Homelessness in the United States is a highly complex and dynamic condition that has evolved over time. The demographic characteristics of persons experiencing homelessness have changed due to, among other things, fluctuations in the strength and nature of the economy, broad population shifts, and changes in societal attitudes toward poor, excluded, and disenfranchised persons. While our understanding about the causes of homeless and what interventions are most effective has improved, there is still much more we need to learn about this complex issue.
This chapter briefly describes what we know about who experiences homelessness, how homelessness can impact health and other outcomes, and current housing interventions and the populations these are intended to serve, including permanent supportive housing. For a brief history of homelessness in the United States, please see Appendix B.
CURRENT STATE OF HOMELESSNESS IN THE UNITED STATES: DATA AND TRENDS
While the numbers have generally been decreasing since 2010, in 2017, more than 550,000 people in the United States were staying in shelters or in places not intended for human habitation on a single night (HUD, 2017a,b). Many more people experience homelessness over longer periods, such as 1 year or more. In 2016, 1.42 million people at some point stayed in a homeless shelter or a transitional housing program (HUD, 2017a). How homelessness is defined, as described below, impacts how these data are collected, and what we know about who is experiencing homelessness, and informs what services are needed.
The definition of “homelessness” has changed over time, and even today, relevant federal agencies define the term differently. Differences in the characterization of homelessness allow agencies to tailor definitions to represent the needs of their unique subpopulations (e.g., homeless unmarried adults, homeless children, or homeless families) and the goals of the agency’s programs and policies. However, having varying definitions can make it more challenging for people to identify and access the appropriate services (Watson, 1984; SAMHSA, 2017). For example, children experiencing homelessness are eligible for services through their local educational agency with funding from the Department of Education,
which uses a definition of homelessness that is broader than that used by the Department of Housing and Urban Development (HUD). In addition, different definitions create challenges in counting individuals experiencing homelessness, tracking the use of homelessness services, and documenting unmet needs (HUD, 2008; Burt et al., 2010).
While it is tempting to make recommendations that a single definition be developed for use across federal agencies and other relevant organizations, this notion was recently abandoned by the U.S. Interagency Council on Homelessness (USICH). In 2010, USICH convened a meeting of experts and stakeholders to discuss the feasibility of adopting standard definitions and a standardized vocabulary as mandated by the Homeless Emergency Assistance and Rapid Transition to Housing Act of 2009. Substantial concerns were voiced that creating a single definition would be too resource intensive for state and local governments to handle and could lead to a loss of resources for local agencies serving unique subpopulations that might not meet the criteria of a standardized definition. Thus, instead of creating a standardized definition, the recommendation was to create a common vocabulary and common data standards that would allow agencies to distinguish the needs of various subpopulations.
A common vocabulary would ensure that a standard terminology would be used in how local agencies define different manifestations of homelessness but still allow for these different manifestations to be defined as homelessness based on a preestablished set of eligibility criteria. This would also increase the ability of local agencies to capture the diversity within their homeless population. A common data standard would also help to ensure that a standard set of information is collected by reporting entities and would facilitate interorganizational data pooling and linkages to characterize the state of the homelessness by pooling data across agencies.
Some progress has been made on common data standards, with one example being efforts to integrate data sources between the Homelessness Management Information System (HMIS)1 and the Runaway and Homeless Youth Management Information System (USICH, 2015b). In addition, the European Typology of Homelessness and Housing Exclusion (ETHOS) has been developed as “a means of improving understanding and measurement of homelessness in Europe”
1 HUD describes the purpose of HMIS “to produce an unduplicated count of homeless persons, understand patterns of service use, and measure the effectiveness of homeless programs. Data on homeless persons [are] collected and maintained at the local level.” HMIS provides sample policies and procedures, training modules, templates and tools, and manuals to support a variety of homelessness services, including the Continuums of Care (CoC) program, HUD-Veterans Affairs Supportive Housing (HUD-VASH) program, and Veterans Homelessness Prevention Demonstration (VHPD) program. For further information see HUD (U.S. Department of Housing and Urban Development). 2018. HMIS Requirements. Available at: https://www.hudexchange.info/programs/hmis/hmis-requirements. Accessed on April 21, 2018; HUD. 2018. HMIS Guides and Tools. Avail-able at https://www.hudexchange.info/programs/hmis/hmis-guides/#coc-resources. Accessed May 17, 2018.
(FEANTSA, 2018). However, there remain many more examples of a lack of data linkages that need to be resolved at the federal, state, and local levels. While it is difficult to precisely quantify the size of the homeless population, HUD has developed several methods for collecting these data, including HMIS and a single point-in-time (PIT) counting system. The best estimate of counting the number of individuals experiencing homelessness is described in Appendix C.
Subpopulations of Individuals Experiencing Homelessness
An assessment of recent data indicates that overall, more men than women experience homelessness. African Americans are significantly overrepresented among persons experiencing homelessness, accounting for 41 percent of the homeless population while constituting only 13 percent of the U.S. population (HUD, 2017b). Nearly 22 percent of the individuals in the PIT count2 were Hispanic/Latino. The numbers of individuals experiencing homelessness among other racial/ethnic minorities is much lower (1.2 percent Asian, 3 percent Native American, 1.5 percent Pacific Islander, and 6.5 percent mixed race). A 2018 study by the Center for Social Innovation’s Supporting Partnerships for Anti-Racist Communities of five communities found significant racial disparities in rates of homelessness. In fact, the study found that “Black residents accounted for nearly 65 percent of people experiencing homelessness in the five communities, even though they accounted for only 18 percent of the communities’ overall population. Nationwide, black people account for 12 percent of the population, but 43 percent of the homeless population” (National Low Income Housing Coalition, 2018).
Of particular interest to the committee is the number of persons who are defined as chronically homeless: that is, individuals or families (which include at least one adult and one child) with disabilities who have either been continuously homeless for 1 year or more or who have experienced at least four episodes of homelessness in the past 3 years (HUD, 2016c). People experiencing chronic homelessness are one of the primary populations that permanent supportive housing (PSH) programs are designed to serve.
In 2017, data from the PIT count indicated that almost three-quarters of the individuals experiencing homelessness on a single night were not chronically homeless (HUD, 2017b) and thus are not the primary focus of this report. In 2017, those who were experiencing chronic homelessness as measured on a single night included 86,962 individuals, nearly 7 in 10 of whom were unsheltered. Half of all people experiencing homelessness on a single night who are living in unsheltered locations live in one of five states having more temperate climates—California, Nevada, Oregon, Hawaii, and Mississippi (HUD, 2017b).
Veterans represent another subpopulation of particular interest. In 2017, 40,056 veterans were experiencing homelessness, accounting for 9 percent of the
2 The PIT count is a count of sheltered and unsheltered persons experiencing homelessness on a single night in January.
Unaccompanied homeless children and youth on a single night totaled 40,799 in 2017 (HUD, 2017b). They are youth under age 25, with the majority between ages 18 and 24. The number of youth experiencing homelessness is particularly difficult to determine with PIT counts. HUD has targeted this group for more focused efforts to produce better PIT numbers. (For more information about a program serving youth experiencing homelessness in San Jose, see Appendix D.) Auerswald et al. (2016) noted that African American youth can be particularly difficult to find, as they are less likely to access services for youth experiencing homelessness.
One group of youth who are at particularly high risk for homelessness are lesbian, gay, bisexual, transgender, and questioning (LGBTQ) youth. Durso and Gates (2012), in surveying service providers for youth experiencing homelessness, estimated that 40 percent of the youth experiencing homelessness that they worked with were LGBTQ. More information about families and youth can be found in Chapter 6.
Homeless families with children numbered 184,661 people in the single-night January 2017 PIT count, accounting for a third of the total population of
people experiencing homelessness (HUD, 2017b). Most families with children experiencing homelessness were sheltered (more than 90 percent).
A recent subpopulation that has grown to be of concern is homeless older adults. The number of persons older than age 65—the aging Baby Boomers—increases daily (Ortman etal., 2014), and some portion of these older persons will experience homelessness. Hahn et al. (2006) examined 14-year trends in the population of individuals experiencing homelessness in San Francisco (n = 3,534) and concluded that the homeless population is aging by about two-thirds of a year every calendar year, consistent with trends in several other cities. Ng et al. (2013) described the added effects of being both elderly and homeless, noting that “with homelessness, the unsafe and unsanitary living conditions aggravate elderly people’s acute and chronic health conditions” (p. 1). Based on the examination of adults age 50 and older in Oakland, California, in 2013–2014 (n = 350), Lee et al. (2016) found that pre-homeless social support appears to protect this group against street homelessness after losing rental housing. A unique aging trend among the homeless population is further discussed in Chapter 5.
Although all individuals experiencing homelessness face health risks, women have a unique array of medical needs, including a range of reproductive health issues. For example, women experiencing homelessness have higher rates of unintended pregnancies when compared to housed women (Crawford et al., 2011; American College of Obstetricians and Gynecologists, 2013). The experience of pregnancy during a period of homelessness is not difficult only for women. Infants born to mothers experiencing homelessness, when compared to infants born to housed women, are more likely to be low or very low birthweight (Merrill et al., 2011; Richards et al., 2011).
HEALTH OF INDIVIDUALS EXPERIENCING HOMELESSNESS
The experience of homelessness can lead to a variety of negative health outcomes. The Institute of Medicine report on homelessness and health (IOM, 1988) described three types of interactions between homelessness and health. There are health problems that precede homelessness and are likely causal factors for homelessness; health outcomes that occur in response to experiencing homelessness; and health problems including chronic illnesses whose treatment is complicated by the experience of being homeless. Each of these is considered in turn. While numerous studies have documented health problems associated with the experience of either spending time in a homeless shelter, or being homeless and living on the street, the committee acknowledges that there may be additional methodological challenges in assessing the health outcomes in this population which are not described here.
Individuals experiencing homelessness also face overwhelming barriers and obstacles to receiving high-quality, continuous, and coordinated health care. Hospitals, clinics, and reimbursement systems are not designed to cope with the special needs of individuals who spend much of their time on the streets and are exposed to extremes of weather, violence, and a lack of safe, secure, stable housing.
Individuals living in shelters and on the streets have a high burden of medical and psychiatric illnesses, often complicated by chronic substance use disorders. They utilize hospital emergency departments for much of their health care and, in general, require more frequent acute care hospitalizations. These frequent hospitalizations are characterized by longer stays while hospitalized (Kushel et al., 2002; Ku et al., 2010; Cheung et al., 2015; Lin et al., 2015). In addition, there are high rates of trauma/victimization, numerous studies documenting evidence of the accumulation of adverse childhood experiences and toxic stress that contribute to serious chronic medical conditions and poor health, including changes in metabolism, immune systems, and executive functioning and cognitive impairment (Cutuli et al., 2015; Lee et al., 2017). These issues, including a new paradigm to better understand the impact of permanent supportive housing on health for those with chronic conditions is further discussed in Chapter 3.
Health Problems Preceding Homelessness
As described below, research indicates that substance use and mental health are both a cause and a consequence of homelessness (e.g., homelessness is related to worsening severity of mental illness and higher-risk behaviors in the case of substance use) (Johnson and Chamberlain, 2008). However, mental illness is a common antecedent to homelessness. The Office of National Drug Control Policy estimates that 30 percent of the population of individuals experiencing chronic homelessness are living with a serious mental illness (SMI) (ONDCP, 2014; SAMHSA, 2017). There is a high prevalence of specific mental illnesses in the population of single individuals experiencing homelessness relative to the general population, including depression (20-25 percent prevalence across studies, as compared to 0.35 percent of the general population) and schizophrenia (5-15 percent prevalence across studies, as compared to 0.35 percent of the general population) (Martens, 2001; Perälä et al., 2007; Toro, 2007). There are systematic reviews that explicitly excluded studies of families because they were so rare that they deemed them a “special population” (Fazel et al., 2008, 2014).
Substance use disorders, especially alcoholism, are also a major problem for individuals experiencing homelessness, as well as an increasingly common cause of death (NIDA, 2013). Baggett et al. (2010) analyzed data from the 2003 Health Care for the Homeless study (n = 966) and found that both drug and alcohol use together was a major predisposing factor for experiencing homelessness. The combination of SMI and substance abuse is common in the population of individuals experiencing homelessness (Salit et al., 1998).
Health Outcomes Due to Homelessness
Spending time in a homeless shelter can also lead to negative health outcomes for individuals experiencing homelessness. Kelly (1985) found that homelessness increases the risk of developing health problems, including diseases of the extremities and skin disorders and increases the possibility of trauma, especially as a result of physical assault or rape. In addition, the Centers for Disease Control and Prevention reported outbreaks of tuberculosis in two homeless shelters, one in Duval County, Florida (CDC, 2012a), and one in Kane County, Illinois (CDC, 2012b). In 1990, McAdam et al. (1990) investigated the prevalence of tuberculosis in a men’s homeless shelter in New York City. Over a 73-month period, the authors screened more than 1,800 men and found an infection rate (positive PPD test, history of a positive PPD test, or active tuberculosis) of nearly 43 percent.
Chak et al. (2011) noted that individuals experiencing homelessness have higher prevalence rates of hepatitis C (HCV), particularly those who are infected with HIV. There are shared routes of transmission for the two viruses, and the authors noted that HCV prevalence rates ranged from 19 percent to 69 percent in patients experiencing homelessness.
Dirmyer (2015) investigated hospital readmission rates for persons experiencing homelessness in Albuquerque, New Mexico. One-third of these patients experienced a 30-day readmission to the hospital over the course of a 3-year period, with the most prevalent cause of readmission being neuropsychiatric disorders. The hospital readmission rate for patients experiencing homelessness was higher than national readmission rates and higher than the rate for Bernalillo County, where Albuquerque is located.
Overall, spending time in either a homeless shelter or being homeless and living “on the street” has diverse untoward health consequences.
Chronic Health Conditions
One of the first comprehensive assessments on the health status of persons experiencing homelessness was in the mid-1980s by the Social and Demographic Research Institute of the University of Massachusetts, Amherst. Data from 19 National Health Care for the Homeless Initiative demonstration projects were reviewed (Wright, 1990; Zlotnick et al., 2013). The prevalence of health conditions in the adult homeless population was compared to that of adults in the general U.S. adult population. Findings indicated that the prevalence of chronic conditions such as asthma, HIV/AIDS, tuberculosis, hypertension, diabetes, and chronic obstructive pulmonary disease was higher in the homeless group than in the general U.S. population (Zlotnick and Zerger, 2009).
Health and the Experience of Homelessness
Several studies have examined the prevalence of cardiovascular disease risk factors and adverse outcomes among persons experiencing homelessness compared to the general population. In 2002, Szerlip and Szerlip compared the medical charts of 100 patients in a homeless clinic in New Orleans, Louisiana, to those of 200 nonhomeless patients who attended another inner-city primary care clinic. They found that individuals experiencing homelessness had a higher prevalence of hypertension and smoking, but there was no difference in diabetes and total cholesterol compared to the general population. Other studies have confirmed the higher prevalence of smoking among homeless populations, but have not found a higher prevalence of hypertension or a difference in diabetes and total cholesterol (Lee et al., 2005). It has been suggested that for many risk factors, it is not their prevalence but the treatment and management of these conditions that is worse among those individuals experiencing homelessness (Jones et al., 2009; Bernstein et al., 2015).
Studies have shown that the prevalence of uncontrolled diabetes is higher among populations experiencing homelessness compared to the general population (Hwang and Bugeja, 2000; Lee et al., 2005). More recent evidence also suggests that the burden of cardiovascular disease is greater among subsets of the homeless population, especially those with mental illness. Among this subset, the 30-year risk of coronary heart disease, including (a) being diagnosed with coronary heart disease, (b) having a myocardial infarction, and (c) having a fatal or nonfatal stroke, is higher among individuals experiencing homelessness who also have a mental illness when compared to the general population. This higher risk was greater in men who were also substance users (Gozdzik et al., 2015).
Individuals experiencing homelessness have higher rates of cancer risk factors (e.g., higher rates of tobacco use), but are less likely to undergo cancer screenings. A study of homeless adults in Los Angeles (Chau et al., 2002) investigated cancer knowledge and screening. Although most of the study population demonstrated understanding of cancer screening, their actual screening rates were lower than for Californians broadly.
Asgary et al. (2014) examined colorectal cancer screening rates, predictors, and barriers in two New York City shelter-based clinics. The authors found that the majority of patients were African American or Hispanic, 76 percent were male, and 60.7 percent were homeless. In addition, “domiciled patients were more likely than homeless patients to be screened (41.3 percent versus 19.7 percent; P < .001). Homeless and domiciled patients received equal provider counseling, but more homeless patients declined screening (P < .001)” (Asgary et al., 2014).
It is not surprising that the experience of homelessness complicates the treatment of health conditions such as diabetes (the need for daily insulin shots) or needed mental health care (due to a lack of community- or shelter-based care delivery).
Mortality Among Individuals Experiencing Homelessness
Individuals experiencing chronic homelessness live shorter lives and, as a group, suffer significant excess mortality. Early studies in this area documented higher premature death rates (three to four times higher) in geographic zones that had a higher prevalence of persons experiencing homelessness, shelters, soup kitchens, and substandard housing compared to the general population (O’Connell, 2005). A recent observational study examined causes of mortality among formerly homeless men in Housing First programs, homeless individuals not in Housing First programs, and the general population (Henwood et al., 2015a). The study found that the causes of death differed between the Housing First group and the homeless individuals who were not in the program.3 Seventy-two percent of the men in Housing First programs died of natural causes, compared to 49 percent of the homeless group. Only 14 percent of Housing First men died due to an accident, compared to 40 percent in the homeless group. Infectious diseases caused 2 percent of deaths in the Housing First group, compared to 13 percent in the homeless group. Death due to hypothermia occurred in 6 percent of deaths in the homeless population, but was not a cause of death for men in the Housing First program.
The findings of more recent studies are consistent with earlier studies. According to data from a study by Baggett et al. (2013), the most common causes of death for individuals who had experienced homelessness in the Boston area were drug overdoses, cancer, and heart disease.4 Individuals in the Baggett et al. (2013) study were observed until either the date of death or until December 31, 2008. Among those who died due to drug overdose, over 80 percent of deaths were due to opioid overdoses, a trend mirrored in society at large (Doe-Simkins et al., 2014).
Studies outside of the United States have helped to establish homelessness as an independent risk factor for mortality. As an example, a study in Glasgow (Morrison, 2009) compared mortality data retrospectively over a 5-year period from 6,757 persons experiencing homelessness in the calendar year 2000 with 13,514 age- and sex-matched controls from the general population. The proportion of those dying in the homeless population was 7.2 percent compared to 1.7 percent in the general population. This four-times-higher rate of dying was independent of age, sex, and prior hospitalization. Cause-specific mortality due to drug-related deaths was seven times higher for those experiencing homelessness.
3 However, the two groups were not matched. Housing First clients are selected for the most disabled adults experiencing chronic homelessness. This is a selection bias in that the group with the worst possible health and psychiatric problems is not equivalent to the general population of individuals experiencing homelessness.
4 The authors note that data limitations made it impossible to determine who was currently homeless and formerly homeless at time of death.
In addition to health-related outcomes described above, studies have examined other outcomes related to the homeless experience, including unemployment, involvement in the criminal justice system, and poor educational outcomes.
FEATURES AND LEVELS OF HOUSING FOR INDIVIDUALS EXPERIENCING HOMELESSNESS
A number of programs have been developed to meet the needs of individuals experiencing homelessness. These programs are funded from diverse sources and by a range of mechanisms. Below is a brief summary of housing options that may be available to individuals experiencing episodic to chronic homelessness, with the primary focus on PSH models. A brief discussion of the financial mechanisms that might be used to support PSH is also provided.
Temporary Housing Models
Individuals and families experiencing temporary or situational homelessness due to job loss, economic hardship, domestic violence, or other short-term emergencies have very different housing needs from individuals experiencing chronic homelessness. There are several interim housing models for persons who experience situational and temporary homelessness.
Emergency shelter programs are for individuals or families who are in need of short-term shelter (Locke et al., 2007). These programs are designed to provide an immediate alternative to sleeping out of doors or in a location not meant for habitation and can include safe places for survivors of domestic violence and their children. This is the most temporary type of housing available and is meant to be a short-term safety net. Emergency shelters offer shelter overnight but often do not provide daytime access to the facility. Emergency shelters can secure funds through HUD to provide their clients with a range of essential support services, including mental health services, child care, case management, and outpatient health services, among others (HUD, 2013a).
Transitional housing provides up to 24 months of housing in supervised settings along with social services to help individuals and families prepare for permanent housing. It can be project based, so that residents move out when they exit the program or transition-in-place by assuming the lease at the end of the program. Transitional housing has been a mainstay of the homeless service system for families and individuals who are not deemed to need or who cannot find places in PSH.
Medical Respite Programs
Medical respite care is for individuals experiencing homelessness who are not yet well enough to be on their own. At the same time, they are not sick enough to continue a hospital stay. Without access to medical respite care, individuals experiencing homelessness are unlikely to successfully manage their post-hospital medical regimen. According to Kertesz et al. (2009), nearly 50 communities in the United States and Canada have created medical respite programs for individuals leaving the hospital while also experiencing homelessness. More recent qualitative data indicate that medical respite programs are useful because they provide linkages to outpatient care (Zur et al., 2016). Doran et al. (2013) systematically reviewed 13 articles in order to investigate the effectiveness of medical respite programs. The lack of evaluations of medical respite programs led the authors to encourage the creation of academic/university partnerships in order to better evaluate these programs.
Permanent Housing Approaches
HUD defines permanent housing approaches to addressing homelessness “as community-based housing without a designated length of stay in which formerly homeless individuals and families live as independently as possible” (HUD, 2018). There are two types of permanent housing: permanent supportive housing (PSH) for persons with disabilities and rapid re-housing. These program models follow the Housing First approach. In some communities, people experiencing homelessness also get priority access to long-term rental assistance in public housing or the private market, with the latter provided primarily by Housing Choice Vouchers. However, these programs typically have waiting lists, so are rarely available to people at the time they experience homelessness. These subsidies do not generally have any associated services.
The early PSH services and programs for individuals experiencing chronic homelessness were “treatment first” (Tsemberis et al., 2004). In this traditional model, individuals experiencing chronic homelessness and substance abuse and/or mental illness were required to be treated for their substance abuse or mental health issues prior to being eligible for permanent housing. This required individuals experiencing homelessness to demonstrate “housing readiness” in order to receive housing. Tsemberis et al. (2004, p. 651) noted that for those individuals experiencing chronic homelessness and desiring housing, the treatment-first approach presents “a series of hurdles” that the individual may not be able to overcome or may be unwilling to overcome to be eligible for housing.
To clarify, Housing First is treated in this report as an intervention where housing is provided to individuals experiencing homelessness with no requirement
for participation in services. Pathways Housing First, described below, is a particular model of HF; all HF programs are not Pathways Housing First. Pathways Housing First is described below, with the more general HF described afterward.
Pathways Housing First
The Pathways Housing First model was created in 1992 in New York City by Pathways to Housing (Tsemberis et al., 2004). At its core, founder Sam Tsemberis believes that housing is a basic human right, and therefore, individuals experiencing homelessness should have immediate access to housing. Unlike previous housing programs for individuals experiencing homelessness, Pathways’ Housing First model did not require efforts toward sobriety or treatment for mental illness prior to accessing housing. The fact that tenants were not required to participate in substance abuse or mental health services has remained an essential feature of PSH programs. Although comprehensive supportive services provided by interdisciplinary “assertive community treatment” teams or intensive case management are available, participation is voluntary. Pathways Housing First focuses on the housing needs of the homeless individual and views housing needs as “paramount” (Pearson et al., 2009). Currently, Pathways to Housing has programs in the District of Columbia, Vermont, and the Philadelphia area along with Canada and a number of European countries.5 The program philosophy is based on several tenets, including and primarily, that housing is a human right and individuals experiencing homelessness are given immediate access to housing, with no preconditions (Tsemberis, 2010).
Housing First Approach
The term “Housing First” is now commonly used in a generic sense, both for PSH programs with low barriers and for other programs much less intensive than PSH, such as rapid re-housing. Martinez and Burt (2006) refer to this as a “low demand” model because housing is made available but abstinence from drugs/alcohol is not a requirement. In 2016, California enacted a new law that encourages state programs to adopt a Housing First model in all programs for housing individuals experiencing homelessness. A statement from the U.S. Interagency Council on Homelessness (USICH, 2017a) says,
Housing First is a proven approach in which people experiencing homelessness are offered permanent housing with few to no preconditions, behavioral contingencies, or barriers. . . . Housing First is an approach that can be adapted by housing programs, organizations, and across the housing crisis response system. The approach applies in both short-term situations, like
rapid re-housing, and long-term interventions, like supportive housing. For crisis services like emergency shelter and outreach, the Housing First approach means referring and helping people to attain permanent housing.
In other words, the Housing First term has been expanded and broadened. In some cases, this dual use of the term “Housing First” has led to confusion.
According to Pleace and Beverton (2013), “from a strategic and policy implementation perspective, it has to be clear what is meant by “Housing First” (p. 23). Housing First might be best viewed as a philosophy of how PSH should be carried out rather than a specific type of housing. Similarly, HUD refers to Housing First as an “approach” (2014, p. 3), and states that “this approach may not be applicable for all program designs” (HUD, 2014, p. 4).
Tsai and Rosenheck (2012) noted that the services component of Housing First needs to address factors other than successful housing outcomes. Because social isolation is a major risk after housing, particularly for those in scattered-site housing, recent research has focused on adding peer support groups for veterans who have formerly experienced homelessness living in supported housing (Tsai et al., 2014) and on the inclusion of trauma treatment for homeless female veterans (Tsai et al., 2012).
Rapid re-housing is a program model that follows the Housing First approach in providing short-term rental assistance and services to families and individuals experiencing homelessness.6 The program also provides housing for individuals and families with other immediate problems such as domestic violence and substance abuse. Individuals experiencing chronic homelessness who are in need of PSH are not a target population for this program.
Some supportive services are provided as part of rapid re-housing programs, the most critical being assistance with identification of housing options (USICH, 2015b). Other services include rent and move-in assistance and case management services. The focus of the services provided in rapid re-housing is to help individuals and families resolve their immediate crises, which are most often financial in nature. HUD describes funding for rapid re-housing as short-term or medium-term, with the focus on the provision of assistance including financial assistance, housing search assistance, and targeted services for a period of 6 months (HUD, 2014). The Department of Veterans Affairs also operates a large rapid re-housing program referred to as the Supportive Services for Veteran Families program (see Box 2-1).
Permanent Supportive Housing
Permanent supportive housing is an umbrella term for the provision of ongoing, long-term housing coupled with supportive services for individuals and families experiencing chronic homelessness, the unstably housed, individuals living with a long-term disability, and individuals and families who face multiple barriers to accessing and maintaining housing. For the purposes of this report, the committee used the following definition: Permanent supportive housing (PSH) is defined as non-time-limited affordable housing matched with ongoing supportive services appropriate to the needs of the tenants. Note that this definition varies slightly from the formal definition of HUD, but this is the definition that the committee agreed on.
The critical components of PSH are the provision of long-term housing and voluntary supportive services for the residents, including access to mental health care and medical services. By providing housing as described above, PSH is designed to provide individuals experiencing chronic homelessness with a place to avoid the extremes of the elements and a stable place for addressing their health needs. The service piece of PSH is in part designed to address health needs by providing ongoing clinical support. PSH is designed to provide stable housing for very-low-income people who would not be able to sustain housing without supportive services. HUD argues that this is the population that needs to be served first in PSH, rather than on a first-come, first-served approach. A notice from HUD states that “PSH needs to be targeted to serve persons with the highest needs and greatest barriers towards obtaining and maintaining housing on their own—persons experiencing chronic homelessness” (2014, p. 2).
There is no set of agreed-upon supportive services that are core to the PSH model. The Corporation for Supportive Housing (Post, 2008) identifies services that PSH typically provides: case management, substance use treatment and mental health counseling, access to health care, support groups, life skills training; community social activities, and assistance with job hunting services. Participation in services, although encouraged, is not mandatory. Three primary approaches for operating PSH include:
- Congregate or “[p]urpose-built or single-site housing: Apartment buildings designed to primarily serve tenants who are formerly homeless or who have service needs, with the support services typically available on site.”
- “Scattered-site housing: People who are no longer experiencing homelessness lease apartments in private market or general affordable housing apartment buildings using rental subsidies. They can receive services from staff who can visit them in their homes as well as provide services in other settings.”
- “Unit set-asides: Affordable housing owners agree to lease a designated number or set of apartments to tenants who have exited homelessness or
who have service needs, and partner with supportive services providers to offer assistance to tenants” (USICH, 2017b).
Given the importance of housing as a social determinant of health, it is critical to find, create, and implement housing for individuals experiencing chronic homelessness. The World Health Organization (WHO) defines social determinant of health as “the circumstances, in which people are born, grow up, live, work and age, and the systems put in place to deal with illness” (NHCHC, 2016). People experiencing homelessness have been significantly impacted by a social determinant of health, leading to chronic health conditions, substance use, mental illness, and increased mortality. This realization led to the development of PSH, as defined above. Specific elements of PSH, as outlined in a Substance Abuse and Mental Health Services Administration (SAMHSA, 2011) evidence-based toolkit on creating and managing PSH programs, include the following:
- Tenants have a lease for their housing and have full rights of tenancy under landlord-tenant law.
- Leases for those individuals with psychiatric conditions are no different from the leases for individuals not having psychiatric conditions.
- Participation in supportive services, such as mental health treatment or substance abuse treatment, is voluntary, albeit encouraged.
- House rules are applied equally for all tenants, regardless of mental health status.
- There is no time limit on the housing, as long as the landlord and the tenant are in agreement about renewing the lease.
- Ideally, tenants are asked for their preferences regarding housing, with options that match the options available to individuals not experiencing homelessness at the same income level. If the housing is single site, however, there may not be other housing options.
- Housing is affordable, with tenants paying no more than 30 percent of their income for rent and utilities.
- The use of supportive services may change over time, depending on the needs of the tenant.
- Tenants choose which supportive services they take advantage of. Different supportive services are provided for different tenants, depending upon their needs.
- Supportive services are designed to promote long-term recovery and sustained access to housing.
- The provision of housing and the provision of supportive services are distinct and are managed by separate agencies.
SAMHSA also promotes “integrated housing,” meaning that PSH tenants should have opportunities to interact with neighbors who are not experiencing
substance abuse and/or mental illness. However, in single-site housing, this is difficult to achieve. (See the section on scattered-site versus single-site housing in Chapter 5.)
FUNDING SOURCES FOR PSH
Funding for PSH is complex and often requires innovative approaches to guaranteeing financing, including braiding together a number of different funding streams. This section reviews the predominant sources of funding used to pay for housing.
Continuum of Care Program
HUD’s Continuums of Care (CoC) program is a potential federal funding mechanism for PSH. CoC refers to a local planning group that coordinates and allocates HUD funding to agencies serving people experiencing homelessness (HUD, 2012b). CoC’s most recent funding competition encouraged the reallocation of existing funds to PSH and rapid re-housing (HUD, 2015d) and provided funding for new PSH projects.
HUD Section 8 Housing Choice Vouchers
Housing Choice Vouchers, more commonly referred to simply as Section 8 vouchers or subsidies, are HUD’s primary means of assisting low-income individuals and families to pay for safe and secure housing. These subsidies are long term and considered permanent housing (Technical Assistance Collaborative, Inc., 2012). Section 8 vouchers have also been used to address the needs of priority populations. The HUD-Veterans Affairs Supportive Housing (HUD-VASH) program, for example, is specifically designed to help meet the needs of veterans who are chronically homeless. The program blends HUD Section 8 vouchers and VA case management and clinical services. The program began in 1992 and funds are administered through local PHAs.
Additional Funding Sources
There are a number of additional federal funding sources that can be leveraged for PSH, including Supplemental Security Income (SSI), Low-Income Housing Tax Credits, and HOME Investment Partnerships. Other federal funding sources, including the Ryan White HIV/AIDS Program Services and SAMHSA grants, are described below. Other innovative non-federal funding sources include Social Impact Bond/Pay for Success models. Under the Social Impact Bond model, investors provide upfront funding to implement a social service project;
the government or a philanthropic organization then contracts to pay back the investors with a small premium if the project achieves its goals.
Ryan White HIV/AIDS Program Services
Several funding opportunities provide housing for low-income individuals experiencing homelessness who are HIV-positive. The Ryan White HIV/AIDS Program provides short-term housing assistance (2 years maximum) and some support services (HRSA, 2016). Funding is given to local communities and state agencies for projects that benefit low-income individuals living with HIV/AIDS.
The Housing Opportunities for Persons with AIDS (HOPWA), managed by HUD’s HIV/AIDS Bureau, has two grant funding streams for PSH for this population. The HOPWA Competitive Grant program and the HOPWA formula grant program provide funding for housing to eligible cities and states (HUD, 2016d). HOPWA assistance may also include support for substance abuse, mental health, nutrition, job training and placement, and assistance with daily living (HUD, 2016d).
Because of the correlation between HIV status and homelessness (Aidala et al., 2007), the provision of housing is an important strategy for improving HIV management, reducing high-risk behaviors, and lowering the possibility of transmission to others (Buchanan et al., 2009). This program also has been important in addressing the disparate impact of HIV/AIDS on racial and ethnic minority groups. African Americans with HIV/AIDS make up 52 percent of those served by HOPWA funding (HUD, 2016d).
SAMHSA provides funds through several grant programs for services for individuals experiencing homelessness, including the Grants for the Benefit of Homeless Individuals–Services in Supportive Housing, a competitive grant program that provides communities with funding for services relating to substance abuse, co-occurring mental health and substance abuse disorders, and other support services. The Cooperative Agreements to Benefit Homeless Individuals is also a competitive grant program that allows communities to provide services within PSH approach. Finally, the Projects for Assistance in Transition from Homelessness (PATH) program is a state block grant program that offers similar supportive services.
Views and perceptions about, definitions of, and the approaches to research and amelioration of homelessness have materially changed over time. Similarly, different types of housing for individuals experiencing homelessness have developed to serve different populations. Individuals or families experiencing short-
term homelessness have different needs than those individuals or families experiencing chronic homelessness. There are a number of forms of housing for individuals and families experiencing homelessness, with varying time limitations and differing levels of service provision.
PSH is designed to provide housing for individuals and families experiencing chronic homelessness, the unstably housed, individuals living with a long-term disability, and individuals and families who face multiple barriers to accessing and maintaining housing. PSH programs have two essential components: the provision of non-time-limited housing, and the provision of an array of voluntary supportive services. Pathways Housing First was an early model created to provide PSH focused on client choice, although the term “housing first” is now used more broadly as a general approach rather than a particular program.
Additionally, a number of federal financing mechanisms support the building and operations of PSH programs; many of these are state Medicaid options for which waivers may be required. A more recent model for funding PSH programs is the Social Impact Bond/Pay for Success, in which a program receives upfront funding from investors, typically a philanthropic organization, who then is paid back by a government agency when and if the program achieves its goals.