Homelessness is not a new phenomenon in the United States, but its visibility has increased, and the composition of the homeless population has changed. Formerly the province primarily of single men, the homeless population today increasingly includes women, children, and families. The exact number of homeless persons has always been difficult to estimate because of differences in who is defined as homeless (eg, living on street [unsheltered] only or persons also living in an unstable housing arrangement), differences in time course (eg, on a given night or within the last year), and the fact that some homeless persons hide from view. Nevertheless, the best estimates are that on a given night in 2016, approximately 550 000 individuals were homeless, of whom 32% were unsheltered and 35% involved families with children.1 Los Angeles City and County had the highest number of unsheltered persons (32 803) of any US metropolitan area.1
Although there is genuine desire to help homeless people, there is also an increasing civic frustration with some of the behaviors of some homeless people, such as open drug use, hypodermic needles scattered in the streets, sidewalks blocked with shopping carts, aggressive panhandling, and urination and defecation in public. An outbreak of hepatitis A among homeless persons and injection drug users in several California jurisdictions has resulted in 600 cases, 395 hospitalizations, and 19 deaths thus far in 2017, illustrating the potential consequences of a lack of sanitary facilities.2 The public expects the government to resolve these problems, but many government officials are concerned about being associated with the issue of homelessness because the problem can seem so intractable. For example, when homeless encampments are forcibly emptied from one area, new encampments develop elsewhere.
Observers of the elaborate tent cities in some US urban areas might be surprised to learn that over the last 15 years, there has been substantial progress on how to house the homeless population. However, the challenge of how to expand successful models to meet this need remains.
There is no single reason that people become homeless. However, it is useful to distinguish those whose homelessness is a function primarily of economic forces (typically the recently homeless) and those who have a range of mental health and addiction problems in addition to economic issues (typically the chronically homeless).
Many metropolitan areas have experienced gentrification with loss of affordable housing. Funding for US Department of Housing and Urban Development (HUD) projects has decreased. For example, the proposed HUD budget for 2018 is $40.7 billion, $3 billion less than in 2010 and unadjusted for inflation.3 Jobs for unskilled laborers have diminished, at least in part, due to automation.
These economic forces compound the problems of people who are struggling with mental illness and addiction. Many are estranged from their families either because their relatives were unable to care for them or because their behavior resulted in eviction from their residence. Decreased use and availability of mental health institutions and a belief that persons with active psychotic disease should not be treated against their will if they are not of imminent harm to themselves or others have resulted in large numbers of mentally ill persons living on the street. Inexpensive potent street drugs, including heroin and methamphetamine, have increased addiction rates. Decreased incarceration of low-level drug users has resulted in more people, who once would have been in prison, being homeless.
Unlike most chronic conditions, homelessness can be cured, and the side effects of treatment (increased medication adherence, improved sanitary conditions, family reunification) are positive. Supportive housing (ie, permanent housing with case management services) is effective for chronically homeless persons.4 Housing First models (ie, programs that house people directly from the streets without requiring that they first become sober and engage in mental health treatment) are more successful than programs that house people only after they have successfully completed treatment.5 Housing First programs have successfully provided and maintained housing even for persons who have been on the street for years. Although the original Housing First programs used single buildings, scattered site programs, which are often easier to establish because they do not require entire buildings, are also effective.6 The notion that many homeless persons prefer to stay on the streets is a myth fueled by inadequate attempts to engage homeless persons and failure to offer options that are responsive to their needs (eg, accommodate spouses, pets).
For persons whose homelessness is primarily due to economic factors, rapid rehousing programs that provide time-limited funding and targeted services to help move homeless persons and their families out of shelters and into permanent housing in the commercial market have proven successful.7 Partial rent subsidies (ie, one-third of the rent) enable people with some income to house themselves or stay housed. Eviction prevention programs (ie, payment of back rent, legal help) are also cost-effective ways to prevent increased homelessness.
Although it is known how to house the homeless, to date, it has been very difficult to bring housing programs to scale. The greatest challenge has been insufficient financial resources. Supportive housing costs differ based on local housing markets, whether the facility is rented or owned and on the intensity of services. But nationally, the cost is approximately $15 000 per person per year. It does not require a formal cost-effectiveness analysis to know that the benefits of housing, in terms of increased well-being to the individual and the society, compare favorably with many medical interventions that are standard of care.
Initially, it was hoped that there would be sufficient cost savings from decreased hospitalizations and emergency department visits4 to pay for housing. However, although there are cost savings, randomized trials of supportive housing have not substantiated savings equal to the cost of housing.
Having Medicaid pay for supportive housing in cases of persons with chronic medical or psychiatric illness would help to solve the funding problem for many homeless persons. Although this may sound controversial, Medicaid already pays for supportive services associated with housing, but does not pay rent. This contrasts with nursing home stays, in which case Medicaid pays much of the full cost (ie, the cost for provided services and the facility stay). If Medicaid were to pay for supportive housing for persons with chronic medical or psychiatric illness, eligibility would be determined by medical necessity, just as is the case for a nursing home, or by need for durable medical equipment. It would be available to those homeless persons who need services to stay successfully housed.
There are understandable concerns that paying for supportive housing for persons with chronic medical or psychiatric illness may place too much burden on Medicaid, especially given its already huge size, complicated mission, and current efforts to curtail expenditures. Also, having Medicaid pay for housing would not help those who are homeless for purely economic reasons.
In the absence of federal funding, localities are struggling with alternative financing solutions. Many communities including Portland, Oregon; Austin, Texas; Santa Clara County, California; and the state of Rhode Island have passed bonds to build affordable housing (ie, housing at below-market rates). Although persons who enter affordable housing are not necessarily homeless, their low income puts them at risk of homelessness, and increasing the supply of housing diminishes the competition for existing units. In 2017, Los Angeles County passed a quarter-cent sales tax to fund services for the homeless, and Los Angeles passed a complementary bond measure of $1.2 billion to fund capital projects to house homeless people. A full array of services is planned including rental subsidies, new buildings, case management, abstinence-based housing for homeless persons completing substance abuse treatment, mental health services, and eviction prevention.
Beyond funding, it is also difficult to site housing programs because of neighborhood concerns, ranging from perceptions of increased crime to decreased property values. The current myriad funding streams for constructing supportive housing projects, combined with challenges in zoning and other approvals, mean that even when approvals can be obtained, construction of new buildings is costly and lengthy.
There is also a gap between entities that provide traditional housing and those that provide medical care that must be bridged. Many nonprofit organizations that provide housing wish to house the neediest people but do not have the capability to care for people with complex health problems. Health care systems that provide care for homeless persons usually have no home to discharge them to. Establishing partnerships between health systems and housing services offers a path forward. By supporting housing providers in their community, hospitals could provide much needed financial support for housing providers and fulfill their community benefit requirements under the Affordable Care Act. Housing providers, in turn, could help to transition patients who are chronically homeless from the hospital or the street into a safer, more humane environment. This would relieve hospitals of patients who may otherwise have prolonged hospital stays that are unreimbursed either because the patient is uninsured or because the patient is no longer acutely ill but cannot be safely discharged.
Bringing housing programs to scale for the entire population of homeless persons will require substantial funding, willingness of residential neighborhoods to allow construction of new facilities, and successful partnerships between housing providers and health care organizations. Although there are numerous successful supportive housing models in the United States, it remains to be seen whether the current public concern and frustration with street homelessness can be translated into an equally strong commitment to solve a solvable problem.
Correction: This article was corrected on November 8, 2017, to correct data involving the scope of homelessness to include the Los Angeles County.
Corresponding Author: Mitchell H. Katz, MD, County of Los Angeles, Department of Health Services, 313 N Figueroa St, Room 912, Los Angeles, CA 90012 (firstname.lastname@example.org).
Published Online: October 31, 2017. doi:10.1001/jama.2017.15875
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported that he serves as a member of the National Academy of Medicine committee on an evaluation of permanent supportive housing programs for homeless individuals.
Disclaimer: This commentary reflects his views and not that of the committee or the National Academy of Medicine.
Additional Contribution: Kenneth W. Kizer, MD, MPH, University of California-Davis, read an earlier version of this article and provided suggestions, for which he was not compensated.
Mitchell H. Katz. Homelessness—Challenges and Progress. JAMA. Published online October 31, 2017. doi:10.1001/jama.2017.15875