Toward Understanding Homelessness: The 2007 National Symposium on Homelessness Research. Employment and Income Supports for Homeless People

By:
David Long, MA, MPP, Abt Associates Inc., Bethesda, MD
John Rio, MA, CRC, Advocates for Human Potential, Cypress, TX
Jeremy Rosen, National Policy and Advocacy Council on Homelessness, Alexandria, VA

Abstract

In this paper, the authors synthesize the findings of recent studies examining the role of mainstream programs such as Social Security Administration (SSA) disability programs, Temporary Assistance for Needy Families (TANF), and Workforce Investment Act (WIA) initiatives in enhancing employment and incomes for people who have experienced homelessness. They also describe the design and outcomes of targeted programs designed specifically to address employment and income support for people who are homeless. While some rigorous evaluations have been done on mainstream programs, the effects of the interventions on the subpopulation that has been homeless are often not addressed. Few rigorous studies have been done on targeted programs. The authors draw several conclusions from the available evidence and outline future research directions to fill important gaps in the research literature.

Introduction

Policymakers face many fundamental questions in deciding how government should respond to the immediate needs of homeless families and individuals  a place to live and the food, clothes, health care, and other items they require. Mainstream federal programs, such as disability assistance, welfare, Medicaid, and assisted housing, provide income and in-kind assistance to meet the needs of low-income people, including those who are homeless. For two reasons, these same programs also provide both incentives and a range of services to encourage employment. First, any assistance that is income conditioned can have the unintended consequence of discouraging work, because earnings reduce or eliminate the assistance for which people qualify. Second, programs that try to meet individuals immediate needs also seek to respond to individuals longer-term need and desire to pursue economic independence. As a result, employment and income supports are closely linked in the mainstream programs that provide the bulk of government assistance to people who are or have been homeless.

This paper synthesizes the findings of recent studies examining the efforts of mainstream programs to increase employment and also reduce immediate hardships among families and individuals who are homeless. These findings indicate that mainstream programs  those serving large populations of which homeless individuals are a small segment  meet many of the basic needs of some groups of homeless people, but struggle to promote employment by these groups while continuing to provide income and other support. They also show that other groups of people who are homeless receive little income or employment support. This is partly due to the difficulty that some individuals and families have in accessing mainstream programs. More fundamentally, however, most mainstream programs do not serve adults unaccompanied by dependent children unless the former have severe and documented disabilities.

This paper also considers programs specifically targeted to homeless people. The homeless assistance programs funded by the U.S. Department of Housing and Urban Development include employment services as fundable activities. The Department of Health and Human Services funds a Transitional Living Program for homeless youth that provides an array of services, including job attainment skills. The Department of Labor funds a Homeless Veterans Reintegration Program, emphasizing employment. In addition, mainstream federal employment programs and demonstrations have particular local grantees that target homeless people. However, the available research on such employment initiatives targeted to homeless people is limited to descriptive reviews and qualitative studies that do not provide hard evidence on the effectiveness of the employment services. We are not aware of experimental studies specifically focused on evaluating employment strategies for any homeless population. In contrast, several experimental investigations have addressed the mainstream programs that serve larger populations, including people who are or have been homeless, and we attempt to tease out of those evaluations implications for homeless people.

Our review of the research evidence is organized by federal agency.^[1]  While the federal government has undertaken efforts to coordinate policies for people who are homeless  through state Policy Academies funded by HHS, through HUDs continuum of care planning and application process, and through the Interagency Council on Homelessness  each agency responds to the segment of the overall homeless population for which it is responsible, providing (or funding state and local governments to provide) the types of income and in-kind support Congress has mandated. For example, the Social Security Administration serves people with disabilities;^[2] the Department of Veterans Affairs serves veterans, particularly veterans with disabilities; the Department of Health and Human Services provides income support for families with children and, through the Substance Abuse and Mental Health Services Administration (SAMHSA), funds activities that support the resilience and recovery of people with mental illness and substance use disorders. HUDs mainstream housing assistance programs serve all types of low-income households, as does the Food Stamps program of the U.S. Department of Agriculture. The federal governments most powerful work support tool, the Earned Income Tax Credit (EITC), is targeted largely to families with children and administered by the Internal Revenue Service.

The Department of Labor is the lead federal agency for supporting employment across the population spectrum. However, because income support and in-kind benefits programs may create disincentives to working, each federal agency that serves low-income people has produced its own work-support strategy, leading to an array of federal programs that have employment as their objective. Those programs often focus on people who face barriers to work  as do many homeless individuals and families. Therefore, before turning to our agency-by-agency review of the research literature, we describe the barriers to employment that may challenge the capacity of mainstream programs to support the work efforts of homeless people.

Barriers to Work Faced by Homeless People

There are assertions and supporting evidence throughout the research literature that all segments of the homeless population  unaccompanied adults, heads of family households, and youth  face significant and multiple barriers to employment. These barriers are personal, programmatic, and systemic. People who are homeless often lack skills in stress management and social interaction, independent living skills, and skills for vocational engagement (Munoz, Reichenbach, & Hansen, 2005), as well as a place to live and financial resources. Barriers such as lack of transportation and educational credentials are prevalent among homeless people in both urban and rural areas (Taylor, 2001). In addition, homeless young adults and youth experience high levels of trauma and typically have poor educational and vocational preparation (Barber et al., 2005).

Mental health and physical health play central roles in the employment and program participation of people who are homeless or at risk for homelessness. Disabilities are well-documented barriers to employment, although the extent of the hindrance varies. For example, the employment of persons with schizophrenia is impeded by a range of specific clinical problems. People with schizophrenia who have greater cognitive impairment experience more difficulty in the labor market and require more vocational support than those with lesser impairment (McGurk et al., 2003).

Substance use disorders, alone or in combination with disabilities, substantially reduce the income people receive from work (Zuvekas & Hill, 2000). Competitive employment is further impeded by receipt of disability payments (and concomitant adverse work incentives) and by race (Rosenheck et al., 2006). Among homeless people with severe mental illness, those with a history of incarceration have more serious problems and show less improvement in community adjustment domains (McGuire & Rosenheck, 2004). Incarceration can decrease the types of employment available after release from jail or prison, and a history of incarceration has been shown to alter how homeless ex-offenders conduct job searches (Cooke, 2004).

The barriers faced by homeless families are generally similar to those of other low-income families, including families on welfare. The key issues are transportation, child care, educational limitations, and substance abuse (Burt & Anderson, 2005; Burt, Aron, & Lee, 1999; Taylor, 2001). Severe mental health problems and histories of incarceration are less common for homeless family heads than they are for homeless adults who are unaccompanied.

In addition to these barriers, the digital divide remains a deep chasm for homeless populations. Competing for jobs today requires some understanding of and comfort and competency with information technology. Miller and colleagues (2005) identified the lack of such facility among homeless men as an important barrier to employment. Because they lacked computer knowledge and feared failure, the majority of study participants had not sought to use computers available through public access.

These limitations help to produce poor labor market outcomes for homeless people. Unemployment among homeless populations is widespread, and the problem is especially great during economic downturns. For example, at the end of 2002, there were 3.2 unemployed workers for every job opening, compared to 1.3 at the end of 2000 (Bernstein & Chapman, 2003), and low-wage job seekers, including people experiencing homelessness, suffered as a result.^[3]  In addition, the jobs that homeless people and tenants of supportive housing most frequently secure are low paying  laborer positions, jobs in the services sector (including food service and hospitality), and clerical or office positions (Isaac, 2001; Rog et al., 1999; Trutko et al., 1998).

As formidable as these barriers may seem, there are consistent reports in the literature that homeless people rise above the barriers and find ways to earn income from employment (Sowell et al., 2004; Theodore, 2000). Indeed, mounting evidence counters the view that homeless people face insurmountable barriers or are simply work shirkers. Given the opportunity, training, and sustained support, even people who have been homeless for long periods or who have experienced frequent episodes of homelessness have succeeded at working. Evidence of homeless individuals desire for jobs and tenacity in working has emerged from case studies and surveys of homeless people (Burt, Aron, & Lee, 1999; Weinberg & Kogel, 1995; Evans, 1998).

Personal characteristics and histories can indicate how well people, including homeless people, will fare in employment. Researchers continue to search for predictive indicators for successful job placement of people with complex problems (McGurk & Mueser, 2006; Hoffman et al., 2003; Macias et al., 2001). Bogard et al. (2001) found that poor single mothers who had experienced an episode of homelessness and had symptoms of depression, but had a work history of full-time employment, left shelters quickly and entered employment after a shelter episode. Identifying such indicators can assist program planners in designing services most likely to meet the needs of a variety of job seekers, so that job seekers and service providers can transcend barriers to employment and achieve vocational objectives.

Income Support and Employment Programs

For each federal agency, we describe mainstream income support and employment programs and their target populations, discussing the barriers to eligibility or participation faced by homeless individuals and families^[4] and, in some cases, evidence that the inability to benefit from the program may contribute to homelessness. We describe relevant efforts that have been made to increase access to the mainstream program for people who are homeless and other hard-to-serve groups. Then we examine ways in which the mainstream program may create disincentives to employment and the agencys efforts to overcome these disincentives through program design or through employment and training initiatives. Finally, we examine the available research to determine what is known about the results of employment and training programs targeted to homeless people or used by homeless people who are part of the broader population group served by the agency. 

Social Security Administration

The Social Security Administration (SSA) focuses on a specific segment of the homeless population: people with serious physical and mental impairments. SSAs mission has always been to provide income support, initially to retired workers and their families and later also to workers with disabilities and their dependents. Most homeless people served by SSA receive Social Security Disability Insurance (SSDI) or Supplemental Security Income (SSI). When SSDI was established in 1956, policymakers expected that individuals who became disabled enough to qualify for benefits would not return to the labor force unless their medical condition significantly improved. Indeed, eligibility for SSDI and SSI has always been contingent on the inability of the applicant to work at any job in the national economy. Disability determination is binary; an applicant is found either disabled or not disabled. As a result, SSA did not initially have a work-support strategy. A strategy modification has gradually evolved during the last 30 years, allowing beneficiaries to earn a limited amount through employment or, on a trial basis, to earn more. Employment-related assistance is offered to all beneficiaries. Earning more than the allowed amount, however, results in termination of benefits, and this has created widespread fear of attempting any work.

SSDI provides monthly benefits to qualifying adults with a significant work history. SSDI benefits vary based on the amount of time worked and the money earned from employment. Using the same disability standard,^[5] SSI is the disability benefit program for low-income people who have never worked, have an insufficient work history to qualify for SSDI, or would receive less in SSDI than the maximum SSI cash grant. Individuals who are disabled are only eligible for SSI if they meet federal income and asset guidelines. Benefits are currently capped at $623 per month for an individual and $934 per month for a couple.^[6]  An individual who has been found disabled and eligible for SSDI or SSI will continue to receive benefits as long as SSAs disability standard is met and the individual does not receive more earnings than program rules allow.

SSA Programs and Homelessness

SSDI and SSI benefits are important both as income supports and as gateways to the receipt of other supports. SSDI receipt provides eligibility for Medicare after a waiting period,^[7] and in most states SSI recipients are immediately eligible for Medicaid. Medicare and Medicaid permit homeless persons to obtain primary and specialty health care as well as prescription drug coverage. These services can help to stabilize people with mental health disorders, and provide proper therapy and other treatment for people with physical disabilities. This care may help people return to work.

A major anti-poverty program targeted to disabled individuals and their families, SSI has been credited with lifting 2.4 million people out of poverty in 2003 (Sweeney & Fremstad, 2005). SSI recipients include many current and formerly homeless people. The National Survey of Homeless Assistance Providers and Clients (NSHAPC) found that 8 percent of homeless people surveyed were receiving SSDI, and 11 percent were receiving SSI.

Receipt of SSI or SSDI can be an important protective factor, preventing people with disabilities from becoming homeless, and may partly explain the finding that homeless people do not have higher rates of disability than other poor people (U.S. Department of Housing and Urban Development, 2007a). Furthermore, both SSDI and SSI are important sources of income for people who move from homelessness into permanent housing. Among formerly homeless people, NSHAPC found that 29 percent received SSI (Burt, Aron, & Lee, 1999).

Barriers to Receiving SSA Income Supports

A recent report found that many homeless persons who are eligible for disability benefits do not receive them. In 2000, the General Accounting Office (GAO, now the Government Accountability Office) estimated that 39 percent of homeless persons reported mental health problems and 46 percent of homeless persons had chronic physical disabilities, far more than the 11 percent receiving SSI. While not all persons with disabilities are eligible for SSDI or SSI, these disparities suggest that at least some eligible individuals who are homeless are not receiving benefits (GAO, 2000). One cause of this gap is a 1997 change in SSI eligibility criteria that prevents receipt of benefits by individuals whose drug or alcohol addiction is a contributing factor material to the determination of disability (Employees Benefits, 2006). A 1999 survey revealed that homeless people losing benefits under the 1997 eligibility changes were more likely than other homeless persons to lose access to both housing and substance abuse treatment services (National Health Care for the Homeless Council and the National Law Center on Homelessness and Poverty, 1999).^[8]

Another reason that many homeless individuals are unable to access disability benefits is that, at the beginning of the application process, they may have difficulty verifying identity or immigration status, because they do not have copies of the necessary documentation. Similarly, homeless persons are often unable to provide documentation of their work history or past medical treatment. Once an application is filed, homeless people are often without a mailing address to receive important communications such as decisions or hearing notices. As a result, their cases can be closed on procedural grounds (Rosen, Hoey, & Steed, 2001).

In recent years, federal agencies have begun programs designed to help homeless persons obtain SSDI and SSI benefits. In 2004, Congress gave SSA funding to operate the Homeless Outreach Projects and Evaluation (HOPE) demonstration grant programs. Each HOPE grantee developed a collaborative relationship with its local SSA office to facilitate the claims of homeless clients. A year later, the Substance Abuse and Mental Health Services Administration (SAMHSA) began its SSI/SSDI Outreach, Access and Recovery (SOAR) technical assistance. The SOAR program provides state agency officials and service provider staff within SOAR states with training and technical assistance designed to improve the work of case managers or other program staff who are handling SSI/SSDI claims on behalf of homeless persons. Both the HOPE and SOAR programs appear to be adopting good practices, but neither initiative has yet been evaluated. However, some preliminary SOAR program data can be discussed.

As of November 2006, eight states had reported outcomes from the SOAR training. Prior to SOAR, only 1015 percent of the homeless SSDI/SSI applicants being assisted by participating agencies were awarded benefits on their initial application. After SOAR, the percentage of successful initial applications increased dramatically. For example, service providers in Montana and the city of Nashville reported 100 percent success rates, while service providers in New York reported a 96 percent success rate. Several other participating states had success rates ranging from 64 percent to 91 percent. In Oklahoma, the success rate lagged at 33 percent, although this was still a notable improvement over prior performance. In addition to higher success rates, the states that documented case processing times all reported significant reductions. In Oregon, cases were approved in an average of 4.5 months, versus 8 months prior to training. In Oklahoma, approval took an average of 80 days, versus 120 days before the training (Policy Research Associates, Inc., 2006).

Rosenheck and colleagues (2000) evaluated outcomes among homeless, mentally ill veterans who applied for SSDI or SSI through a special outreach program. Veterans who were awarded benefits were compared with those who were denied benefits. Beneficiaries were more willing to delay gratification, as reflected in scores on a time preference measure. Three months after the initial decision, beneficiaries had significantly higher total incomes and reported a higher quality of life. They spent more on housing, food, clothing, transportation, and tobacco products, but not on alcohol or illegal drugs. The authors concluded that receipt of disability payments is associated with improved quality of life and is not associated with increased alcohol or drug use.

SSA Employment Supports

Once disabled people begin receiving SSI or SSDI, the probability of their becoming employed is greatly reduced (Rosenheck et al., 2006; Resnick et al., 2003). The work disincentives of SSI and SSDI are well documented (Stapleton & Burkhauser, 2003).^[9]  There also is widespread fear among recipients of both SSI and SSDI that, by becoming employed and earning too much money, they risk losing eligibility for continued benefits, including health insurance. In addition, severely disabled individuals face substantial barriers to employment.

A number of studies have shown, however, that disabled people who are homeless and receive vocational services can achieve promising employment outcomes (Shaheen, Williams, & Dennis, 2003; Zlotnick, Robertson, & Tam, 2002; Picket-Schenk et al., 2002; Cook et al., 2001; Quimby, Drake, & Becker, 2001; Rog et al., 1999; Becker et al., 1999; Trutko et al., 1998). The best research evidence on the effectiveness of employment services for SSI and SSDI recipients comes from Project NetWork, which was evaluated using a rigorous random assignment research design (Kornfeld et al., 1999). SSA implemented NetWork, beginning in 1991, to provide rehabilitation and employment services to SSI and SSDI applicants and recipients. Four program models were tested in eight sites around the country. NetWork significantly increased the earnings of both SSI and SSDI recipients, but these impacts declined in magnitude over time. The least intensive intervention tested  the referral manager model  also appeared to be the least effective in improving earnings and other outcomes. The evaluation did not isolate the impacts of NetWork on sample members who had experienced homelessness.

SSAs Ticket to Work (TTW) program has sought to make a wider range of employment and training services accessible to beneficiaries. TTW gives eligible beneficiaries tickets that may be used to obtain employment-related services from participating providers. The eligible providers are called Employment Networks (ENs). Beneficiaries may choose to work with any approved EN from a range of service providers in public and private sectors. The most common providers are state vocational rehabilitation agencies.^[10]  The TTW Program is designed to provide the specific services needed to meet a beneficiarys employment goals and ultimately move him or her off disability insurance. The maximum allowable payment to an EN for a SSDI beneficiary is approximately $20,000.

In addition, the Ticket to Work legislation improved work incentives in several respects, most notably by allowing people who work to maintain Medicare and/or Medicaid coverage even as their income rises. The legislation also funded counselors to provide reliable information to beneficiaries about pertinent SSA rules and opportunities. Finally, people who lose benefits altogether due to significant work can get those benefits reinstated in an expedited manner if their disability returns and they must reduce or stop their work activity as a result.

A large-scale, rigorous evaluation of TTW is currently underway (Thornton et al., 2004). The evaluations most important interim finding is that the participation rates of SSI and SSDI beneficiaries in the Ticket program are extremely low. The cause of low participation is said to be the TTW payment system. ENs are paid for services they provide in two ways. One, the outcome payment system, provides higher payments, but only when a beneficiary leaves the rolls due to earnings. The other option provides smaller outcome payments, but allows up to four milestone payments for services while the beneficiary is still on the rolls. Recently proposed changes in the EN payment system allow for higher milestone payments.

U.S. Department of Health and Human Services

The U.S. Department of Health and Human Services (HHS) oversees a wide range of programs that target various populations, some including people who are homeless. Temporary Assistance to Needy Families (TANF) is extended to low-income families with dependent children, who automatically become eligible for Medicaid, one of the two largest federal health insurance programs, and several types of social services. Welfare reform efforts during the 1980s and 1990s have left TANF and related programs with a strong emphasis on employment and promoting family self-sufficiency. HHSs Administration on Children and Families also funds transitional living programs for homeless youth.

The programs managed by HHSs Substance Abuse and Mental Health Services Administration (SAMHSA) focus on people with mental illness and people with substance abuse disorders, including people who are homelessness. SAMSHA targets homeless people directly through the Projects for Assistance in Transition from Homelessness (PATH) formula grants and a number of current and past demonstration programs, including Access to Community Care and Effective Services and Supports (ACCESS). Many of SAMHSAs employment efforts are collaborative initiatives with other federal agencies (DOL, HUD, and VA).

Temporary Assistance for Needy Families

Ten years ago, Temporary Assistance for Needy Families (TANF) replaced Aid to Families with Dependent Children (AFDC) as the nations welfare program. Federal funds are distributed to states, each of which operates its own TANF program under broad federal standards. In addition to cash assistance, TANF offers several types of support services to low-income families, including many homeless families. TANF benefits are paid monthly and are time-limited; while there are exceptions, the general rule is that no family may receive federally funded assistance for more than five years.

TANF and Homelessness.  The NSHAPC indicated that 52 percent of homeless families were receiving welfare (it was still AFDC at the time the survey was done) (Burt, Aron, & Lee, 1999). No recent research has updated this percentage, but given the substantial declines in TANF caseloads, it seems likely that fewer homeless families are receiving these benefits. A recent study found that families reaching their TANF time limit or sanctioned for failing to comply with TANF rules are among those particularly vulnerable to housing instability and homelessness (Mills et al., 2006).

The TANF program requires that non-disabled adult recipients work or take part in job training or other educational programs as a condition of receiving benefits. Specific options vary by state; for example, some states have allowed TANF recipients to attend college while receiving benefits, while others do not allow it unless recipients are also working.

Research evidence shows that mandatory work participation programs, which sanction uncooperative TANF recipients by reducing or eliminating their monthly TANF grants, can increase homelessness. One study, based on a random assignment evaluation of a mandatory work program in Connecticut, found that 2.6 percent of the program group reported being homeless and living on the streets and 9.9 percent had to live with family or friends (doubling up). The rates were 1.5 percent and 6.4 percent respectively for the control group, both significantly lower than for the program group. Qualitative data led the evaluators to conclude that these impacts on homelessness were caused by these penalties and by TANF time limits (Bloom, Riccio, & Nandita, 2002).

TANF Employment Supports.  States spend a significant portion of their TANF funds on work supports. Just over one-third of all TANF spending goes to cash benefits, while 18 percent is spent on childcare, 8 percent on work support or employment programs, 2 percent on transportation, and 24 percent on other services (Coven, 2005). When recipients go to work, they may have access to childcare, transportation, and other resources needed to get to work and maintain employment. The benefit calculation process also provides an incentive for TANF recipients to work; benefits are reduced by less than one dollar for each dollar in earned income that a family receives.

However, availability of many work supports is dependent on sufficient funds to provide them. Since 1996, TANF payments to states have been provided in block grants at flat annual amounts. Due to inflation, these funds are not worth as much as they were 10 years ago. As a result, states have been forced to cut TANF spending, and many of the first cuts to be made came in work supports such as childcare and transportation (Coven, 2005).

The TANF disability rules are much less restrictive than the SSI standards. Typically, a doctors letter attesting to an inability to work is sufficient to have an adult exempted from TANF work requirements. In most cases, states have not focused their employment support and job training efforts on households headed by persons with disabilities. Instead, those households have been exempted from time limits and allowed to continue receiving cash assistance. The impact of these policies has not been carefully studied.

Employment services provided by TANF are the subject of a vast evaluation research literature, which has been summarized by Blank (2007). Several studies suggest that employment, training, and support services provided to the most disadvantaged TANF recipients, including those who have been homeless, can produce positive results. One example is the evaluation of Welfare-to-Work Strategies, which rigorously evaluated the impacts of 20 programs on employment and other outcomes, and analyzed impacts for particularly disadvantaged recipients (Michalopoulos & Schwartz, 2000). This and other studies (Danziger & Seefeldt, 2002) have looked at the effectiveness of employment supports for very disadvantaged TANF families, although none have isolated the impacts of employment services on homeless families.

The Job Opportunities for Low-income Individuals (JOLI) program is a discretionary grant program administered by the, HHS Administration for Children and Families Office of Community Services.^[11]  Its purpose is to provide technical and financial assistance to create employment and business opportunities for individuals receiving TANF and for other low-income individuals, including homeless people, with incomes not exceeding 100 percent of the official federal poverty guidelines. Over a three-year period, grantees are expected to help low-income participants achieve self-sufficiency through business expansion, new business ventures, micro-enterprise development, or other non-traditional strategies. Among the 47 JOLI grants to communities across the country from 1998 to 2001, six projects included homeless people among their targeted populations, but no project specifically aimed at or tailored its services for homeless individuals. Largely qualitative evaluations have been conducted for individual JOLI grants, but generalizable research results are not available.

Other Efforts to Help Homeless Families Become Self-Sufficient.  A number of initiatives have sought innovative strategies to stabilize homeless families, most of which are current or former TANF recipients, and help them move towards self-sufficiency. These initiatives typically utilize TANF funds together with resources from HUD, other agencies, and sometimes private sources.

One such initiative is Sound Families, a transitional housing venture in the Seattle area funded by the Gates Foundation, as well as government agencies. Reporting on the 292 families in Sound Families, of which nearly two-thirds were still TANF recipients at intake, Bodonyi et al. (2004) reported positive changes in the primary caregivers resources between entry into transitional housing and exit. Employment increased from 27 percent at intake to 50 percent at exit; TANF reliance decreased from 64 percent at intake to 44 percent at exit. However, the average hourly wage earned at intake and exit did not significantly increase, holding steady at approximately $9.25, which is several dollars below the self-sufficiency wage needed to support a family with only one child in the Puget Sound region.

A follow-up report on the Sound Families initiative suggests a possible relationship between preentry employment and program completion. Data indicate that those unsuccessfully exiting the project were slightly more likely to be receiving TANF and less likely to have income from any level of employment. They were more likely to be receiving Medicaid and less likely to be receiving child support (Bodonyi et al., 2006).

Another initiative, Hearth Connection in Minnesotas Twin Cities, is using a network of service providers to test a managed care approach to ending homelessness for families and single adults. The pilot program tries to coordinate assistance to families and individuals from the Minnesota Family Investment Program (MFIP), which is Minnesotas TANF program, and the Social Security Administration while also providing supportive housing (see additional discussion of supportive housing under U.S. Department of Housing and Urban Development). Four years into the pilot, respondents in family programs reported support from food stamps, TANF, earned income, rent support, and child support, while singles were more likely to report receiving income from Social Security and SSDI (National Center on Family Homelessness, 2004). With many of their survival needs met, few participants worked. Those who did not have jobs noted that their daily activities were limited to walks, watching television, and occasionally talking with friends. Many participants wished they could find and keep a job, even a non-paying job. These feelings were echoed by service providers, who wanted more funds and resources for creating employment opportunities (National Center on Family Homelessness, 2006).

SAMSHA Demonstrations for Homeless People with Mental Illness.  Among SAMHSAs demonstrations, the Access to Community Care and Effective Services and Supports (ACCESS) demonstration program, funded for five years beginning in 1993, produced the most noteworthy results (Randolph et al., 2002; Morrissey et al., 2002; Rosenheck et al., 2002; Pickett-Schenk et al., 2002; Goldman et al., 2002). The ACCESS initiative sought to improve performance and integration of services to reduce homelessness among persons with serious mental illness. More than 35 journal articles used client-level data from the 18 demonstration sites across nine states (Illinois, Connecticut, Missouri, North Carolina, Pennsylvania, Texas, Virginia, and Washington). Of particular interest for this paper are those investigations that addressed themselves to the employment status of the participants (Lam & Rosenheck, 2000; Cook et al., 2001; Pickett-Schenk et al., 2002; Min, Wong, & Rothbard, 2004).

The research design of the ACCESS project involved a comparison of two service delivery models and the results of these in reducing homelessness as well as making improvements in well-being, such as decreased hospitalizations, increased quality of life, and reduced service use. The first intervention was an effort to integrate services at a systems level in conjunction with outreach and case management. The second condition included an outreach and case management strategy absent a systems-level integration component. Researchers used client-level data and a systems-level evaluation to measure the impacts.

In an examination of nearly 5,000 demonstration participants, researchers found that despite receiving relatively few job-related services, modest but significant increases in employment occurred across 12 months of participation. Those employed at 12 months were more likely to have received job training and job placement services, suggesting a relationship between receipt of employment services and positive vocational outcomes (Cook et al., 2001).

A review of the work histories of 7,228 homeless individuals with mental illness in the ACCESS demonstration supports this conclusion. Earlier studies indicated that lengthy periods of homelessness posed a major barrier to employment for the mentally ill population (Lehman et al., 1995; Ratcliff, Shillito, & Poppe, 1996). However, in the ACCESS study, having been homeless for a long time was not a significant predictor of employment status.

Severe psychiatric illness, however, did emerge as a significant predictor of employment status. Clients with a mental health diagnosis but who appeared not to be hindered by this obstacle  those who had not used mental health services, had never been hospitalized, and did not have schizophrenia  were more likely to have worked for pay in the month prior to enrolling in ACCESS. Most interesting, however, is that, holding all other factors equal, clients who received job training assistance prior to enrollment into ACCESS were one and a half times more likely to be employed at the programs conclusion, while people who also received help in finding a job were two and a third times more likely to find employment. These results suggest that homeless people with serious mental illness can use vocational services, and that receipt of these services is significantly associated with an increased likelihood of being employed (Pickett-Schenk et al., 2002).

U.S. Department of Veterans Affairs

The Department of Veterans Affairs (VA) is one of the federal governments most important providers of employment services and income and in-kind assistance directed to the homeless population, especially unaccompanied adults. NSHAPC estimated that about one-quarter of the adult homeless population has served their country in the Armed Services (Burt, Aron, & Lee, 1999). The Department reports that, on any given day, as many as 200,000 veterans (male and female) are living on the streets or in shelters, and perhaps twice as many experience homelessness at some point during the course of a year (U.S. Department of Veterans Affairs 2006). The VA also provides cash assistance and health care coverage to veterans, including those who are or have been homeless, although NSHAPC estimated fairly low percentages of homeless veterans receiving veterans disability payments or veterans pensions (Burt, Aron, & Lee, 1999). Two VA programs support employment and training interventions for homeless veterans: the Capital Grant and Per Diem program and the Compensated Work Therapy program.

The Capital Grant and Per Diem program funds community-based agencies to provide transitional housing or service centers for homeless veterans. Under the Capital Grant Component, the VA may fund up to 65 percent of the project for the construction, acquisition, or renovation of facilities or to purchase vans to provide outreach and services to homeless veterans. The Per Diem Component is available to grantees to help offset operational expenses. Programs may apply for Per Diem Only funding. Per Diem funds can be used for a variety of services, including education and job training.

No specific evaluations have been done to determine whether stays in transitional housing programs funded by the Capital Grant and Per Diem program increase the likelihood that homeless veterans will become employed. However, many Grant and Per Diem residents are receiving job training assistance through the Homeless Veterans Reintegration Program (HVRP), which is operated by the Department of Labor. HVRP has produced positive employment outcomes for program participants (see U.S. Department of Labor section below).

The VA also administers a Compensated Work Therapy (CWT) program for impaired, at-risk, and homeless veterans who have multiple challenges, including psychiatric and substance abuse issues, physical limitations, ex-offender status, and/or family relationship issues. Participants in CWT programs are referred by a primary-care clinical team that has assessed the veteran and determined that he or she could benefit from the program. Vocational rehabilitation specialists then work with the veteran to address barriers that stand in the way of achieving full-time gainful employment in the private sector.

Compensated Work Therapy has two components, Transitional Residence (CWT/TR) and Veterans Industries (CWT/VI). Transitional residences are community-based supervised group homes in which veterans live while working for pay in the Veterans Industries program. Veterans in the CWT/TR program work about 33 hours per week, with approximate earnings of $732 per month, and pay an average of $186 per month toward maintenance and upkeep of the residence. The average length of stay is about 174 days.

Veterans Industries is a vocational rehabilitation program that endeavors to place veterans in competitive jobs and to provide workplace supports as needed. VA contracts with private industry and the public sector for work done by these veterans. CWT/VI staff provide vocational rehabilitation services; employment supports and case management; work site analysis; and consultation with businesses regarding assistive technology, accommodation, and guidance in addressing Americans with Disabilities Act (ADA) regulations compliance. An evaluation of the program found that these services did not produce significantly greater improvement among participants than other clients on any of seven outcome measures, one of which was employment (Rosenheck, Stolar, & Fontana, 2000).^[12]

U.S. Department of Housing and Urban Development

The U.S. Department of Housing and Urban Development (HUD) provides mainstream rental housing assistance to very low-income and low-income families and individuals. Housing assistance comes in three basic forms: tenant-based housing choice vouchers, public housing developments owned and operated by local housing authorities, and rental developments owned and operated by private landlords who have subsidy contracts with HUD under the project-based Section 8 program. HUD also administers McKinney-Vento grant programs targeted specifically to homeless people. HUDs homeless assistance programs include competitive funding for local providers of transitional housing for homeless people and permanent supportive housing for homeless people with disabilities, as well as a formula-based Emergency Shelter Grant program.

Mainstream Housing Assistance

HUD-funded housing assistance is a major source of in-kind income support, freeing up cash income for other needs by enabling assisted households to pay only 30 percent of their income for rent and utilities. A recent experimental design evaluation of the Housing Choice Voucher program found that families using vouchers were much less likely to live in unstable housing situations than families assigned to the control group and that having the availability of a voucher virtually eliminated homelessness among sample members (Mills et al., 2006). Housing assistance programs also are widely used by homeless families and individuals as a means to leaving homelessness for permanent housing.

The main barrier to the use of assisted housing by people at risk of homelessness or trying to leave homelessness is that housing assistance is not an entitlement. Altogether, it serves about 5 million households, while a similar number of renter households with incomes below 50 percent of local median income have severe or worst case housing needs and would likely use housing assistance if they could get it (US Department of Housing and Urban Development, 2007b). Thus, homeless people face stiff competition from other households on waiting lists for assisted housing. The housing authorities that administer the voucher and public housing programs may establish priorities for assisting homeless people, but such priorities are not common (Khadduri & Kaul, 2005.) One of the purposes of the Community of Care planning process required of HUD McKinney-Vento grantees is to bring local housing authorities into the effort to end or reduce homelessness by encouraging them to create such priorities and to find ways to address other barriers to program participation, such as histories of evictions from rental housing.

Employment Supports in Mainstream Housing Assistance Programs.  When homeless people are able to use HUD mainstream programs to leave homelessness, does that help them work? Like other means-tested programs, housing assistance may discourage work. Receiving housing assistance reduces the need for income to buy housing, and tenants must pay 30 percent of most earnings as their contribution toward their subsidized housing, which amounts to a 30 percent tax on top of other taxes residents pay on their earnings (Olsen et al., 2005).

However, the work disincentive effect of public housing and housing vouchers is not as strong as is widely thought. Corcoran and Heflin (2003) studied current and former welfare recipients receiving housing assistance and how they differed from those not receiving assistance on various potential barriers to employment. They found that housing assistance is not associated with the probability of receiving welfare or being sanctioned for noncompliance with the work requirement. Additionally, they found that support for the relationship between housing assistance and work outcomes is weak. Housing assistance has no effect on the probability of being employed, the natural log of weekly earnings, the percentage of months observed working, or the percentage of months observed receiving welfare.

The recently completed Housing Choice Voucher Demonstration randomly assigned more than 9,000 current and former TANF families to program and control groups in six sites. The demonstration, which included many families that were unstably housed or homeless, found that the reduction in labor supply caused by housing vouchers was modest and temporary (Mills et al., 2006).

The Family Self-Sufficiency (FSS) program is HUDs program to help voucher families obtain employment that will lead to economic independence and self-sufficiency.^[13]  Housing authorities work with welfare agencies, schools, businesses, and other local partners to develop a comprehensive program that gives participating FSS family members the skills and experience to enable them to obtain employment that pays a living wage. FSS has not been rigorously evaluated, but both case studies and analyses of FSS participant outcomes have been encouraging (Sard, 2001; Ficke & Piesse, 2004). No one has studied the extent to which formerly homeless families who use vouchers participate in the FSS program.

A demonstration project, Jobs Plus, sought both to reduce public housings role as a barrier to employment (by limiting increases in rent due to increased earnings) and increase its role as a bridge (by offering attractive, place-based services in public housing). The evaluation found statistically significant impacts on public housing residents earnings, whether or not the residents lived in their homes for extended periods. The overall effects were determined primarily by large, sustained impacts in Dayton, Los Angeles, and St. Paul, where implementation of Jobs Plus was strongest (Bloom, Riccio, & Nandita, 2005; Riccio & Orenstein, 2003).

In addition to housing assistance programs without a specific target population, HUD also funds the development of housing dedicated for occupancy by elderly people through the Section 202 program and housing for people with disabilities through the Section 811 program. Section 811 housing can be one of the permanent housing options for people with disabilities. Another residential program, Housing Opportunities for People with AIDS, is among the programs often identified in particular communities as preventing homelessness or helping people leave homelessness.

McKinney-Vento Homeless Assistance Programs

Among HUDs programs explicitly targeted to homeless families and individuals, the McKinney-Vento Supportive Housing Program can be used for transitional housing; permanent housing; and for services, including employment supports, that are part of a particular residential program or that serve a broader homeless population. The McKinney Vento Shelter Plus Care program provides permanent housing for people with disabilities through a rent subsidy structured in the same way as mainstream assisted-housing programs (residents pay 30 percent of their income and the subsidy pays the balance of the housing cost). Grantees must match the Shelter Plus Care rent subsidy with funding for services that comes from other sources. Services usually include case management focused on working and increasing employment income.

Employment is an important objective for transitional housing programs, both as an aspect of self-reliance and because the limited availability of both mainstream assisted housing and supportive housing means that moving to permanent housing often means having enough income to pay for it. Employment is important for permanent supportive housing because of its importance for self-reliance and self-esteem and because residents of supportive housing often want to move on to mainstream housing.

HUD has established performance goals under the Government Performance Reporting Act (GPRA) for its McKinney-Vento programs that give employment gains the same importance as housing stability. Reviewed annually, the programs are expected to increase the percentage of clients working by 11 percent from entry to exit from the program. This is expected to change in 2007 to a performance measure that expects an additional 18 percent of exiting clients to be employed compared with the percentage employed at entry. Without a rigorous impact evaluation, we cannot judge the extent to which increased employment is attributable to the HUD-funded homeless assistance programs.

We also do not have systematic knowledge of the type of employment services and work supports that are provided to homeless people in connection with residential programs for homeless families and individuals or to other homeless people (those in emergency shelters or unsheltered) as part of local continuums of care. Vocational training, job search assistance, and job placement all are used by particular providers and in particular communities, but no one has documented the patterns or assessed the quality and appropriateness of the services provided to particular groups of clients as part of local strategies to address homelessness.

Supportive Housing and Employment-Related Services.  In Philadelphia, researchers tracked 96 leavers of permanent supportive housing and found that less than a quarter engaged in paid employment. The study examined the experience of some 943 residents of permanent supportive housing in the City during the period from 2001 to 2005 (Wong et al., 2006).

Many supportive housing organizations appear to recognize the important role employment can play  as a source of income from which tenants pay rent; as a viable platform from which to enter the workforce; and as an important occupying activity in the face of potential idleness. Employment services are frequently part of the supportive services linked to permanent supportive housing (Burt & Anderson, 2005). Supportive housing projects in New York, Chicago, San Francisco, and other cities combine housing with social services and employment and educational services (Rio et al., 1999). Supportive housing for homeless people, including single adults and heads of family households with complex needs, frequently offers job-retention supports to help residents maintain their attachment to the workforce.

Research evidence on the effectiveness of supportive housing generally does not focus on the separate effect of employment-related services (Martinez & Burt, 2006; Nelson et al., 2005; Rog, 2004; Tsemberis, Gulcur, & Nakae, 2004; Rosenheck et al., 2003; Rog & Randolph, 2002; Culhane, Metraux, & Hadley, 2001). However, Long and Amendolia (2003) evaluated the cost-effectiveness of employment and training supports provided as part of the Next Step: Jobs Demonstration, which introduced intensive new services to existing supportive housing programs operating in New York City, Chicago, and San Francisco. The demonstration was conducted between 1996 and 2000 with funding from the Rockefeller Foundation. Demonstration services included basic and life skills training, GED and ESL classes, vocational training, on-site employment (mostly in program-sponsored businesses), and job development and placement services. The net impacts of the services were estimated using a comparison group research design and regression adjustments for demographic and pre-enrollment experience differences between program and comparison groups. The study concluded that the benefits of the employment supports  resulting primarily from statistically significant increases in earnings and reductions in the use of transfer program  far exceeded their costs to the Foundation and government programs.

U.S. Department of Labor

The U.S. Department of Labor (DOL) concentrates on the employment and training needs of all Americans, including those who are homeless. Many of its mainstream programs operate under the Workforce Investment Act (WIA), which funds state and local workforce development systems.^[14]  These systems place great emphasis on universal access to services and local planning and control  principles that help determine the services that are provided  and not provided  to homeless people. WIA created a national infrastructure in which state workforce boards oversee the local Workforce Investment Boards (WIBs) that control the funding and employment services within their jurisdictions. At the heart of the system are some 3,500 One-Stop Career Centers operated by 600 WIBs.^[15]  States and localities are given broad discretion to design and operate their systems to meet state and local needs.^[16]

One-Stop Employment Centers

WIA-funded programs are intended to provide access to employment-focused assistance to all individuals in need of help, including hard-to-serve people such as homeless families and individuals. A person needing employment assistance should be able to obtain, through WIAs One-Stop Centers, information and services from the various federal agencies that provide help finding jobs and advancing in the labor market. However, the WIA system is not designed to pay special attention or deliver specially tailored services to people who are homeless. Indeed, DOL does not currently count the number of homeless adults or homeless dislocated workers served by WIA and apparently has no plans to do so in the future (counting homeless clients is not mentioned in DOLs Strategic Plan for 20062011). Under WIAs predecessor, the Job Training Partnership Act (JTPA), about 2 percent of the 151,580 individuals served in 1998 by JTPAs adult programs were homeless; and 2.4 percent of JTPA participants in 1994 were homeless (GAO, 2000). A subsequent report estimated that 416,000 WIA participants received training in program year 2003 (GAO, 2005). If the estimated percentage of homeless participants in JTPA applies to WIA, then perhaps 8,00010,000 people who are homeless are served annually under WIA. However, given WIAs demanding performance requirements (which discourage programs from working with hard-to-serve clients), the percentage of WIA participants who are homeless might be less than under JTPA.

Participants seeking assistance through One-Stop Centers are supposed to receive services designed to meet individual needs. However, evaluation research findings offer several reasons to think this may not be happening in the case of homeless individuals. First, while customer choice is promoted through Individual Training Accounts (ITAs), which allow individuals to choose the kind of training they want, resources for these accounts are limited.^[17]  GAOs assessment of ITAs (2000) concluded that (1) the dollar value of the ITAs may not be sufficient to meet the training needs of homeless individuals who require more intensive services, (2) the network of qualified providers may not include enough providers with expertise in meeting the needs of hard-to-serve populations, and (3) homeless people may find the vouchers difficult to use and may not be in a position to choose the training programs most suitable for their needs. A more recent study, which focused on persons seeking employment assistance from One-Stop Centers in Chicago, found that most job seekers who were homeless did not receive the assistance they needed (Chicago Coalition for the Homeless, 2005).

Second, most One-Stop Centers are not able to provide special guidance to homeless individuals about how to use the available resources. Only a few centers have staff with suitable training to work with homeless clients. Selected WIBs in about half the states have received special funding for Disability Program Navigators (DPNs), staff who are specially trained to facilitate access to services and benefits for workers who are disabled.

DOL also funds the Comprehensive Employment Program (CEP) in eight sites (made up of WIBs or groups of WIBs) around the country. The CEPs provide employer-linked services, such as job carving and restructuring, to a small number of people with disabilities. It is unlikely that more than a fraction of the disabled people served by these special programs are homeless (no estimates are available), but a largely qualitative evaluation of these services (Holcomb & Barnow, 2004) offers hope that approaches like these would be successful if operated on a larger scale.

Third, the One-Stop Career Center system, as currently designed, discourages centers from serving homeless people. DOLs Office of Disability Employment Policy conducted an evaluation of pilot employment programs to identify policy and practices concerns that would improve employment outcomes for people with disabilities served by the centers. Project staff told the evaluators that the current indicators of performance for workforce investment activities under WIA are a disincentive to serving customers with disabilities, including people who are homeless. Several project sites noted that state and federal policies, such as WIA performance measures, impede the ability of One-Stop Centers to provide needed services to hard-to-serve clients.^[18]  Thus, the demonstration projects have tended to enroll participants they gauge will be successful in finding employment, helping the projects meet the performance measures. The One-Stop Centers were consequently reluctant to enroll participants with disabilities and other severe problems. Some members of the WIA system also said they were reluctant to participate in these pilot projects, because they feared they would be penalized as their staff members took on time-consuming responsibilities that negatively affected their WIA-defined performance (Elinson et al., 2005). Such observations are consistent with those reported elsewhere (GAO, 2000).

A number of individual WIBs and One-Stop Centers have sought to do more for the homeless population. Some have partnered with homeless assistance agencies or have developed employment and training services for job seekers who are homeless, often using McKinney Supportive Housing Program (SHP) dollars. Others have enrolled homeless job seekers in WIA services or have secured funding to enhance their core services to meet the needs of homeless populations. As part of a national evaluation of WIA implementation, cases studies focused on One-Stop Centers serving homeless job seekers in three areas: Tucson, Arizona, Portland, Oregon, and Portland, Maine. These sites were selected because of their experience and innovative strategies for working with people who are homeless. All were connected to community networks of homeless service providers, which appeared to increase their effectiveness. The three sites had service strategies that emphasized job placement over job training, but also provided supports (including assessment and case management services) intended to produce greater success for homeless individuals. Finally, the sites had local political support that helped promote effective partnership strategies and the linking of mainstream resources (Henderson-Frakes, 2004).

Alternatives to conventional job services have also been used as employment strategies for people who are homeless: day labor programs (now referred to as contingent labor) and social purpose business ventures. In a review of 27 staffing services, including services that provide temporary jobs for homeless people, researchers suggest that these services can play a unique role in addressing the needs of disadvantaged job seekers in ways more conventional job services do not. They conclude that access to these jobs is less difficult for those with barriers to employment than jobs they are able to find at One-Stop Centers. The jobs offered do not require a long-term commitment; the brokering service agency has an investment in the jobs and acts as a protective buffer for those who would find it difficult to interact directly with employers. Community organizations provide temporary-job brokering specifically to disadvantaged workers, often in conjunction with career counseling, transportation assistance, and other supports; in some instances the brokering agency offers health care benefits to workers (Carre et al., 2003).

Other Mainstream DOL Employment Programs

In addition to One-Stop Centers, other Department of Labor programs that sometimes include homeless people within their service populations are the Adult and Dislocated Worker Program and the WIA Youth Program. The Adult and Dislocated Worker Program serves people who have been terminated or laid off, have received a notice of termination or layoff, or are eligible for or have exhausted unemployment insurance. The Youth Program provides services to low-income youth (age 1421) with barriers to employment. Eligible youth are deficient in basic skills or are homeless, runaway, pregnant or parenting, offenders, school dropouts, or foster children.

Another DOL initiative that included homeless people among its service population was the Welfare-to-Work (WtW) Grants program, which was designed to help the most disadvantaged TANF recipients leave the rolls and become employed. The program was the subject of a large-scale evaluation (Mills et al., 2006).^[19]  The sites in the WtW Grants demonstration generally provided services, such as job search assistance, geared to moving welfare recipients rapidly into jobs in the labor market. The WtW enrollees generally had characteristics associated with disadvantages in the labor market  including being an unmarried parent with young children, having little education or work experience, and experiencing work-limiting health problems  although the enrollees in some sites appeared to be less disadvantaged than those in other sites. Two years after entering WtW programs, 4 in 10 enrollees were working  a much higher proportion than at program entry  and nearly two-thirds worked at some point in the second year. Receipt of TANF and poverty rates both declined substantially.

Three of the sites in this demonstration (in Chicago, Ft. Worth, and Nashville) funded organizations that specialized in serving homeless families and persons with mental or physical disabilities. In addition, enrollees in three other sites (Boston, Milwaukee, and Phoenix) exhibited high rates of homelessness shortly after program entry. The evaluation findings indicate that these rates fell significantly by the end of the second year in Boston and Phoenix, but not Milwaukee. A significant reduction in homelessness was also found for Yakima, Washington.

DOL Employment Programs for Homeless People

The Homeless Veterans Reintegration Program (HVRP) awards funds on a competitive basis to eligible applicants such as state workforce boards and local WIBs; public agencies; commercial entities; and nonprofit organizations, including faith-based and community-based organizations.^[20]  Administered by DOLs Veterans Employment and Training Service, it is a modest program with an annual appropriation of about $25 million supporting 87 grantees in 2006. It funds services to assist in reintegrating veterans who are homeless into meaningful employment within the labor force and to stimulate the development of effective service delivery systems that address the complex problems facing homeless veterans. In the Departments last program report on HVRP, 13,725 homeless veterans were served, 61 percent of those seeking assistance entered employment, and 58 percent retained employment for six months. The key program-effectiveness metric used in the report, average wage at placement, was $10.11 (greater than both the average wage in previous years and DOLs goal for HVRP). In program year 2004, the HVRPs average cost per placement ($2,152) was less than the average cost per participant during the same year under other DOL programs for special population groups, the Dislocated Worker program ($3,318) and the Employment Opportunities for Youth and Adults with Disabilities program ($2,882).^[21]

DOL, jointly with HUD, is sponsoring a federal five-year demonstration for Ending Chronic Homelessness through Employment and Housing, in partnership with Local Workforce Investment Boards in Boston, Indianapolis, Portland, Oregon, San Francisco, and Los Angeles.^[22]  Projects are expected to make use of evidenced-based and best practices to address mental health and addiction disorders and provide customized employment  a job placement service in which the service provider negotiates with the employer to customize a job to reflect the special needs of the client. Nearly all participants have been served with grant resources; few were enrolled in WIA-funded services. Housing retention rates for the comparable period are not available. However, at mid-course of this demonstration, 484 chronically homeless, mostly single, adult men had been served. Half were African American, and 63 percent reported a psychiatric or emotional disability at intake. In terms of income supports, 25 percent were receiving SSI, 21 percent were receiving food stamps, and 14 percent were working either full- or part-time at intake. After 33 months of grant awards, projects reported an entered-employment rate of 51 percent (competitive) and a 22 percent rate at which participants were placed in non-competitive employment activity. For participants for whom data were available, those in competitive employment worked an average of 28 hours per week and started at $9.06 per hour (Palan, Elinson, & Frey, 2006).

U.S. Department of Agriculture

The U.S. Department of Agriculture (USDA) administers the Food Stamps program, which enables low-income people, including those who are homeless, to buy food. At the local level, the program is often implemented in the same offices where people go to apply for TANF, Medicaid, and other public assistance benefits. A similar policy perspective is shared across the federal agencies that administer these programs. However, while TANF is targeted to families and Medicaid is directed to both families and to individuals with disabilities or high medical expenses, food stamps are, in principle, available to the entire homeless population.

Food Stamps

Homeless families are typically eligible for food stamps based on income alone, as are non-disabled single adults over 50. However, non-disabled adults between the ages of 18 and 50 face significant time limits in the receipt of food stamps. People in this category may only receive food stamps for 3 months out of a 36-month period, unless they are working at least 20 hours per week or enrolled in a job training program. As in the TANF program, the disability standard is not nearly as stringent as that used by the SSDI and SSI programs. A letter from a doctor, provided to the food stamp office, is typically sufficient proof of disability  meaning that work requirements will not apply.

Food stamps are a significant income support for individuals and families attempting to leave homelessness. Often most of a TANF grant or an SSI benefit goes to pay rent in permanent housing; receipt of food stamps may allow a household to reach a subsistence level. Food stamps also are an important source of income support for people who leave homelessness with a job. A household may earn up to 130 percent of the federal poverty line and still be eligible for food stamps. The food stamps benefit is reduced by no more than 36 cents for every dollar earned.

The NSHAPC indicated that 31 percent of homeless single adults were receiving food stamps, compared to 71 percent of people in homeless families (Burt, Aron, & Lee, 1999). No research has reviewed the reasons for this disparity. However, it can be inferred that program restrictions on the receipt of benefits by non-disabled adults between the ages of 18 and 50 are responsible for much of this disparity.

Homeless persons who are living on the street can find it difficult to interact with local food stamp offices. They typically do not have the documentation necessary to apply for benefits or a mailing address at which to receive important notices. In addition many applicants for food stamps are incorrectly told that they are ineligible for benefits if they are living in a homeless shelter, particularly if the shelter serves meals. However, this is incorrect. Residents of homeless shelters are exempted from the general rule that residents of institutions providing more than 50 percent of daily meals are ineligible for food stamps, and often persons who are homeless will want food stamps to supplement shelter meals for their children or to purchase food while they are away from a shelter at work (Rosen, Hoey, & Steed, 2001; GAO, 2000).

A recent study tested the effect of housing and substance use treatment histories on the receipt of food stamps. The study showed that people who were homeless or unstably housed were less likely to receive food stamps than people in stable housing. Conversely, people receiving current substance use treatment were more likely to receive food stamps (Nwakeze et al., 2003). This study indicates that people who are stably housed are more likely to obtain government benefits.

Many homeless families believe that food stamps are linked to the TANF program, particularly its time limits. As a result, they are reluctant to apply for food stamps, not wanting to have months of food stamp receipt count against their lifetime five-year TANF time limit.

There are a number of key recommendations for improving homeless persons access to food stamps. The first, and perhaps most important, is outreach. States are advised to put caseworkers in hospitals, homeless shelters, food pantries, and soup kitchens. Another recommendation is for states to train workers at homeless shelters, so those workers can ensure that their eligible clients receive food stamps. Finally, agencies must exhibit flexibility in working with homeless persons. A homeless person should be able to pick up a letter at the local office instead of having it mailed, and if a homeless person needs help in obtaining documentation, agencies should fulfill their obligation to provide that assistance. These recommendations can be expected to increase homeless persons access to food stamp benefits (K. Gale Consulting, 2003; Rosen, Hoey, & Steed, 2001; GAO, 2000).

Employment Programs Linked to Food Stamps. For single non-disabled adults who must register for work, USDA has a Food Stamps Employment and Training program (FSET). Funds are awarded in each state based on the number of work registrants in that state. In addition to these funds, states may draw additional FSET resources through the 50:50 matching program in which states use non-federal revenue to match federal dollars. In FY2005 total FSET funds expended were $466,599,435, including federal and state resources.

Delivering services through a variety of local entities, such as TANF offices, community-based service providers, or One-Stop Career Centers, FSET focuses on placing food stamp recipients who are able-bodied adults without dependents in employment and training slots to enhance their ability to gain unsubsidized private sector employment. About 30 percent of food stamp recipients are exempt from mandatory participation because of disability, taking care of a child under six, or being employed 30 or more hours a week, but they can voluntarily access employment and training services without the sanctions that may be applied to mandatory participants.

Several communities  including Seattle,^[23] Houston, and Boston  have used FSET resources to help homeless job seekers who are food stamp recipients. Anecdotally, it seems that participants who participate in employment services voluntarily are more job-ready than participants mandated to participate and referred from the local food stamps office.

Reports about FSET focus on information about how states operate their programs (Botsko et al., 2001). There are no FSET evaluations or reports specifically evaluating the inputs and results of employment and training services for homeless job seekers who receive food stamps, or whether the program helps participants get a job. In preparing its report to Congress, the GAO surveyed 15 states and found the entered employment rate ranged from 15 percent of enrolled participants to 62 percent (GAO, 2003).

Internal Revenue Service

The Internal Revenue Service (IRS) has a very different mission from the other agencies: compliance with federal tax laws. One policy instrument contained in those laws, the Earned Income Tax Credit (EITC), is considered by many to be the countrys most important anti-poverty tool. Essentially, the EITC pays a family with children 40 cents for every dollar of earned income up to $11,000. This is an important source of income, and a powerful work incentive, to heads of families.

Earned Income Tax Credit

The Earned Income Tax Credit (EITC) provides a refundable tax credit to qualifying low-wage workers. Individuals or families who do not work are ineligible for the EITC. Many people with low earned income find that when they claim the credit they do receive money back from the federal government (National Law Center on Homelessness & Poverty, 1998). Single individuals are eligible for a very small EITC  up to $380. The bulk of the credits are received by households with at least one dependent child; households with two dependent children will receive a credit that may reach the maximum of $4,400, depending on the family income level.

The credit is calculated by reviewing a households annual income from work. It rises as work income grows from zero to approximately $11,000. The credit stays the same between $11,000 and $15,000. After $15,000 the credit gradually declines, until it is phased out at approximately $37,000.

There are two ways to apply for the EITC. Most commonly, a household will apply for the EITC as part of its annual federal tax return. If the household is found eligible for the EITC, the credit is applied to their tax liability, and any remaining amount is refunded to the household. It is also possible to apply for the EITC in advance. If a households earnings for the year are predictable and within the range of EITC eligibility, IRS forms can be filled out so that a portion of the EITC is included in a workers paycheck every two weeks. However, this can be tricky, as it requires workers to know their annual income at the beginning of the year. If the IRS determines that a household received advance EITC payments that it was not entitled to, those payments will be due on the next annual federal tax return.

EITC can be a strong work support for low-wage earners, including people who are homeless. EITC payments are not treated as income under federal programs such as public housing, the Housing Choice Voucher Program, Section 8 subsidies, TANF, food stamps, or Medicaid.

EITC refunds have allowed homeless persons to buy cars to provide reliable transportation to work, pay a security deposit and the first months rent on apartments, and pay off student loans to reduce monthly expenses (National Law Center on Homelessness & Poverty, 1998). One housing policy expert has recommended that policymakers consider expanding the EITC and other tax provisions that benefit low-income workers to improve the ability of those workers to afford housing (Stegman et al., 2003).

Approximately 75 to 86 percent of eligible workers claim the EITC each year. NLCHP estimates that this includes between 44,000 and 79,000 homeless workers. At the same time, several significant factors prevent additional eligible homeless workers from claiming the EITC. First, homeless workers are often unaware of the EITC; second, homeless workers are frequently unable to document some or all of their earned income; and third, many homeless workers do not have a mailing address or bank account at which to receive a tax refund check (National Law Center on Homelessness & Poverty, 1998).

Although a significant portion of eligible workers do claim the EITC, NLCHP found that many homeless persons were simply unaware of it. A number of homeless workers do not earn enough to owe taxes, but would still qualify for the EITC. However, their lack of knowledge of the program prevents many from filing tax returns. This problem could be remedied by additional outreach, both directly from the IRS and by ensuring that the IRS educates employers about the credit, so that those employers can encourage their employees to apply (National Law Center on Homelessness & Poverty, 1998).

Homeless workers can often have significant problems in documenting their annual income  a requirement for filing taxes, including filing for the EITC. This can be a problem for several reasons. First, homeless persons often move frequently during the year, as they cycle in and out of homelessness or from one shelter to another. As a result, they may have multiple addresses and multiple jobs. When this occurs, employers may not know where to send a W2 form at the end of the year. Or, a W2 form is received by a homeless person but the paperwork is lost or stolen before that person files his or her taxes in the following year. And, in many cases, homeless persons have day labor jobs or perform other short-term work, employment that may go unreported to the IRS. Consequently, these workers do not file tax returns. Agencies can and should work with consumers to help surmount these barriers. For example, a case manager could help a worker obtain a replacement W2 form, or might be able to persuade a recalcitrant employer to provide a W2 and report earned income to the IRS (National Law Center on Homelessness & Poverty, 1998).

Even though the IRS has ruled that workers do not need a mailing address in order to claim the EITC, lack of an address remains a barrier  particularly if a homeless person is reluctant to file for the EITC because he lacks an address where a refund check can be mailed or a bank account in which to deposit funds. To remedy this problem, homeless service providers can allow consumers to use the agency address as a mailing address. In addition, case managers should explore the availability of low- or no-cost bank accounts where consumers can deposit their EITC refund as well as any additional income (National Law Center on Homelessness & Poverty, 1998).

Conclusions

We draw three general conclusions from our examination of the research evidence. First, both mainstream and targeted programs offer promise. Second, looking across the income support and employment programs of the various federal agencies, it is clear that specific groups within the homeless population receive much more assistance than other groups. Third, the research evidence about the effectiveness of programs in reaching the homeless population with income support and in encouraging employment and self-sufficiency while support is provided is plainly weak. In each case, steps could be taken to learn and accomplish more than we have so far.

Mainstream vs. Targeted Programs

Mainstream programs reach far more people than do programs specifically targeted to segments of the homeless population. They are available everywhere, their rules and benefits are relatively well understood by counselors and advocates, and all have grappled with the need to encourage employment and self-sufficiency. Their main drawback is that they are not tailored to the particular needs of homeless people. Several mainstream programs have demonstrated that this shortcoming can be addressed. In particular, the available evidence suggests that SOAR has reduced the barriers that people who are homeless face in accessing disability assistance. The efforts of selected One-Stop Career Centers appear to have increased involvement of homeless individuals in employment activities, although there is no evaluation evidence to gauge their effectiveness. Funding for such mainstream initiatives geared to the homeless, and for evaluating the effectiveness of these initiatives, would be a wise investment.

The rationale for targeted programs is clear: the programs can be tailored to the needs of people who are homeless, and program resources can be focused on the homeless population as opposed to other low-income groups. Targeted programs can also assemble resources for homeless families and individuals from multiple mainstream programs into an accessible package. Unfortunately, there is relatively little evaluation evidence on the effectiveness of such efforts, particularly for employment outcomes. An exception, the evaluation of the Next Step Jobs employment and training services for supportive housing residents, provides reasons to be optimistic that targeted efforts would produce substantial effects. A more systematic, larger-scale evaluation of such services would be extremely valuable.

Employment and Income Support of Specific Groups

One group within the homeless population, severely disabled heads of families, includes people with mental disabilities (such as schizophrenia and mental retardation) and physical disabilities (such as blindness, musculoskeletal problems, and HIV/AIDS). A family in this category is eligible for income support from SSDI if its head has a sufficient earnings history, or for SSI if he or she does not (assuming that the familys income is not too high to qualify for SSI). Both of these mainstream programs provide monthly cash benefits and Medicare or Medicaid coverage. Regardless of whether they receive disability benefits, families may qualify for food stamps, TANF, assisted housing, veterans benefits, unemployment insurance (UI), and other programs.

In some ways, the picture for family heads with disabilities who are homeless is encouraging. If they receive income support from SSI and/or SSDI, they are eligible for extensive employment, training, and rehabilitative services from the Ticket to Work program. Many have access to additional services through vocational rehabilitation, TANF, or other programs. The pertinent evidence on the effectiveness of these services is limited, but the results of research on hard-to-serve welfare recipients provide reason for optimism.

In other ways, the picture we see for this group is decidedly discouraging. One reason is that many people with severe disabilities do not receive SSI or SSDI and consequently may receive few services and little income support. The other reason is that, among SSI and SSDI beneficiaries, very few people currently use the employment-related services for which they are eligible.

A second group includes severely disabled single individuals. Individuals with documented disabilities can qualify for SSDI or SSI, which in turn gives them access to Medicaid or Medicare, Ticket to Work and other vocational rehabilitation services. Individuals may also qualify for VA aid and other forms of assistance. As with families, however, eligibility does not ensure program utilization. A small numbers of eligible individuals currently utilize these services. Findings from evaluations of employment services for SSDI and SSI beneficiaries suggest that, when services are used, improvement in employment outcomes can be achieved.

The third group, families whose heads are not severely disabled (according to the Social Security Administrations standards for disability assistance^[24]) includes many families whose heads are not living with a spouse. The families are eligible for TANF and food stamps so long as they meet income eligibility requirements. Families on TANF automatically receive Medicaid and other assistance. However, TANF is time-limited, and families lose eligibility after five or fewer years. Indeed, since welfare time limits were first imposed in 1996,^[25] a steady stream of families has lost eligibility for welfare and, at least temporarily, has experienced homelessness.^[26]  Some of these heads have physical or mental impairments limiting their functioning. These families may qualify for exemptions from welfare time limits as well as assistance from other programs, including mainstream assisted housing when it is available.

While the income support provided to this group is not as generous as that available to severely disabled individuals and family heads, the evaluation research evidence on the effectiveness of employment services is more plentiful and promising. There is extensive evidence from research on services provided to welfare families, and these services generally appear to be as effective for hard-to-serve welfare recipients as for others. In addition, there is encouraging qualitative evidence from some initiatives targeted specifically to homeless families.

The last group, homeless individuals who are not severely disabled, is generally eligible for the fewest services and the least income support. They do not qualify for SSI, SSDI, TANF, or the employment services and public health insurance that accompany such assistance. This is also a heterogeneous and changeable group. Homelessness among this group more often consists of a single episode or is brief and often due to an event such as a job loss, medical emergency, house fire, or natural disaster.

Although challenged by a lack of permanent housing, these individuals apparently have employment prospects similar to those of housed individuals competing for employment in the labor market. However, there is very little hard evidence on the effectiveness of employment services in helping people who are homeless improve their labor market prospects. Numerous qualitative studies have highlighted promising program approaches to working with people who are homeless.

Limitations of the Available Research Evidence

A range of income support and employment programs is available to the homeless population. We know that many families and individuals who have experienced homelessness are served by these programs. We also know that some people who need assistance never gain access to the programs. We cannot assess the extent to which programs reach the homeless population. The best available evidence, the National Survey of Homeless Assistance Providers and Clients, is now more than a decade old. Changes in both the programs (program features, eligibility requirements, and outreach efforts) and the homeless population (due to immigration, the introduction of welfare time limits, and many other developments) leave the surveys findings clearly out of date.

There is research evidence, some of it very good, on the effectiveness of mainstream programs in encouraging employment and self-sufficiency in the people they serve. We have highlighted several rigorous evaluations of mainstream programs employment interventions. Unfortunately, none of these evaluations has isolated the impacts of employment-related services on homeless people, who represent a small fraction of mainstream programs' clienteles. Identification of homelessness, including family and individual homelessness, should become a standard part of research that targets a broader population,. For example, SSA and HHS could collect data on homelessness in their new research projects testing employment initiatives for SSI, SSDI, and TANF recipients. Similarly, while it may not be feasible to count the number of homeless job seekers served in the mainstream workforce investment system, a study using sampling techniques could provide insight into how well homeless people are being served by the system.

The evidence on programs specifically targeted to homeless people is limited. To date, none of these programs has evaluated its employment and training interventions rigorously.

The research we have examined suggests that now would be a good time to fill these gaps in our understanding of how to help the homeless population. A survey of the homeless population, measuring their use of income support and employment programs, would be enlightening. Substantial changes in eligibility requirements for and benefits of several mainstream programs, most notably welfare and disability assistance, have been fully implemented. New survey findings could provide a good picture of where we currently stand in meeting the immediate needs of the homeless population.

Finally, many of the studies we have cited have promising findings. The findings are, for the reasons we have given, far from conclusive. However, we now have a stronger basis for taking policy research related to homelessness to the next level  rigorous evaluation research. Rigorous studies of mainstream programs can and should isolate program impacts for the homeless population and assess ways to increase those effects. Rigorous studies of programs targeted specifically to homeless people can realistically be undertaken. Such research will build our policy knowledge base, enabling policymakers to do more for the families and individuals who encounter homelessness.

References

Barber, C. C., Fonagy, P., Fultz, J., Simulinas, M., & Yates, M. (2005). Homeless near a thousand homes: Outcomes of homeless youth in a crisis shelter. American Journal of Orthopsychiatry, 75(3), 347355.

Becker, R. E., Meisler, N., Stormer, G., & Brondino, M. J. (1999). Employment outcomes for clients with severe mental illness in a PACT model replication. Psychiatric Services, 50(1), 104106.

Bernstein, J., & Chapman, J. (2003). Falling through the safety net: Low income single mothers in the jobless recovery. Washington, DC: Economic Policy Institute.

Blank, R. M. (2007). What we know, what we dont know, and what we need to know about welfare reform. Working Paper #07-19. Ann Arbor: University of Michigan, National Poverty Center.

Bloom, H. S., Riccio, J., & Nandita, V. (2005). Promoting work in public housing: The effectiveness of Jobs-Plus. New York: MDRC.

Bodonyi, J., Orlando, L., Yancey, B., & Lamberjack, R. (2006). Evaluation of the Sound Families Initiative, Early exits: Lessons learned from families asked to leave transitional housing programs. Seattle: University of Washington School of Social Work, Northwest Institute for Children and Families.

Bodonyi, J., Orlando, L., Yancey, B., Brennan, K., & Erwin-Stewart, N. (2004). Sound Families evaluation: Preliminary findings. Seattle: University of Washington Schoo l of Social Work, Northwest Institute for Children and Families.

Bogard, C. J., Trillo, A. Schwartz, M., & Gerstel, N. (2001). Future employment among homeless single mothers: The effects of full-time work experience and depressive symptomatology. Women & Health, 32(1-2), 137157.

Botsko, C., Gabor, V., Schreiber, S., & Pachikara, S. (2001). State use of funds to increase work slots for food stamp recipients: Report to Congress (Food Assistance and Nutrition Research Report No. 15). Washington, DC: U.S. Department of Agriculture, Food and Rural Economics Division, Economic Research Service.

Brown, V. L., & Riley, M. A. (2005). Social support, drug use, and employment among low-income women. American Journal of Drug and Alcohol Abuse, 31(2), 20323

Burt, M., Aron, L. Y., & Lee, E. (1999). Homelessness: Programs and the people they serve. Washington, DC: The Urban Institute.

Business Cycle Dating Committee. (2001, November). The business-cycle peak of March 2001. Cambridge, MA: National Bureau of Economic Research.

Carre, F., Herranz Jr., J., Seavey, D., Vickers, C., Aull, A., & Keegan, R. (2003). Employment brokering for disadvantaged job seekers: Improving the temp and the temp-to-perm job experience and enhancing job opportunities (Report of the National Study of Alternative Staffing Services). Boston: University of Massachusetts Boston by arrangement with the President and Fellows of Harvard College.

Chicago Coalition for the Homeless. (2005). Failing to deliver: One-stop employment centers. Chicago, IL: Author.

Cook, J. A., Leff, H. S., Blyler, C. R., Gold, P. B., Goldberg, R. W., Mueser, K. T., et al. (2005). Results of a multisite randomized trial of supported employment interventions for individuals with severe mental illness. Archives of General Psychiatry, 62, 505512.

Cook, J. A., Pickett-Schenk, S. A., Grey, D., Banghart, M., Rosenheck, R. A., & Randolph, F. (2001). Vocational outcomes among formerly homeless persons with severe mental illness in the ACCESS program. American Psychiatric Association, 52, 10751080.

Cooke, C. L. (2004). Joblessness and homelessness as precursors of health problems in formerly incarcerated African American men. Journal of Nursing Scholarship, 36(2), 15560.

Corcoran, M., & Heflin, C. (2003). Barriers to work among recipients of housing assistance. Cityscape Journal of Policy Development and Research, 6(2), 7387.

Coven, M. (2005). An introduction to TANF. Washington, DC: Center on Budget and Policy Priorities.

Culhane, D. P., Metraux, S., & Hadley, T. (2001). The impact of supportive housing for homeless persons with severe mental illness on the utilization of the public health, corrections and emergency shelter systems: The New York-New York initiative. Housing Policy Debate, a Journal of the Fannie Mae Foundation.

Danziger, S. K., & Seefeldt, K. S. (2002). Barriers to employment and the hard to serve: Implications for services, sanctions, and time limits [Special Issue]. Focus, 22(1), 7681.

Drake, R. E., McHugo, G. J., Bebout, R. R., Becker, D. R., Harris, M., Bond, G. R., & Quimby, E. (1999). A randomized clinical trial of supported employment for inner-city patients with severe mental disorders. Archives of General Psychiatry, 56, 627633.

Elinson, L., Frey, W. D., Beemer, M. A., Riley, J. K., & Kruger, H. R. (2005). Evaluation of disability employment policy demonstration programs: Demonstration program progress to date: A synthesis of key findings, issues, and lessons learned: Final report. Rockville, MD: Westat.

Employees Benefits. You are disabled and drug addiction or alcoholism is a contributing factor material to the determination of disability. 20 C.F.R.§ 416.214 (2006).

Evans, W. (1998). Day labor. In K. Hiser & F. Yang (Eds.), We are home now, Our journey from homelessness and other stories. San Francisco: Miller Freeman.

Ficke, R. C., & Piesse, A. (2004). Evaluation of the Family Self-Sufficiency Program: Retrospective analysis, 19962000. Rockville, MD: Westat.

Fragile Families and Child Wellbeing Study. (2003). Variations in maternal and child wellbeing by TANF eligibility and participation (Fragile Families Research Brief, No. 19). Princeton University, Bendeim-Thoman Center for Child Wellbeing.

Goldman, H. H., Morrissey, J. P., Rosenheck, R. A., Cocozza, J., Blasinsky, M., & Randolph, F. (2002). Lessons from the evaluation of the ACCESS program. Psychiatric Services, 53, 967970.

Henderson-Frakes, J. (2004). Serving the homeless through the One-Stop system: A case study. Oakland, CA: Social Policy Research Associates.

Hoffmann, H., Kupper, Z., Zbinden, M., & Hirsbrunner, H-P. (2003). Predicting vocational functioning and outcome in schizophrenia outpatients attending a vocational rehabilitation program. Social Psychiatry and Psychiatric Epidemiology, 38(2) 7682.

Holcomb, P., & Barnow, B. S. (2004). Serving people with disabilities through the Workforce Investment Acts One-Stop Career Centers. Washington, DC: The Urban Institute and Johns Hopkins University.

Holzer, H. J. (1998). Employer demand for welfare recipients and the business cycle. (Discussion Paper No. 1185-99). Madison, WI: University of Wisconsin Institute for Research on Poverty.

Isaac, N. (2001, July). Personal communication with former director of employment services, Lakefront SRO, Chicago, IL.

Katherine Gale Consulting. (2003). Holes in the safety net: Mainstream systems and homelessness. San Mateo, CA: Charles and Helen Schwab Foundation.

Khadduri, J., & Kaul, B. (2005). Permanent housing for homeless families: A review of opportunities and impediments. Paper prepared for U.S. Department of Health and Human Services, Homeless Families Typology Development.

Kornfield, R. J., Wood, M. L., Orr, L. L., & Long, D. A. (1999). Impacts of the Project NetWork Demonstration: Final report. Cambridge, MA: Abt Associates.

Lam, J. A., & Rosenheck, R. A. (2000). Correlates of improvement in quality of life among homeless persons with serious mental illness. Psychiatric Services, 51, 116118.

Lehman, A. F., Kernan, E., DeForge, B. R., & Dixon, L. (1995). Effects of homelessness on the quality of life of persons with severe mental illness. Psychiatric Services, 46(9), 922926.

Long, D. A, & Amendolia, J. (2003). Next Step Jobs  Promoting employment for homeless people: A cost-effectiveness analysis of the Next Step Jobs Initiative. Cambridge, MA: Abt Associates.

Macias, C., DeCarlo, L. T., Wang, O., Frey, J., & Barreira, P. (2001). Work interest as a predictor of competitive employment: Policy implications for psychiatric rehabilitation. Administration and Policy in Mental Health and Mental Health Services Research, 28(4), 279297.

Mancuso, D. C., Lieberman, C. J., Lindler, V. L., & Moses, A. (2003). TANF leavers: Examining the relationship between the receipt of housing assistance and post-TANF well-being. Cityscape Journal of Policy Development and Research, 6(2), 123138.

Martinez, T. E., & Burt, M. (2006). Impact of permanent supportive housing on the use of acute care health services by homeless adults. Psychiatric Services, 57, 992999.

McGuire, J. F., & Rosenheck, R. A. (2004). Criminal history as a prognostic indicator in the treatment of homeless people with severe mental illness. Psychiatric Services, 55, 4248.

McGurk, S. R., & Mueser, K. T. (2006). Cognitive and clinical predictors of work outcomes in clients with schizophrenia receiving supported employment services: 4-year follow-up. Administration and Policy in Mental Health and Mental Health Services Research, 33(4), 598606.

McGurk, S., Mueser, K., Harvey, P., LaPuglia, R., & Marder, J. (2003). Cognitive and symptom predictors of work outcomes for clients with schizophrenia in supported employment. Psychiatric Services, 54(8), 11291135.

Michalopoulos, C., & Schwartz, C. (2000). National evaluation of Welfare-to-Work strategies  Work works for whom: Impacts of 20 Welfare-to-Work programs by subgroup. New York: MDRC.

Miller, K. S., Bunch-Harrison, S., Brumbaugh, B., Kutty, R. S., & FitzGerald, K. (2005). The meaning of computers to a group of men who are homeless. American Journal of Occupational Therapy, 59(2), 191197.

Mills, G., Gubits, D., Orr, L., Long, D., Feins, J., Kaul, B., et al. (2006). Effects of housing choice vouchers on Welfare families. Cambridge , MA : Abt Associates, Inc.

Min, S., Wong, Y. I., & Rothbard, A. B. (2004). Outcomes of shelter use among homeless persons with serious mental illness. Psychiatric Services 55(3), 284289.

Morrissey, J. P., Calloway, M. O., Thakur, N., Cocozza, J., Steadman, H. J., & Dennis, D. (2002). Integration of service systems for homeless persons with serious mental illness through the ACCESS Program. Psychiatric Services, 53(8), 949957.

Munoz, J. P., Reichenbach, D., & Hansen, A. M. (2005). Project Employ: Engineering hope and breaking down barriers to homelessness. Work, 25(3), 24152.

Nakashima, J., McGuire, J., Berman, S., & Daniels, W. (2004). Developing programs for homeless veterans: Understanding driving forces in implementation, Social Work Health Care, 40(2), 112.

National Center on Family Homelessness. (2004). The population being served by the Minnesota Supportive Housing and Managed Care Pilot: A quantitative portrait. Newton Center, MA: Author.

National Center on Family Homelessness (2006). The Minnesota Supportive Housing and Managed Care Pilot: Qualitative evaluation: Year four. Newton Center, MA: Author.

National Health Care for the Homeless Council and the National Law Center on Homelessness & Poverty. (1999). The effects of SSI & SSD benefits termination as seen in HCH projects. Nashville, TN: National Health Care for the Homeless Council, Inc., HCH Clinicians' Network.

National Law Center on Homelessness & Poverty. (1998). Due credit: Increasing homeless workers earnings through the Earned Income Tax Credit. Washington, DC: Author.

Nelson, G., Clarke, J., Febbraro, A., & Hatzipantelis, M. (2005). A narrative approach to the evaluation of supportive housing: Stories of homeless people who have experienced serious mental illness. Psychiatric Rehabilitation Journal, 29(2), 98104.

Nwakeze, P. C., Magura, S., Rosenblum, A., & Joseph, H. (2003). Homelessness, substance misuse, and access to public entitlements in a soup kitchen population. Substance Use and Misuse, 38(3-6), 645668.

Olsen, E. G., Tyler, C. A., King, J. W., & Carrillo, P. E. (2005). The effects of different types of housing assistance on earnings and employment. Cityscape: A Journal of Policy Development and Research, 8(2), 163187.

Orr, L., Feins, J. D., Jacob, R., Beecroft, E., Sanbonmatsu, L., Katz, L. F., et al. (2003). Moving to Opportunity for Fair Housing Demonstration Program: Interim impacts evaluation. Washington, DC: U.S. Department of Housing and Urban Development, Office of Policy Development and Research.

Palan, M. E., Elinson, L., & Frey, W. D. (2006). Evaluation of disability employment policy demonstration programs: Task 5: Follow-up site visit report  Fiscal year 2003 Demonstration program: Ending chronic homelessness through employment and housing. Rockville, MD: Westat.

Pickett-Schenk, S. A., Cook, J. A., Grey, D., Banghart, M., Rosenheck R. A., & Randolph, F. (2002). Employment histories of homeless persons with mental illness. Community Mental Health Journal, 38(3), 199211.

Policy Research Associates, Inc. (2006). Preliminary outcomes from the SOAR (SSI/SSDI Outreach, Access, and Recovery) TA initiative. Retrieved from http://www.prainc.com/SOAR/about/SOARPreliminaryOutcomes.pdf

Quimby, E., Drake, R. E., & Becker, D. R. (2001). Ethnographic findings from the Washington, D.C. vocational services study. Psychiatric Rehabilitation Journal, 24(4), 368374.

Randolph, F., Blasinsky, M., Morrissey, J. P., Rosenheck, R. A., Cocozza, J., & Goldman, H. H. (2002). Overview of the ACCESS program. Psychiatric Services, 53(8), 945948

Ratcliff, K. A., Shillito, L. S., & Poppe, B. J. (1996). The employers role in the job success of people who are homeless. Psychiatric Rehabilitation Journal, 19(3), 8790.

Resnick, S. G., Neale, M. S., & Rosenheck, R. (2003). Impact of public support payments, intensive psychiatric community care and program fidelity on employment outcomes for people with severe mental illness. Journal of Nervous & Mental Disease, 191(3), 139144.

Riccio, J. A., & Orenstein, A. (2003). Are welfare recipients in public housing really harder to employ? In B. Sard (Ed.), A place to live, a means to work: How housing assistance can strengthen welfare policy. Washington, DC: Fannie Mae Foundation.

Rio, J., Russell, D., Dudasik, S., & Gravino, L. (1999). Supportive-housing based employment services. American Rehabilitation, Spring-Summer.

Rog, D. (2004). The evidence on supported housing. Psychiatric Rehabilitation Journal, 27(4), 334344.

Rog, D. J., Holupka, C. S., Hopper, K., Lester, R., Roy, K. & Davidson, C. (1999). Next step: Jobs evaluation/documentation report. Washington, DC: Vanderbuilt Institute for Public Policy Studies.

Rog, D., & Randolph, F. (2002). A multisite evaluation of supported housing: Lessons learned from cross-site collaboration. New Directions for Evaluation, 94, 6172.

Rosen, J., Hoey, R., & Steed, T. (2001). Food stamps and SSI benefits: Removing access barriers for homeless people. Journal of Poverty Law and Policy (March-April): 679696.

Rosenheck, R. A., Kasprow, W., Frisman, L., & Liu-Mares, W. (2003). Cost-effectiveness of supported housing for homeless persons with mental illness. Archives of General Psychiatry, 60, 94051.

Rosenheck, R. A., Lam, J., Morrissey, J. P., Calloway, M. O., Stolar, M., & Randolph, F. (2002). Service systems integration and outcomes for mentally ill homeless persons in the ACCESS Program. Psychiatric Services, 53, 958966.

Rosenheck, R., Leslie, D., Keefe, R., McEnvoy, J., Swartz, M., Perkins, D., et al. (2006). Barriers to employment for people with schizophrenia. American Journal of Psychiatry, 163, 411417.

Rosenheck, R., Stolar, M., & Fontana, A. (2000). Outcomes monitoring and the testing of new psychiatric treatments: Work therapy in the treatment of chronic post-traumatic stress disorder. Health Services Research, 35(1 Pt 1), 133151.

Sard, B. (2001). The Family Self-Sufficiency Program: HUDs best kept secret for promoting employment and asset growth. Washington, DC: Center on Budget and Policy Priorities.

Sard, B., & Harrison, T. (2001). The increasing use of TANF and state matching funds to provide housing assistance to families moving from welfare to work  2001 supplement. Washington, DC: Center on Budget and Policy Priorities.

Sard, B., & Lubell, J. (2000). The increasing use of TANF and state matching funds to provide housing assistance to families moving from welfare to work. Washington, DC: Center on Budget and Policy Priorities.

Sowell, R. L., Bairan A., Akers T. A., & Holtz C. (2004). Social service needs and case management implications for individuals accessing a faith-based suburban homeless shelter. Lippincotts Case Management, 9(2), 7286.

Shaheen, G., Williams, F., & Dennis D. (Eds.). (2003). Work as a priority: A resource for employing people who have a serious mental illness and who are homeless (DHHS Publication No. SMA 03-3834). Rockville, MD: SAMHSA, Center for Mental Health Services.

Stapleton, D. C., & Burkhauser, R. V. (Eds.). (2003). The decline in employment of people with disabilities: A policy puzzle. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

Stegman, M., Davis, W., Quercia, R. (2003). Tax policy as housing policy: The EITC's potential to make housing more affordable for working families. Washington, DC: Brookings Institution, Center on Urban and Metropolitan Policy.

Sweeney, E. P., & Fremstad, S. (2005). Supplemental Security Income: Supporting people with disabilities and the elderly poor. Washington, DC: Center on Budget and Policy Priorities.

Taylor, L. C. (2001). Work attitudes, employment barriers, and mental health symptoms in a sample of rural welfare recipients. American Journal of Community Psychology, 29(3), 443463.

Theodore, N. (2000). Homeless who cant make enough to get ahead. Chicago: Day Labor Project, Chicago Coalition for the Homeless.

Thornton, C., Livermore, G., Stapleton, J., Kregel, T., Silva, B., ODay, T. et al. (2004). Evaluation of the Ticket to Work Program: Initial evaluation  report. Washington, DC: Mathematica Policy Research.

Trutko, J. W., Barnow, B. S., Beck, S. K., Min, S., & Isbell, K. (1998) Employment and training for America's homeless: Final report of the job training for the Homeless Demonstration Program. Washington, DC: U.S. Department of Labor.

Tsemberis, S., Gulcur, L., & Nakae, M. (2004). Housing first, consumer choice, and harm reduction for homeless individuals with a dual diagnosis. American Journal of Public Health, 94(4), 651656.

U.S. Department of Housing and Urban Development (2007a). The annual housing assessment report to Congress. Washington, DC: Author.

U.S. Department of Housing and Urban Development (2007b). Affordable housing needs 2005: Report to Congress. Washington, DC: Author.

U.S. Department of Veterans Affairs (2006). Community Homelessness Assessment, Local Education and Networking Group (CHALENG) report. Washington, DC: Author.

U.S. General Accounting Office. (2000). Homelessness: Barriers to using mainstream programs (GAO/RCED-00-184). Washington, DC: U.S. Author.

U. S. General Accounting Office. (2003, March). Food Stamp Employment and Training Program: Better data needed to understand who is served and what the program achieves (GAO-03-388). Washington, DC: U.S. Author.

U. S. Government Accountability Office. (2005, June). Workforce Investment Act: Substantial funds are used for training, but little is known nationally about training outcomes (GAO-05-650). Washington, DC: Author.

Weinberg, D., & Koegel, P. (1995). Impediments to recovery in treatment programs for dually diagnosed homeless adults: An ethnographic analysis. Contemporary Drug Problems, Summer, 193236.

Wong, Y-L. I., Hadley, T. R., Culhane, D. P., Poulin, S. R., Morris, R. D., Cirksey, B. A., & Brown, J. L. (2006). Predicting staying in or leaving permanent supportive housing. Washington, D.C.: U.S. Department of Housing and Urban Development, Office of Policy Development and Research.

Zlotnick, C., Robertson, M. J., & Tam, T. (2002). Substance use and labor force participation among homeless adults. American Journal of Drug and Alcohol Abuse, 28(1) 3753.

Zuvekas, S. H., & Hill, S. C. (2000). Income and employment among homeless people: The role of mental health, health and substance abuse. Journal of Mental Health Policy and Economics, 3(3), 153163.

Endnotes

[1]     The review covers several initiatives sponsored by private foundations. These initiatives are considered in the context of the federal agency with primary responsibility for the population segment served by a particular project. For example, the Next Step Jobs initiative, funded by the Rockefeller Foundation to serve supportive housing residents, is discussed in the U.S. Department of Housing and Urban Development section.

[2]     The Social Security Administration also serves long-term workers who have reached retirement age, a population unlikely to experience homelessness.

[3]     The U.S. was in recession from March through November 2001 (Business Cycle Dating Committee, 2001). For a discussion of the evidence of harm to disadvantaged populations, such as welfare recipients, see Holzer (1998).

[4]     GAO (2000) provides a good summary of eligibility barriers to mainstream programs experienced by homeless people.

[5]     SSDI and SSI have similar disability determination criteria  adults will be found disabled if they are unable to perform substantial gainful activity (SGA) for at least 12 months due to one or more medically determinable disabilities. To be found eligible for benefits, an applicant under either program must be unable to perform SGA in any former jobs as well as any job that exists in the national economy. This is a high standard, considerably more stringent than the standard used to determine eligibility for HUDs housing and homeless assistance programs, so many homeless individuals who qualify for HUD assistance based on a disability will not be eligible for SSDI or SSI. Inability to perform SGA is often equated with inability to work at all, which is inaccurate. SGA is actually defined as a certain threshold of work earnings, below which an individual is disabled under SSA standards and above which an individual  although he or she might have a disability  is not considered disabled under SSA rules. Whether a person is performing SGA is dependent on the amount of money earned, the number of hours worked, and the type of work being performed. This means that it is possible for people to work and still receive SSDI or SSI.

[6]     2007 Social Security changes are available at http://www.ssa.gov/pressoffice/factsheets/colafacts2007.htm. While some income and assets are excluded from being factored into basic eligibility determinations and determinations of monthly benefits, receipt of too much non-excluded income and assets can result in temporary ineligibility for monthly benefits or even removal from the programs rolls.

[7]     SSDI beneficiaries do not begin receiving cash payments until 5 months after qualifying for benefits; Medicare coverage begins 24 months after benefits start.

[8]     Seventy-six percent of survey respondents who were paying for their own housing lost that housing when their disability benefits were cut off; and 29 percent of respondents who were in a drug treatment program when their benefits were cut off were required to leave that program. In Cook County, Illinois, 74 percent of people surveyed who had lost their benefits also lost Medicare or Medicaid coverage.

[9]     The incentive to work is also low at the time people apply for SSI and SSDI. Work at more than a minimal level calls the severity of the disability into question, potentially resulting in a denial of benefits. If too much income is earned, SSI eligibility criteria may not be met. As a result, applicants for disability benefits often avoid work while their applications are pending.

[10]     State vocational rehabilitation agencies also receive substantial funding from the Rehabilitation Services Administration (RSA), U.S. Department of Education. The funding provided by RSA is authorized under the WIA legislation discussed in the section on U.S. Department of Labor programs. Most notably, the state vocational rehabilitation (VR) agency system has been the main provider of employment-related services to disabled individuals and, until 1996, the only option disability insurance recipients had for publicly financed rehabilitation services (the VR agencies are also by far the most widely used TTW providers as well). VR agencies typically assess disabled individuals, develop rehabilitation plans, and then purchase needed services for their clients or leverage resources at other agencies or service providers. VR staff also counsel clients about their potential eligibility for disability insurance and other program benefits.

[11]     JOLI is authorized under Section 505 of the Family Support Act of 1988, Public Law 100-485, as amended by Section 112 of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Public Law, 104-193.

[12]     In their analysis, Rosenheck, Stolar, and Fontana evaluated the interaction of participation in CWT and changes in outcomes over time using propensity scoring and hierarchical linear modeling. In addition to employment, the outcomes addressed by the analysis include two measures of clinical improvement in posttraumatic stress disorder (PTSD), violence, an alcohol problem index, a drug problem index, and a medical problem index. All sample members had PTSD; other impairments and characteristics varied within this sample.

[13]     FSS was established in 1990 by section 554 of the National Affordable Housing Act. There is a smaller FSS program for public housing residents.

[14]     The Workforce Investment Act (WIA) of 1998, Public Law 105-220, replaced the Job Training Partnership Act (JTPA) that had been in place from 1982 to 2000. Title I of the legislation authorizes the Workforce Investment System and the dislocated worker programs described in this section; Title II reauthorizes adult education and family literacy programs; Title III amends the Wagner-Peyser Act mentioned in this section; Title IV reauthorizes Rehabilitation Act programs (operated by the Rehabilitation Services Administration in the U.S. Department of Education), such as rehabilitation services, projects with industry, and independent living centers; and Title V contains general provisions.

[15]     Workforce investment areas are defined by the state boards, and are generally different from the regions used by other mainstream and targeted programs. For example, the geography of HUDs continuums of care frequently overlap, but rarely share the same boundaries.

[16]     State and local Workforce Investment Boards develop strategic plans and set priorities to meet workforce needs. The majority of board members on both state and local boards are business people. Most WIA funding flows to the states and local areas by formula. The mainstream WIA program includes three funding streams: youth, adult, and dislocated workers.

[17]     DOL, as part of its 2007 budget request, proposed creating career advancement accounts, providing workers more ownership of their education and training. The proposal, based on Individual Training Accounts and experiences with community colleges, is intended to help meet the Administrations goal of achieving higher enrollment rates in training.

[18]     WIA services are divided into three tiers: core, intensive, and training. Core services are available to all job seekers, including access to job listings, information about careers and the local labor market, and limited staff assistance with job search activities. Intensive services are only available after core service efforts are exhausted, and include life skills workshops, case management, and comprehensive assessments leading to the development of an individual employment plan. Training services, such as employer-linked programs and classroom-based skills training leading to a specific occupation, can only be accessed by individuals who have failed to obtain or maintain employment through core and intensive services. Operators of these programs are expected to meet certain performance outcomes established by the state and negotiated with the DOL. Failure to meet performance measures can result in decreased funding to the WIB.

[19]     The evaluation was originally intended to be a random assignment experiment, but, because of low local program enrollment, this research design became unfeasible.

[20]     HVRP was initially authorized under Section 738 of the Stewart B. McKinney Homeless Assistance Act in July 1987. It is currently authorized under Title 38 U.S.C. Section 2021, as added by Section 5 of Public Law 107-95, the Homeless Veterans Comprehensive Assistance Act of 2001.

[21]     U.S. DOL, PART Review for FY 2008.

[22]     For more information about this initiative go to http://www.dol.gov/dol/audience/aud-homeless.htm or www.csh.org/cheta

[23]     See Seattle Jobs Initiative report 8/18/06, Food Stamp Employment and Training  Lessons Learned from Community Partners Engaged in King County Third Party Match Pilot Providers.

[24]     SSA evaluates disabilities on the basis of medically determinable impairments, the extent to which these impairments limit the individuals ability to work, and whether the limitations are expected to last at least a year. The severity of disabilities in each physical and mental diagnostic category is assessed in terms of two types of criteria: medical findings and impairment-related functional limitations.

[25]     The federal Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 enacted this and other changes in welfare. The time limits began to affect families in subsequent years.

[26]     Based on a large survey of TANF-eligible families, 14 percent of involuntary TANF leavers and 8 percent of voluntary leavers report being evicted from their residences. Such families may live with friends and relatives, find other housing, or experience homelessness (living in emergency shelters or on the streets). See Fragile Families and Child Wellbeing Study (2003).