Improving Our Federal Response To Homelessness

American Enterprise Institute

From the the book “A Safety Net That Works,” edited by Robert Doar

Share on Facebook  Tweet  Send email

Homeless assistance programs constitute the safety net of last resort for individuals and families who have fallen through all the cracks. As a result of extreme poverty, depleted support from relatives, and ineffective help from other safety-net programs, the homeless are unable to meet their basic housing needs. For many individuals, mental illness and substance abuse are linked to homelessness as well. It is essential that our programs catch the vulnerable populations that may otherwise go without shelter and provide the services necessary to lift them back up.

This chapter first describes who the homeless are, distinguishing between the single adults with high rates of substance abuse and mental illness, who frequently sleep on the streets, and the families, who most often are found in shelters. Then it summarizes our major federal efforts to confront homelessness. Finally, it offers recommendations to better prioritize assistance to the most vulnerable and, at the same time, provide more flexibility to local communities to try innovative ideas and boost program performance.

Who Are the Homeless?

Homelessness, as defined by the US Department of Housing and Urban Development (HUD), includes people who sleep in (1) places not intended for human habitation, such as parks, abandoned buildings, sidewalks, and vehicles; (2) emergency shelters, which provide nightly or short-term stays; and (3) transitional housing programs, which provide stays for 6 to 24 months.1 On a single night, just over 170,000 people are found sleeping in unsheltered locations, while close to 400,000 people are found in shelters (emergency shelters and transitional housing).2 That implies that 1 in 1,849 Americans are found on the streets and 1 in 819 Americans are found in a homeless shelter at a single point in time. Over the course of an entire year, 1 in 214 Americans will sleep in a shelter on at least one night.3 Homelessness is serious but not common, and for most, it does not last long.

In understanding who experiences homelessness and what that experience is like, it is important to distinguish between single adults and families. While 43 percent of homeless single adults are found on the street, only 10 percent of homeless families are found in unsheltered locations (see Figure 1).4 Furthermore, shelters for single adults are much more likely to have shared sleeping arrangements without the ability to come and go, while families more often receive private rooms for extended periods of time.

Homeless single adults and families also differ substantially in their personal characteristics. Among single individuals accessing shelter throughout the year, 71 percent are male, 75 percent are over the age of 30, and 47 percent have a disabling condition (e.g., a substance abuse problem or severe mental illness).5 Among adults in families, 78 percent are female, 53 percent are at or below the age of 30, and 21 percent have a disabling condition. Among children in families, 51 percent are below the age of six.6 In broad terms then, single adults experiencing homelessness tend to be middle-aged men who often have an addiction or a severe mental illness, while the typical homeless family is a single mother with young children.

Our Federal Homelessness Programs

Programs to assist the homeless are funded by federal, state, and local governments, as well as private sources. The federal government alone funds 15 separate programs across 8 different agencies, with the bulk of funding administered by HUD, the Department of Veterans Affairs (VA), and the Department of Health and Human Services (HHS). Altogether, the federal government spent $4.2 billion on these programs in 2014 (see Figure 2).

The largest and most influential of the federal programs is HUD’s Homeless Assistance Grants program, which allocated $2.1 billion dollars to local communities in 2014.7 This represents 50 percent of all federal funds spent on homelessness programs overall and 75 percent of all funds not specifically dedicated to veterans. Table 1 summarizes the interventions funded in part by HUD.

Emergency shelter is the safety net of last resort for people who find they have nowhere else to sleep. It is noteworthy then that only 12 percent of HUD’s homelessness budget can potentially be used for emergency shelter, with a portion of those funds used for other interventions as well.8 As a result, emergency shelters are largely funded by local governments and private sources.

The majority of HUD’s budget is used for longer-term, service-rich interventions. Transitional housing provides time-limited housing and supportive services intended to help people overcome problems and achieve self-sufficiency. Permanent supportive housing provides indefinite housing paired with supportive services for people with disabling conditions such as severe mental illness or substance abuse problems.

Other less intensive interventions funded by HUD include rapid re-housing and homelessness prevention. Rapid re-housing provides short- or medium-term rental subsidies along with case management to individuals or families soon after they become homeless. The intervention is intended to provide only the housing assistance and services needed to avert costly stays in shelter and prevent the recipient from returning to homelessness. Homelessness prevention is generally short-term financial assistance that helps families remain in their current housing.

These interventions are typically carried out by nonprofit service providers, which for the most part apply for HUD funding through local Continuums of Care (CoCs). Each CoC represents a specific geographic area (a city, county, or collection of counties), and its board includes representatives of local service providers, government officials, and at least one currently or formerly homeless individual. CoCs collect and rank funding applications, conduct annual counts of their homeless populations, maintain data systems that track individuals and families across participating service providers, and monitor the performance of providers.9 HUD also requires CoCs to allocate interventions to people using coordinated entry systems that prioritize the most intensive interventions to the most vulnerable people and less intensive interventions to those with fewer needs.10

While CoCs have significant autonomy to determine how funds are allocated across providers, HUD rewards CoCs that target particular subpopulations and use certain types of interventions. In recent years, HUD has encouraged CoCs (via higher scores on applications and thus more funding) to offer permanent supportive housing to the chronically homeless—individuals or families who have been homeless for the past year or four separate times during the past three years and who have a disabling condition.11

As a result, the national inventory of permanent supportive housing has increased by 69 percent since 2007.12 With these units, service providers are rewarded for employing a Housing First approach—providing housing and services without requirements that people maintain sobriety or engage with treatment.13 For people who do not require intensive services, HUD encourages CoCs to offer rapid re-housing in lieu of extended stays in emergency shelter or transitional housing.14

Agencies besides HUD have smaller homelessness budgets. The VA distributes $1.3 billion across six programs for transitional housing, health care, and supportive services for homeless veterans. HHS distributes $487 million across four programs for health care and youth-specific programming. Other agencies that distribute small amounts of funding include the Department of Homeland Security ($120 million), the Department of Education ($65 million), the Department of Labor ($38 million), and the Department of Justice ($25 million).15

Lastly, a special agency called the Interagency Council on Homelessness is charged with coordinating the collective federal response to homelessness. In 2010, the council released a strategic plan known as Opening Doors, which sought to end homelessness in a decade.16 Specifically, it set out to end homelessness among veterans and the chronically homeless by 2015 and among families by 2020. The word “end” is used loosely, however. The original plan called these goals “aspirational,” and an amended plan in 2015 clarifies that ending homelessness means preventing it “whenever possible” and otherwise making it a “rare, brief, and non-recurring experience.”17 Nonetheless, the rhetoric in combination with HUD funding priorities may have been instrumental in driving large increases in permanent supportive housing and investment in housing and services for veterans over the past several years.

While the federal government plays an extremely important role in assisting the homeless, state, local, and private efforts are important as well. There is tremendous variation across the country in the availability and quality of local efforts. For example, New York City is one of the few places in the country to offer immediate shelter to all residents who need it as a legal right, and for families, shelter often means a private apartment unit. Perhaps as a result, 22 percent of all homeless families in the country are found in New York City alone.18 Other cities often have waiting lists for shelters without legally mandated minimum quality standards and have much lower rates of family homelessness.19

Local variation in service quality not only complicates federal efforts but also brings to light a more fundamental problem; better options for the homeless increase the number of people who will take them up, including many who would have otherwise been housed. This means that services must be targeted carefully and must avoid promising benefits that are more generous than is necessary.

How to Improve Our Efforts

Our system of homelessness programs constitutes the final safety net for people who have no place else to go. It is a failure of that system when vulnerable human beings nonetheless sleep on our streets. While counts of the homeless have fallen since beginning in earnest in 2007, recent work suggests that much of that reduction may be due more to changes in methodologies and the quality of street counts than to how many homeless people there actually are.20 Meanwhile, a number of major cities have reportedly seen recent spikes in the numbers sleeping on the street, leading several to declare a homelessness state of emergency.21 Rather than double down on plans to end homelessness with old solutions, we should invest in innovative ideas that push progress forward, while ensuring that resources are prioritized to the people who need them most.

Better homelessness policy starts with making a fundamental distinction—homeless families are different than homeless single adults, and they require wholly different policy responses. Homeless families generally live in private rooms in shelters. They most often need temporary housing assistance to get back on their feet. Homeless single adults generally sleep on the street or in congregate shelters, and they often suffer from severe mental illness or substance abuse problems. They are more likely to require longer-term, service-rich interventions.

It makes little sense to lump these groups together. It is important that federal funds are intentionally allocated to each group separately and that they support the package of interventions best suited to each group.

Solutions for each group are discussed next.

Single Adults. The federal government should have a strong, proactive response for homeless single adults. This group includes the most vulnerable individuals within the homeless population—individuals who suffer from severe mental illness or chronic addiction and who live on the streets as a way of life. Acting alone, localities have strong incentives to avoid the most vulnerable, perhaps in hopes that they move somewhere else, rather than bring them in to provide the expensive services they often need. But simply increasing federal funding will not solve the problem. The federal government must insist that funds are prioritized to the most vulnerable and that service providers achieve the best possible results with the resources they receive.

Fortunately, significant progress has been made on prioritization. HUD requires communities to implement coordinated entry systems that assess the needs of individuals experiencing homelessness and prioritize them based on vulnerability. But HUD has not fully embraced coordinated entry. Driven by a goal of ending chronic homelessness, HUD rewards CoCs for targeting permanent supportive housing to chronically homeless individuals.

Chronic status is a crude distinction that should be abandoned as long as CoCs are using a vetted coordinated entry process for allocating assistance. An individual with a severe mental illness and a co-occurring addiction problem who has been homeless for 11 months is not considered chronically homeless, but this individual may be much more vulnerable and in greater need of supportive housing than a chronically homeless individual who has been homeless for more than a year but solely has an addiction problem.

Even perfect prioritization is not sufficient, however, if the people who are most vulnerable do not actually use the services they are offered. HUD should take a more proactive role in ensuring that CoCs successfully bring in the most vulnerable. This may in part mean increasing funding for high-quality outreach efforts. But it also means finding ways to hold CoCs accountable for outreach quality.

For example, CoCs could be required to submit a list of all individuals encountered sleeping on the street who suffer from severe mental illness or a chronic addiction problem, ordered by their vulnerability. The CoC would then be required to explain why any individuals are still sleeping on the street if they were deemed more vulnerable than others who were actually brought into supportive housing.

Better prioritization and outreach is only half the battle for single adults, however. More attention must be paid to increasing the quality of services themselves. If service providers can more effectively address underlying problems such as addiction or family disconnection, or if they can help clients receive treatment needed to improve their mental health, clients can more quickly be transitioned out of supportive housing programs and into private living environments with family or on their own, freeing up resources for future people in need. These outcomes are also inherently valuable.

Unfortunately, HUD has stifled innovation by focusing more on encouraging specific service models than on service-provider performance. In particular, HUD encourages service providers to adhere to a Housing First approach, which does not require sobriety or treatment for clients to maintain housing. This approach is not inherently bad, and it may be the best one for some individuals. But it has not been shown to be superior to other models in reducing substance abuse or improving mental health; it may even be less effective than certain models that have yet to undergo rigorous testing with randomized controlled trials.22

A better way to unleash innovation is to focus on performance.23. Service providers should be offered substantial flexibility in their service models, but they should be held accountable for their performance in helping their clients achieve desired outcomes. Performance measures should take into account the vulnerability of clients served so as not to punish service providers for accepting hard-to-serve individuals. This approach requires a renewed commitment to collecting high-quality data so that individual outcomes can be reliably tracked.

Over the past decade, HUD has taken the important step of requiring CoCs to establish data systems that track individuals through homeless assistance programs. But even better data are needed to monitor broader outcomes of individuals, including records from hospitals, emergency rooms, detox centers, mental health facilities, and jails. These records can indicate the costly use of other public services, as well as problems associated with poor mental health, addiction, and unstable living situations. HUD should offer additional support and flexibility to CoCs to build these broader data systems and evaluate service providers on performance measures.

Larger-scale innovation that altogether rethinks how homeless services are provided should be encouraged as well. For example, I discuss in a recent proposal how homelessness policy could be reoriented around smartphones and big data.24 Homeless individuals could be given free smartphones and full service plans in return for providing daily information on their sleeping locations, health status, and other outcomes. Research could be revolutionized with access to detailed, longitudinal data on an otherwise hidden population. Innovative interventions could be delivered to individuals via smartphones, and using randomized controlled trials, their effects could more easily be tracked. Statistical algorithms could help homeless individuals make data-based choices about service use and other major decisions. Newly created Homeless Innovation Centers staffed by data scientists and homelessness experts could be charged with dynamically creating unique packages of interventions for each homeless individual, in conjunction with implementing a dynamic research agenda. Flexibility from other federal restrictions could be offered for CoCs wishing to experiment with this or other innovative ideas.

Families. A strong federal response to homelessness among families is less necessary than that needed for single adults. Families rarely sleep on the streets, and they almost never do so as a way of life. Family shelters are not highly desirable places in which to live, but they usually offer families private rooms without forcing them to leave in the morning, as is more often the case with single adults. As a result, the number of homeless families in a community depends largely on whether quality shelter is actually made available. Rather than keep families off the streets, shelters are more likely to keep families out of doubled-up situations that may or may not be problematic. Nonetheless, the federal government has an important role to play in empowering communities to provide an efficient housing safety net for families.

Prioritization of families for the appropriate intervention is just as necessary for families as it is for single adults. Diverting families who have safe places to stay with relatives or who can be better served by mainstream programs is a necessity if resources are to be available for families who otherwise face unsafe housing situations, especially in unsheltered environments. Meanwhile, service-intense interventions should be reserved for families who would actually benefit from them. HUD already requires that CoCs use coordinated entry systems but again impedes efficient prioritization by artificially prioritizing chronically homeless families.

The actual interventions offered to homeless families should primarily emphasize helping families to quickly regain housing. Toward this end, HUD has encouraged CoCs to increasingly rely on rapid re-housing, which by providing temporary rental assistance, can be a cost-effective way to help families get back on their feet. But HUD should also encourage CoCs to experiment with modified ways of offering rapid re-housing assistance that could better increase self-sufficiency without major cost increases.

For example, families could be allowed to share housing (or “double up”) with others, with cost savings used to provide the family financial incentives for maintaining employment. HUD could also allow CoCs to experiment with providing greater autonomy to caseworkers assigned to families in rapid re-housing. Caseworkers could, for instance, require job-seeking activities or set expectations for other steps that would bolster self-sufficiency.25

At the same time that HUD has encouraged CoCs to invest more heavily in rapid re-housing, it has also discouraged investment in transitional housing. Given that transitional housing is expensive and its effectiveness is uncertain, this is a sensible decision.26 However, if CoCs implement reliable methods of identifying specific transitional housing programs that achieve better outcomes than other forms of assistance, HUD should offer federal resources to such programs as well. Finally, HUD should continue to fund homelessness-prevention assistance, but it should remain cautious given the inherent difficulty in predicting which families would otherwise enter shelters.


Our homeless assistance programs are supposed to act as the final safety net for people who would otherwise have no other place to go. It is a failure then when, despite the billions of dollars spent by the federal government alone on programs for the homeless, tens of thousands of people with severe mental illness and chronic substance abuse problems live on the streets. Meanwhile, homeless families do not typically face life on the streets, but it is important to provide efficient assistance to those who otherwise would be housed in unsafe environments. Federal and local government must do better.

Better federal policy for single adults requires fully embracing coordinated entry systems, holding communities accountable for actually serving the most vulnerable living on the streets, and rewarding service providers for their performance in helping clients achieve a broad set of outcomes. At the same time, the federal government should provide more flexibility to localities to meet these objectives in innovative ways, rather than stifle innovation with blunt political goals. For homeless families, a full commitment to coordinated entry and experimentation with cost-effective variants of rapid re-housing assistance is needed. Ultimately, improving our homelessness policies requires clear thinking about who needs what, accountability for results, and a culture of innovation.

To read the complete collection of essays, please click here.

Source: American Enterprise Institute