Policy Recommendations to Address Housing Shortages for People with Severe Mental Illness

Adults with mental health issues whose sole source of income is Supplemental Security Income—approximately $783 per month—are unable to pay monthly rent, which averages $1,022 per month, before accounting for other costs of basic living (2). Our system of financial support for our most disabled citizens—those with serious and persistent mental illness—assumes that the individual has other methods of support. This limited income source, without guarantees of other support, especially housing, cruelly predestines individuals with mental illness to a possible life of housing instability, including homelessness and cycling in and out of institutions.

In 2009, Newman and Goldman (3) cited evidence that individuals experiencing severe and persistent mental illness could live in independent housing and noted that social service supports should be available for those who need them. Although these authors concluded that the evidence base on effectiveness of various approaches to housing was slim, their conclusions nevertheless encourage policies promoting supportive housing. They also noted that access to housing was limited and that would-be tenants with mental illness reported discriminatory practices. Newman and Goldman called for expansion of opportunities and resources and for more research.

There are no set-asides or prioritization systems in federal housing programs for people with mental illness—the missing component. The lack of a coordinated directive from federal policy makers and a misapplication of federal nondiscrimination laws, such as Section 504 of the Americans With Disabilities Act (ADA) and the Olmstead settlements, perpetuate housing limitations for people with serious mental illness.

In 2020, Leopold (4) updated the research base and concluded that even in the absence of strong research answers to questions about comparative effectiveness of specific policies, experts have proposed a number of practical policy solutions. Leopold called for an expansion of funding in the federal government’s 811 program, and Mathis (1) proposed expanding options in Medicaid for housing. Expansion of resources, although needed, does not provide the missing piece in federal policy to address the problem. There are no set-asides or prioritization systems in federal housing programs for people with mental illness—the missing component. The lack of a coordinated directive from federal policy makers and a misapplication of federal nondiscrimination laws, such as Section 504 of the Americans With Disabilities Act (ADA) and the Olmstead settlements, perpetuate housing limitations for people with serious mental illness. This article analyzes the problem and offers policy proposals to increase federal funds while prioritizing this severely underserved population.

Approximately 116,000 people with mental illness are homeless in the United States (5), and, although estimates vary, as many as 843,000 people with mental illness are incarcerated (6, 7). These figures do not account for the thousands couch surfing, those living with aging parents who are unable to cope with the episodic nature of the illness, or those who remain stuck in other institutional settings.

The United States lacks a national database or federally coordinated data collection system to determine the number and type of housing units needed to house the nearly one million individuals with mental illness estimated to be homeless, incarcerated, or poorly housed (5–7).

The number of housing units needed to address this population is unknown. The United States lacks a national database or federally coordinated data collection system to determine the number and type of housing units needed to house the nearly one million individuals with mental illness estimated to be homeless, incarcerated, or poorly housed (5–7). No national database exists, because a federal policy to eradicate the problem has never developed.

Many states and local communities have developed their own housing goals, but often these have not addressed the need. In more recent years, goals have been embedded in the federal policy to end homelessness or to meet legal settlements between the U.S. Department of Justice (DOJ) and local governments to uphold the Supreme Court’s 1999 Olmstead decision, which found that the unjustified segregation of people with disabilities is a form of unlawful discrimination under the ADA. These efforts have worthy goals, but they do not address persons with mental health illness who do not meet the federal homeless definitions or who are overlooked in Olmstead settlements.

Laws developed to protect the rights of disabled persons have had unintended consequences in housing for people with mental illness. Section 504 of the ADA does not allow a qualified individual to be denied access to or discriminated against in federally funded programs or services on the basis of the person’s disabilities. This has been interpreted to mean that under no circumstance should one disability class have preference over another in housing access. However, Section 504 allows distinctions based on disabilities, if the distinction provides “equal access to housing” (8). U.S. Department of Housing and Urban Development (HUD) officials have stated that if Congress provides an authorizing statute or if the Administration provides an executive order, federal housing programs can limit occupancy to people with one particular disability (9).

Olmstead settlements have caused some local governments to believe that prioritizing federal housing resources for any particular disability group is a violation of law. Yet, the DOJ, which litigates these cases, found that the plaintiffs in M.G. v. Cuomo had standing to make a claim that New York State did not provide adequate community-based mental health housing and services for people with mental illness leaving state prison. The DOJ stated that without mental health housing, this disability group was being segregated upon release (in homeless shelters) or was at risk of reinstitutionalization, both violations of federal law (10).

The lack of a directive from the federal government, with the patchwork of funding, makes solving the shortage of housing for persons with mental illness a challenge, whether in New York City or Des Moines. A federal and state compact with robust financial supports and specific goals prioritizing this population is needed.

A federal and state compact with robust financial supports and specific goals prioritizing this population
is needed.

Capital is used to acquire land or buildings, construct new units, or rehabilitate existing units. There is a myriad of costs in these tasks, including architectural; engineering; legal; environmental; investor requirements; federal, state, and local regulatory requirements; licensing; fees; insurance; utilities; property taxes; and debt services.

The Low-Income Housing Tax Credit (LIHTC) program is the largest source of capital funding in the United States for housing that is built for low- or moderate- income households. A Treasury Department program, created by the Tax Reform Act of 1986, the LIHTC program gives state and local agencies the equivalent of approximately $8 billion in annual budget authority to issue tax credits for the acquisition, rehabilitation, or new construction of rental housing targeted to lower-income households (11).

The median cost to create one, one-bedroom apartment using LIHTC is nearly $186,000 in 2020 dollars (12). That per unit cost is heavily influenced by location and density of a project. Building an apartment building in Los Angeles costs more than in Ocala, Florida; 25 units is more costly per unit than 250. Other factors contribute to capital cost per unit, such as a lack of raw materials, newly imposed import tariffs on items such as appliances, or a legal battle against local citizens who object to a project arguing “Not In My Back Yard” (NIMBY), a common occurrence in affordable housing development, arising from neighbors or businesses objecting to a proposed project in what they consider their neighborhood and based on bias, misunderstanding, or fear.

In all jurisdictions across the country, there is no one source of capital for housing development, which requires a developer to braid as many as eight to 12 funding sources to cover all the costs associated with any project.

Unfortunately, an LIHTC award never finances 100% of a project. In all jurisdictions across the country, there is no one source of capital for housing development, which requires a developer to braid as many as eight to 12 funding sources to cover all the costs associated with any project. The time involved in cobbling together this patchwork can also be onerous. New developments can take 2 to 6 years. Time is money.

Competition for capital resources is high, resulting in additional costs if applications are repeated multiple times before award. The good news is that all of this complexity has spawned an industry of sophisticated affordable housing developers. The bad news is that most do not make a habit of developing housing for people with mental illness.

“Operations” refers to the expense of keeping the apartments functional and taking care of the asset over time. Landlords also like to make a profit, including nonprofit landlords. Rents are the primary source of revenues for operations.

Mom-and-pop landlords can set their rents at any level that their local markets will bear. Housing developments that use public funds normally must use fair market rents (FMRs) to set their rent levels. FMRs are established by HUD. Often additional compliance requirements translate into the need for rents that are high enough for a landlord to cover costs. What this means for low-income people with mental illness is that rental subsidies are critical to their housing success.

Rental subsidies come in two types: project-based rental assistance (PBRA) and tenant-based rental assistance (TBRA).

Rental subsidies come in two types: project-based rental assistance (PBRA) and tenant-based rental assistance (TBRA). PBRA subsidies are attached to a particular unit within a specific building. The subsidy stays with the unit or “project” no matter who the eligible tenant is at any given time. These are helpful to projects that serve extremely low-income tenants and that may specialize in meeting goals, such as to end homelessness. A consistent source of rents, such as PBRA subsidies, can financially stabilize a project. TBRA subsidies do as they sound—the eligible household receives the rent subsidy and uses it in any affordable apartment or house where a landlord is willing to accept it. These are helpful to the household by providing choice and portability, if there is an adequate supply of willing landlords whose units have passed muster with HUD’s housing quality standards.

HUD provides the bulk of rental subsidies for eligible households through mainstream housing choice vouchers (HCVs)—formerly referred to as Section 8 vouchers. Other HUD rental subsidies are embedded in rental units owned by local public housing authorities, in homeless assistance programs, in the HUD 811
program (described below), and in special, short-term interventions, such as the nonelderly disabled category of HCVs. Some local and state governments provide rental subsidies to address a local issue, although many are intended for temporary assistance and expire after 3, 6, or 12 months. Nowhere are rental subsidies an entitlement. Waitlists are long, and there are challenges for people with criminal histories or ongoing issues with recovery to obtain a place in line.

When a disabled person signs a lease and lives independently, often his or her independence is actually reliant on some level of support. Supports can vary depending on the individual’s need and desires. Supports may include help moving into a unit; provision of household items; help with access to community- based services, health care, and treatment; support with building a community of friends; and help finding employment. Service supports help a person with mental illness maintain housing, such as by advocating with a landlord when a tenant knowingly or unknowingly violates the lease or assisting the tenant to complete the rental subsidy recertification paperwork.

Therefore, state and local governments or local philanthropy supplement Medicaid, which leads to inconsistencies across states and locales in the levels of services that a person with mental illness receives to live independently.

Medicaid is the primary source of revenue for some of these supports. In some locations, Medicaid does not cover what are called “tenancy supports,” only the obvious medically related services. Therefore, state and local governments or local philanthropy supplement Medicaid, which leads to inconsistencies across states and locales in the levels of services that a person with mental illness receives to live independently. Services that are offered and accepted by the tenant on a voluntary basis are also most effective in sustaining housing (13).

All three legs of the housing stool—capital, operations (rent subsidies), and services—are oversubscribed or underfunded to the need. The sophisticated affordable housing developers have a choice of developing new units for a variety of moderate- or low-income households. Private landlords in most markets look for tenants with higher incomes or better tenancy histories. It is more complicated when developing for, and then renting to, persons with episodic health issues. As a result, nonprofits or local governments have typically delivered housing for this population at insufficient quantities.

From 1960 to 1990, the federal government invested funds in two legs of the stool, capital and operations (as rent subsidies), for new housing for “handicapped” people as a subset of an elderly housing program called 202. This was the first federal attempt at developing housing for people with disabilities. The need for this housing continued to be evident, and in 1990, the Cranston-Gonzalez National Affordable Housing Act provided for a special program, Supportive Housing for Persons with Disabilities, referred to as 811 (“eight-eleven”) (14).

Through 811, nonprofit organizations applied directly to HUD for capital grants, accompanied by 5-year commitments for rental subsidies as PBRA. Housing types for development could include group homes, independent living facilities, multifamily rental units, condominium units, and cooperative housing with limitations on size—not more than eight or 16 units depending on the type. The nonprofit receiving the 811 funds committed to providing the third leg of the stool—services. The 811 program also awarded stand-alone TBRA subsidies but at a smaller amount.

This model provided nonprofit groups, many of which were small organizations, a single source application for two of the funds hardest to come by for affordable housing. This was a boon to local providers wanting to house their own clients and for those in areas with little or no capital. The commitment for 5 years of rental subsidy with the capital allowed clients to skip the long waiting lists for other rent subsidies.

What did 811 produce? Actual numbers of units created for persons with disabilities since 1960 are unknown, but HUD estimates that 30,000 units were created from 1990 through 2010—or 1,500 new units a year (15). That is a paltry and low national production pipeline. The units had to be available to any disability class as long as the person was age 18 to 61. As a result, many units did not house people with mental illness, further diluting the impact for this population.

Little has been published outlining successes of the early 811 model.Instead, issues arose that made people question its efficacy and cost. A 2019 Congressional Research Services report to Congress on federal housing programs stated that federal officials worried about the cost per unit (14). According to 811 program experts, advocates for the disabled community argued for more integration, others wanted to attract the sophisticated developers who used LIHTC, and Congress wanted to shift its role in housing production to states (personal communication, Sperling A., National Alliance for the Mentally Ill; Sloane L., Technical Assistance Collaborative, Feb. 2021). The TBRA (tenant-based subsidy) component of 811 needed regular reauthorizations to preserve existing units. In other words, rent subsidies needed to be maintained beyond the initial funding period so that existing tenants did not lose their housing. This squeezed the program from producing new units, because annual Congressional appropriations were limited. In addition, the 2007–2008 housing market collapse caused all affordable housing industry stakeholders to rethink financing of housing.

With advocates’ assistance, and with an eye on leveraging the largest capital source—that is, LIHTC—811 was amended. The 2010 Frank Melville Supportive Housing Investment Act did not do away with the original model, although some housing providers thought so. In fact, the Melville Act amended 811 to provide
an incentive for states to be more involved and to create greater opportunities for a small number of units dedicated to people with disabilities in large-scale affordable housing projects.

The 811 program’s competitive grants to states source the three legs of the stool differently, according to a state’s plan. Capital is entirely from other sources, most notably from LIHTC. Operations are 811, using project rental assistance (PRA), with important caveats: the services must be guaranteed from the state, primarily from Medicaid, and units dedicated to people with disabilities are maxed at 25% in a project.

Twenty-seven states are contracting with HUD for 811 PRA. Nearly 5,500 811 PRA units were identified during the years 2015 to 2020 (federal funds did not flow until 2013) (16). The units, integrated into existing affordable housing projects, are providing an important alternative to other options for persons with disabilities. In some states, where an Olmstead settlement exists, 811 PRA is a positive step toward meeting integration requirements. Having another federal rent subsidy is an important incentive for a state to invest in the other two legs (capital and services) to meet requirements.

Some believe 811 PRA is producing more units. However, the rate of “production” is actually slightly less than with the former model. Nationally, 5,500 units in the course of 5 years is only 1,100 annually. Although 5,500 units were identified, only 1,901 households are living in the 811 units (16). All new programs take time to ramp up, and in time other “identified units” may be filled. But is this a success for people with mental illness? In addition to this low rate of actually housed people, there have not been consistent annual congressional appropriations for 811 to significantly expand new housing units.

A finding that needs further monitoring, according to 811 program evaluators, is that people with mental illness are ineligible for admission to units at higher rates than other disability groups (17).

Once again, there is evidence that as with other subsidies not earmarked for people with severe mental illness, many units are going to people with other disabilities.
A finding that needs further monitoring, according to 811 program evaluators, is that people with mental illness are ineligible for admission to units at higher rates than other disability groups (17). HUD’s 811 PRA success story is of a man in Massachusetts who moved from a nursing home into an 811 apartment (16). His disability was physical.The sample client for Ohio’s 811 program training modules is a person with a physical disability; and providers who talk of the success of the 811 PRA are those who work with people with development disabilities (personal experience of the author, who completed Ohio’s 811 PRA program’s online training curriculum, Feb. 2021). The federal government’s website on the ADA touts the successes of Olmstead settlements, but the stories are primarily about individuals with developmental disabilities (18).

The 811 program’s competitive grants to states source the three legs of the stool differently, according to a state’s plan. Capital is entirely from other sources, most notably from LIHTC. Operations are 811, using project rental assistance (PRA), with important caveats: the services must be guaranteed from the state, primarily from Medicaid, and units dedicated to people with disabilities are maxed at 25% in a project.

The 811 program currently is not structured or funded for the necessary impact on people with mental illness who need housing. Structurally, Congress would again need to change 811 language to accomplish any significant impact. Language changes should require prioritization of people with mental illness or a set-aside of units for this population and additional local mandates for awards, such as collaboration on assessing and housing the most vulnerable or those with mental illness who cycle in and out of public institutions. Documented local involvement in the process for assessment should cross disciplines and systems, including criminal justice, hospitals, mental health, and housing.

Language changes should require prioritization of people with mental illness or a set-aside of units for this population and additional local mandates for awards, such as collaboration on assessing and housing the most vulnerable or those with mental illness who cycle in and out of public institutions.

Capital should continue to be available for a variety of housing types, allowing for local decision making based on current inventories while retaining the option of integration within larger housing projects and supporting consumer choice in housing as much as possible, given limited resources. HUD should require local or state assessment of need so that capital and rent subsidy requests match the local need and tenant choice in the types of housing available. To achieve decent production levels, funding within 811 for the two legs of the stool— capital and rental subsidies—would need to be boosted to five to ten times recent appropriations. The nearly one million people with mental illness who need housing now or will need housing on institutional release (5–7) cannot be assisted by a few thousand new units.

To achieve decent production levels, funding within 811 for the two legs of the stool—capital and rental subsidies—would need to be boosted to five to ten times recent appropriations.

To meet the third leg of the stool—service supports—in a manner that many of these individuals will likely need, Congress should direct federal agencies to review all existing programs and awards to increase service supports and collaboration with HUD’s 811 program.

For example, new 811 grants could be made only where Medicaid expansion and Medicaid waivers or state plan amendments are in place for tenancy supports, which would help ensure housing success. Other service support funding streams could be found in federal departmental programs within the DOJ and the Substance Abuse and Mental Health Services Administration.

Two federal initiatives, the federal Plan to End Homelessness and the Community Living Project, have taught us that anything less than a federal directive is a hobbled response to a critical social problem. A federal policy and commitment to develop housing for people with mental illness must include federal authorizing legislation, media and public relations investments, federal technical assistance on best practices, federal interagency cooperation, requirements for local collaborative bodies that are charged with investing in and monitoring audacious goals, federal funds that leverage mainstream and philanthropic funds, data collection, and research and evaluation. The federal policy to address the housing shortage for people with mental illness could start with 811 but must include far more than one HUD program.