as in group homes) for these people and allowed the money to be used more flexibly to provide them with supportive housing (independent housing with supportive services). In June 1998, the Wilder Research Center published a 1-year evaluation report on the demonstration project. The report evaluated (1) the effectiveness and quality of the supportive housing and services provided and (2) the cost- effectiveness of this supportive housing compared with that of the housing and services provided in other institutional settings. Officials at the Corporation for Supportive Housing, which coordinates the demonstration project, said that the Wilder evaluation was the first study that ever quantified and compared the cost of supportive housing with the costs of alternative public-sector service systems. The cost of the housing and services provided by the demonstration’s supportive housing were compared with the costs that the public sector would have incurred to provide these residents with shelter and services. Public-sector costs were estimated from data provided by systems such as the state criminal justice system (for costs associated with correctional facilities), county detoxification centers (for costs associated with providing detoxification services), and the state Department of Human Services (for costs associated with prior residential care, hospital stays, General Assistance Page 47 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix IV Minnesota grants, and other social service grants). The evaluation reported that, compared with the other systems, the demonstration project improved the quality of life for participants and reduced the costs of caring for them. Anishinabe Wakiagun. Anishinabe Wakiagun is a permanent supportive housing program for Native American men and women who are chronic inebriates. The goal of the project is to provide a safe and stable alternative to the street for this population, while improving the civic atmosphere and reducing the amounts of money spent on detoxification units, emergency rooms, and jails. The project opened in September 1996 and is located in Minneapolis. The Hennepin County Office of Planning and Development evaluated the Anishinabe Wakiagun program for the period from September 1996 through March 1998. As part of this evaluation, the following two outcome goals were analyzed: (1) reducing the population’s use of detoxification and emergency rooms and (2) stabilizing the population’s housing status. For each of the residents, the evaluation compared their history 1 year before they were admitted into the program with their status while they were in the program. It evaluated data on their use of detoxification units, use of hospital emergency room facilities, and booking in the adult detention center. Other Evaluations. The Wilder Research Center has also conducted or is conducting the following evaluations of other homeless assistance programs in Minnesota: • A 6- and 12-month follow-up evaluation of homeless people who are currently living in transitional housing. The objective of the evaluation is to gauge the effectiveness of transitional housing in moving homeless people into permanent housing. • An evaluation of what happens to youth once they have left Project Foundation, an emergency shelter for homeless youth in Minneapolis. • An evaluation of Rebuilding Our Own Futures (ROOF), a transitional housing program for families. The study evaluated outcome measures such as participants’ success in obtaining permanent housing, increasing income, and maintaining children’s school attendance. Page 48 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix V Comments From the Department of Health and Human Services Note: GAO comments supplementing those in the report text appear at the end of this appendix. Page 49 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix V Comments From the Department of Health and Human Services See comment 1. Page 50 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix V Comments From the Department of Health and Human Services See comment 2. See comment 3. Page 51 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix V Comments From the Department of Health and Human Services Page 52 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix V Comments From the Department of Health and Human Services The following are GAO’s comments on the Department of Health and Human Services’ (HHS) letter dated June 9, 1999. 1. We agree that HHS has undertaken several initiatives to address GAO’s Comments homelessness; however, we disagree with the Department that our reporting of state and local officials’ perceptions about its lack of involvement in addressing homelessness is not adequately substantiated or lacks specificity and documentation. The observations we have reported are based on interviews we conducted with more than 50 state and local officials in four different locations across the country. The consistent nature of their comments clearly suggests that many at the state and local level believe that HHS needs to do more to address the needs of homeless people. 2. We agree that there is a need to obtain more information on the barriers created by federal, state, and local policies. This information can be used by federal agencies to better coordinate their efforts and help them implement changes that can eliminate some of these barriers. However, this issue was not within the scope of this assignment. We plan to address this issue in a future review. 3. As we stated in comment 1, HHS has made some efforts in this area, but, according to our review, they are not perceived as adequate by some state and local officials. Page 53 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix VI Comments From the Department of Housing and Urban Development Note: GAO comments supplementing those in the report text appear at the end of this appendix. See comment 1. See comment 2. Page 54 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix VI Comments From the Department of Housing and Urban Development Page 55 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Appendix VI Comments From the Department of Housing and Urban Development The following are GAO’s comments on the Department of Housing and Urban Development’s (HUD) letter dated June 2, 1999. 1. We revised the report to clarify the role of the Interagency Council on GAO’s Comments the Homeless. 2. After reviewing HUD’s comments, we deleted the sentence cited because it was not the primary concern of the state and local officials with whom we spoke. (385760) Page 56 GAO/RCED-99-178 Integrating and Evaluating Homeless Assistance Programs Ordering Information The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. Orders by mail: U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists. For information on how to access GAO reports on the INTERNET, send an e-mail message with "info" in the body to: email@example.com or visit GAO’s World Wide Web Home Page at: http://www.gao.gov PRINTED ON RECYCLED PAPER United States Bulk Rate General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. G100 Official Business Penalty for Private Use $300 Address Correction Requested
Homelessness: State and Local Efforts to Integrate and Evaluate Homeless Assistance Programs
Published by the Government Accountability Office on 1999-06-29.
Below is a raw (and likely hideous) rendition of the original report. (PDF)