Health Serious Mental Illness
February 23rd, 2021 28 Minute Read Report by Stephen Eide, Carolyn D. Gorman

Medicaid’s IMD Exclusion: The Case for Repeal

Inpatient psychiatric care forms a crucial part of America’s mental health system. Though most mental health services are provided on an outpatient basis, treating some serious mental illnesses requires a hospital setting. Inpatient treatment may be provided in a general hospital unit or a specialized psychiatric hospital. Within the context of Medicaid, specialized psychiatric hospitals are known as “Institutions for Mental Diseases,” or IMDs.

Federal law generally prohibits IMDs from billing Medicaid for care given to adults between the ages of 21 and 64 at a facility with more than 16 beds. This “IMD Exclusion” has been in place, in some fashion, since Medicaid was enacted in 1965. The intent was to prevent states from transferring their mental health costs to the federal government and to encourage investments in community services. The IMD Exclusion achieved its desired effect by contributing heavily to what’s popularly called “deinstitutionalization,” the transformation of public mental health care from an inpatient-oriented to an outpatient-oriented system.

This report argues that the IMD Exclusion has outlived its usefulness and should be repealed. It discourages states from investing in inpatient care, hampering access to a necessary form of treatment for some seriously mentally ill individuals. As a result, these individuals end up repeatedly in the emergency departments of general hospitals, “boarded” for lack of access to available beds, and overrepresented among the homeless and incarcerated populations. More broadly, the exclusion discriminates, through fiscal policy, against the seriously mentally ill.

Concerns that repealing the IMD Exclusion would lead to a mass re-institutionalization of the mentally ill are overblown. The population of public psychiatric hospitals today stands at about 5% of what it was before deinstitutionalization. Individuals in need of mental health care have access to a much greater diversity of programs and public services than existed before the 1960s, when institutional care was often the sole option. Strong legal regulations also now exist that did not exist when Medicaid was first passed—most notably, the “integration mandate” of the Supreme Court’s Olmstead ruling, which requires mentally ill individuals to be provided services in the community when those services are appropriate, are not of objection to patients, and can be reasonably accommodated.

Interest in repealing the IMD Exclusion has increased recently in response to a concern over bed shortages for the seriously mentally ill and persistent challenges with mental illness-related homelessness and incarceration. There have also been signs of bipartisan interest in a full and clear repeal. Under the Biden administration, mental health-care reform, beginning with the repeal of the IMD Exclusion, may present an opportunity for substantive bipartisan policy reform.



About 5% of the adult population is afflicted with a serious mental illness, which the National Institute of Mental Health (NIMH) defines as one that causes “serious functional impairment, which substantially interferes with or limits one or more major life activities.”[1] (Two psychiatric diagnoses commonly associated with serious mental illness are schizophrenia and bipolar disorder.) About one-quarter of adults with serious mental illnesses are on Medicaid; among adults aged 18–64 on Medicaid, 8.2% have a serious mental illness.[2]

Federal law requires all state Medicaid programs to cover inpatient hospital services and mental health services.[3] But states generally may not bill Medicaid for services provided to anyone aged 21–64 who is a patient in an Institution for Mental Disease (IMD).[4] The Centers for Medicare & Medicaid Services (CMS) defines an IMD as a hospital, nursing facility, or other institution with more than 16 beds—or more than 50% of the total beds in the facility—that is devoted to the diagnosis, treatment, and care of individuals with a mental illness (developmental disabilities, senility, and neurological disorders are not considered “mental diseases” in this context).[5] The IMD Exclusion pertains to both mental health-related and standard medical-surgical services. The exclusion also applies to services provided outside the IMD to a patient in an IMD.[6]

States determine which health-care facilities qualify as an IMD based on federal criteria.[7] One such criterion is the total percentage of hospital beds dedicated for psychiatric treatment. New York City’s Bellevue Hospital has one of the largest concentrations of psychiatric beds of any medical facility in New York State (approximately 330 inpatient beds for acute psychiatric needs).[8] But it is not an IMD because those beds constitute less than 50% of the hospital’s total. Also, an IMD can be a public or private facility and need not admit patients on an involuntary basis. Public-private partnerships have also been established, such as with the Sheppard and Enoch Pratt Hospital (Sheppard Pratt) in Maryland.[9] Residential treatment centers with more than 16 beds are also subject to the IMD Exclusion.


Though its precise features and application have changed slightly over time, the IMD Exclusion has been in force since Medicaid was first enacted in 1965.[10] At that time, state governments housed hundreds of thousands of patients in specialized psychiatric institutions. Since the 19th century, funding for care in specialized psychiatric institutions had been a state responsibility.

In the mid-19th century, the federal government seriously considered funding public mental institutions from the proceeds of public land sales (similar to the later Morrill Land-Grant Act program for public colleges). However, President Franklin Pierce’s 1854 veto of the Bill for the Benefit of the Indigent Insane reaffirmed states’ responsibility for funding mental health.[11] Pierce believed that federal funding would discourage states from continuing to make investments in “establishments of local beneficence” (referring to state psychiatric hospitals).[12]

That state responsibility became increasingly burdensome over time. By 1940, the institutionalized mentally ill population was about 188 times larger than it had been in 1840.[13] A sizable portion of that population was made up of older Americans with senile dementia or psychosis related to syphilis, patients with nonpsychotic ailments, or those with no discernible mental illness in need of long-term support.[14] From 1939 to 1950, states’ spending on mental health care rose by close to 160% in real terms.[15] “The states,” notes Ann Braden Johnson, “faced a grim future in 1950: based on their uniform experience over the past century, they could expect more admissions of more people who would stay longer, at prices that could only rise, presumably by the same or even more enormous increments.”[16] State officials were expressing the same anxiety about the costs of mental health services as state officials in the 21st century do about public pension costs.

New Deal policymakers challenged the traditional assumption of state responsibility for social programs; NIMH was established in 1949. Nevertheless, debate existed over what fiscal responsibility the federal government would assume for mental health care. In postwar America, mental health advocates and psychiatrists viewed state mental institutions as an expensive failure. They were founded to cure mental illness, but by the mid-20th century, they seemed mainly to be providing custodial care.[17] New York, in the late 1940s, was devoting almost one-third of the entire state budget to a system of public mental hospitals that were, on average, 21% over capacity and decrepit.[18]

As scholars such as Michael Katz and David Rothman have documented, states built mental hospitals in the 19th century as part of a broad movement to develop institutional solutions to social challenges.[19] Other examples include almshouses and orphanages. The federal government developed the modern welfare state as a replacement for the 19th-century state-led system. From an early stage in that development, it showed a reluctance to subsidize institution-based social programs run by states.[20]

Institutionalized Americans were initially excluded from Social Security. In 1950, Congress modified that prohibition but kept it in force for individuals in “mental institutions.”[21]

The first major federal commitment to mental health came with President John F. Kennedy’s Mental Retardation and Community Mental Health Centers Construction Act of 1963, which authorized tens of millions in annual funding for community-based services. Enthusiasm for treating mental illness outside hospitals had begun growing out of the perceived success that military psychiatrists had in treating trauma-afflicted soldiers during World War II and also with the advent of antipsychotic drugs such as chlorpromazine.[22] Even so, limitations of community services were acknowledged. Many viewed these services as complementary to hospitals, which they assumed would remain essential to caring for chronic patients who lacked homes or families to return to.[23]

The IMD Exclusion, as part of the Medicaid program enacted in 1965, aligned with prevailing anti-institutional sentiments. Congress sought to encourage investment in community-based, noninstitutional modes of mental health care, leave states responsible for what long-term institutionalized care would still be necessary, and control costs. Half of all hospital beds nationwide, around mid-century, were devoted to psychiatric patients.[24] Some scholars have estimated that, without the IMD Exclusion, Medicaid’s initial annual cost would have been nearly 80% larger.[25]

In 1955, about 560,000 Americans were committed to public mental institutions. At present, there are fewer than 40,000.[26] The number of state mental hospitals peaked at about 350 and has since declined to about 210.[27] To be sure, deinstitutionalization began a decade before the IMD Exclusion, but the exclusion accelerated its pace. In the decade before Medicaid’s enactment, the population of public mental hospitals had been declining approximately 1.5% per year. After Medicaid became law, the rate of decline rose to 6% per year.[28] General hospitals, which traditionally had not been major providers of inpatient psychiatric services— but which could bill Medicaid for that purpose— added thousands of beds during the 1970s and 1980s.[29] Nursing homes and group homes, both of which can bill Medicaid, expanded dramatically in the wake of Medicaid and the IMD Exclusion.[30]

Over the past 40 years, community services have risen from one-third of state mental health agencies’ budgets to three-fourths.[31] Welfare programs, expanded insurance coverage, the advent of managed care, and a broader array of professionals who provide services (social workers, psychiatric nurses, etc., in addition to psychiatrists) have made mental health care more affordable for more people. This has allowed for a substantial increase in the number of reimbursable services demanded, creating more reason for healthcare providers to supply those services. The structure of mental health-care financing used to be that states raised funds from general revenues to devote to direct services; now, states generally raise funds to match federal Medicaid funds.[32]

Problems with the Status Quo

It is broadly accepted among scholars that Medicaid financing influenced deinstitutionalization.[33] More controversial is the benefit of the IMD Exclusion to a mental health-care system dramatically transformed from that of 50 years ago. Unrestricted federal Medicaid funds for community-based services has not negated the need for IMD-based services. Psychiatric services should be arranged across a “continuum of care”[34] to meet the needs of different mental illnesses across a spectrum of seriousness and of people in different stages of recovery from mental illness.

“Boarding” is a clear manifestation that this continuum of care is lacking. Boarding occurs when patients with psychiatric symptoms have been assessed and admitted to a hospital but left in the emergency room (or some other equally unsuitable location, such as a hallway) for lack of a bed that is available and suited to their needs.[35] Boarding is not unique to psychiatric care. But according to the American College of Emergency Physicians, an advocacy organization, “it takes three times as long to find an inpatient bed for a psychiatric patient rather than [for] a medical patient after the decision to admit has been made.”[36] A survey of 328 emergency department directors by that organization published in 2008 (since which year the number of public psychiatric beds has declined) found that 79% of those directors boarded psychiatric patients.[37] Mental illnesses place a major burden on emergency rooms generally: serious thought disorders and mood disorders account for more than a million visits every year.[38]

A 2017 report by the National Association of State Mental Health Program Directors described boarding as “a widespread problem that is on the rise” and identified the IMD Exclusion as one of its main drivers.[39] Some hospitals board patients for a day or longer.[40] For those patients who are given access to beds, their stays can be cut short in the interest of rapidly turning beds over.[41]

News reports of mental illness-related tragedies regularly report that the mentally ill victim or assailant had spent years cycling through various public systems— criminal-justice and homeless-services systems especially. [42] Patients are often discharged from general hospitals before receiving necessary care or without being connected to follow-up care.[43] Their encounters with community mental health, homeless services, criminal justice, and general hospital health-care systems plainly failed to stabilize them. The incarcerated population has a higher rate of serious mental illness (at least 15%)[44] than the adult population more generally (about 5%),[45] and some surveys of the incarcerated population indicate that those with mental disorders have been found to have committed more serious offenses than those without mental disorders.[46]

Meanwhile, an estimated 25% of homeless adults have a serious mental illness.[47] Every major work about the “modern” homelessness crisis—which began around 1980—has extensively discussed serious mental illness.[48] Scholars of the earlier “Skid Row” era, even those writing as late as the early 1970s, devoted considerably less analysis to the issue.[49] It was during the 1970s that public psychiatric beds declined by about 250,000, or roughly 40% of the total beds lost since 1955.[50]

The IMD Exclusion, as a special exception to Medicaid coverage for a clinically necessary service, would be most justifiable if inpatient systems had far more beds than they needed. Indeed, inpatient systems were widely seen to be overcapacity in the mid-20th century. But it is difficult to see how that is still the case. According to the federal Substance Abuse and Mental Health Services Administration (SAMHSA), only three states have 10 or more public psychiatric hospitals, and only four host 50 or more general hospitals with psychiatric units.[51] In a recent Government Accountability Office (GAO) survey, 47 of 50 state Medicaid officers said that IMD frustrates their ability to provide a full continuum of care; 34 states characterized it as a “significant challenge.”[52] According to a 2008 analysis coauthored by current American Psychiatric Association president Jeffrey L. Geller, states should maintain 40 to 60 psychiatric beds per 100,000 persons.[53] At present, no states meet that mark; every jurisdiction other than Wyoming and Washington, D.C., has fewer than 20 psychiatric beds per 100,000 persons.[54]

Nationwide, there are fewer than 12 beds per 100,000 persons.[55] Only five other OECD countries have fewer psychiatric beds per 100,000 than the U.S.[56] Bed counts at public hospitals that specialize in psychiatric care, in short, have declined in every decade since the 1950s; the 2010s were no exception.[57] The brunt of fiscal austerity measures that government took in response to the 2009–10 recession was felt, in many states, more by psychiatric hospitals than community mental health services.[58]

Recent Developments

After the 2012 mass shooting in the Sandy Hook Elementary School in Newtown, Connecticut, Congress undertook a reform effort that eventually resulted in the Helping Families in Mental Health Crisis Act. That bill, first introduced in 2013,[59] passed the U.S. House of Representatives in July 2016 and contained a near-full repeal of the IMD Exclusion, giving states the ability to contract with Medicaid managed-care organizations to cover patients receiving treatment in specialty psychiatric hospitals.[60] However, when much of the bill was incorporated into the final 21st Century Cures Act that President Barack Obama signed into law in December 2016, the repeal provision had been weakened to a stipulation directing CMS to instruct state governments about “opportunities to design innovative service delivery systems” for Medicaid-eligible mentally ill adults and emotionally disturbed children.[61] The Trump administration fulfilled this directive with a letter sent to state Medicaid directors in November 2018.[62]

Also in 2016, CMS issued final regulations allowing states to receive federal matching funds to make capitation payments to Medicaid managed-care organizations on behalf of beneficiaries receiving short-term (under 15 days per month) IMD services “in lieu of” similar services available under state Medicaid plans.[63] Put simply, if IMD care is cost-effective and medically appropriate, Medicaid beneficiaries can be covered for short-term IMD stays. This was formally codified into law as part of the 2018 Substance Use-Disorder Prevention that Promotes Opioid Recovery and Treatment for Patients and Communities Act.[64] Clinical and financial benefits must be demonstrated and the specific plan services for which IMD care substitutes must be identified.

A separate mechanism through which IMD care can be partially covered is through Section 1115 waivers (named for Section 1115(a) of the Social Security Act). The secretary of Health and Human Services is empowered to give states authority to make changes to Medicaid that are deemed to further the program’s overall purpose without increasing costs. These Section 1115 waivers can be used, for example, to expand eligibility for certain populations, change benefits or reimbursement rates, or respond to emergencies such as the Covid-19 pandemic. Most states make use of these waivers, but they are not without problems.

The waivers must be renewed, and they are subject to changing stipulations from one administration to another, or even during a single administration. In the case of mental illness, a 1115 waiver for IMD payment was only recently made available, in November 2018. States with approved 1115 waivers meant to expand Medicaid eligibility for a different population—such as individuals with disabilities—must reapply again separately. These waivers provide payment specifically for short-term stays of acute care and limit the permissible number of days in an IMD, while still requiring actions to “ensure a continuum of care is available to address more chronic, on-going mental health care needs of beneficiaries” with serious mental illness, which can include long-term care.[65]

According to the Kaiser Family Foundation, seven states and Washington, D.C., have applied for, and six have received approval for, IMD Exclusion waivers for seriously mentally ill populations as of January 26, 2021.[66] As set out by the Trump administration’s guidance letter, states may receive Medicaid reimbursement for IMDs whose average length of stay for patients does not exceed 30 days. However, in the first state waivers to be approved, an individual limit of 60 days was required.[67] This is problematic because a facility-wide average length of stay under 30 days allows flexibility for some patients in need of longer treatment to remain covered; an average accounts for those patients who are treated in under 30 days (which is most patients), while an individual limit of 60 days does not.

The Medicaid managed-care regulations and Section 1115 waivers have given states important flexibility to pursue IMD-based care but not to the degree that a full repeal would, or to the degree that would be medically necessary for some patients. The 15-day-per-month limit for Medicaid managed care can be arbitrary. In the case of patients who have visited more than one facility in the same month, it can be difficult for hospitals to account for what services a patient received elsewhere and how much can be billed to Medicaid. Antipsychotic medications can take four to six weeks to have full effect and getting the dosage or medication(s) right may take longer for some patients.[68] Section 1115 waivers can also get tangled up in partisan debates. The Trump administration, for example, granted states waiver authority to impose work requirements for Medicaid, which was criticized by progressive groups and is expected to be rescinded under the Biden administration.[69]

At the same time, it’s notable that the Obama and Trump administrations both issued guidance letters for waivers for the IMD Exclusion. Even before its 21st Century Cures Act, the Obama administration granted states waiver authority to modify the IMD Exclusion in order to respond to the opioid crisis.[70] Thirty states and the District of Columbia have exercised that authority, and, as of January 26, 2021, an additional four states are awaiting approval.[71] Use of the expanded waiver authority, as well as the new funding options for Medicaid managed care, affects the accuracy of past cost calculations of full IMD repeal.

Though President Joe Biden has not specifically called for repealing the IMD Exclusion,[72] some of his competitors for the 2020 Democratic presidential nomination did, including Vice President Kamala Harris.[73] Further evidence of bipartisan support for repealing the IMD Exclusion may be found in the 422–2 margin by which the Helping Families in Mental Health Crisis Act passed the House of Representatives in 2016.[74]

Conclusion: The Case for Repeal

Medicaid’s IMD Exclusion was crafted for an entirely different era. During the last half-century, America built a system of community-based mental health services that did not exist in 1965. Income-support programs for the disabled, assertive community treatment, clubhouse programs, supportive housing, assisted outpatient treatment, supported employment, peer support services—these either did not exist in the 1950s, or they operated on a much smaller scale than now.[75] Nevertheless, a small subset of severely mentally ill individuals still needs inpatient treatment on a short-term, intermediate-term, and long-term basis. The IMD Exclusion inhibits those individuals’ access to medically appropriate care. As is implied even by supporters of the IMD Exclusion who argue that it prevents “needless hospitalizations,”[76] medical need, not financing, should primarily shape public mental health care.

The IMD Exclusion punishes states for their historical commitment to providing mental health care. The 19thcentury asylums, for all their faults, entailed significant expenditures at a time when tax bases were far weaker than they are now. Had state governments never made any special commitment to the mentally ill and left them consigned to jails and poorhouses, Congress may well not have felt the need to make an exception for IMD care when it enacted Medicaid in the 1960s.

Today, legal and economic restrictions against “needless hospitalization” exist that did not 50 years ago. Since Medicaid’s passage, states across the nation adopted “dangerousness” (to oneself or others) as the standard criterion for civil commitment, and Congress passed the 1980 Civil Rights of Institutionalized Persons Act, which regulates the quality of inpatient care. Most important, the U.S. Supreme Court imposed an “integration mandate” through its decision in Olmstead v. L.C. (1999).[77] Olmstead held that unjustified segregation of disabled persons constitutes discrimination in violation of the Americans with Disabilities Act. As such, the ruling requires mentally ill individuals to be provided services in the community when those services are appropriate, are not of objection to a patient, and can be reasonably accommodated.[78] The holding was a reflection of two judgments: placing individuals in an institutional setting who can manage and benefit from being in the community perpetuates unwarranted assumptions that persons so isolated are incapable of or unworthy of participating in community life; and confinement in an institution severely diminishes the everyday life activities of individuals, including family relations, social contacts, work options, economic independence, educational advancement, and cultural enrichment.

Nevertheless, Olmstead did not outlaw institutional-based care. Indeed, Justice Anthony Kennedy emphasized in his concurring opinion that “it would be a tragic event … were the Americans with Disabilities Act of 1990 (ADA) to be interpreted so that States had some incentive, for fear of litigation, to drive those in need of medical care and treatment out of appropriate care and into settings with too little assistance and supervision.”[79] The Olmstead standard requires the placement of disabled people into “the most integrated setting appropriate to their needs,” and for some people, that will mean IMDs.[80]

Modern IMDs are not designed as isolation wards; on the contrary, they are open to the point where they’ve been characterized as “uniquely vulnerable” to spreading Covid-19 infections during the current pandemic.[81] Policies on seclusion and restraint are drastically changed from the pre-deinstitutionalization era.[82] “Snakepit”–type scandals associated with mentally ill people being held in poor-quality or brutal institutional settings have become more common among jails and prisons than mental hospitals.[83] Instead of serving as further justification for the IMD Exclusion, as some assert,[84] Olmstead and related legal regulations are best seen as evidence that fiscal disincentives for institutional care are no longer justified in the way they may have been in 1965.

The larger purpose of Olmstead is to prevent social discrimination against the mentally ill and other disabled Americans. An even more specific focus on preventing social discrimination against the mentally ill may be seen in parity regulations that require health plans to provide behavioral health benefits that are no more restrictive than coverage generally available for traditional medical and surgical benefits.[85] If Medicaid’s core function is to attend to the health-care needs of low-income Americans, and some of those needs must be met in an IMD, it does not seem consistent to carve out separate financing arrangements for those forms of care versus others. In any event, given the parity laws, laws restricting civil commitment, and court decisions, mass involuntary reinstitutionalization is simply not a realistic danger. This is aside from the reality that fewer individuals would need to be institutionalized long-term based on diagnosis alone, given that many who constituted the institutionalized population previously now benefit from antipsychotic medications and other advances in modern medicine that make it possible to be treated in an outpatient capacity.[86]

Finally, the structure of today’s Medicaid system—which operates mainly in a managed-care environment, as opposed to the fee-for-service environment of previous decades, makes it irrational to think that cost-conscious insurance companies and managed-care organizations would allow for anything other than the minimum necessary inpatient stays, given the expense. Today’s Medicaid managed care (health insurance that is publicly funded but privately administered)[87] will pay for 15 days of treatment in an IMD.[88] But even with no day limit, as would be the case were the IMD Exclusion repealed, Medicaid managed-care organizations would provide significant downward pressure on long-term psychiatric care out of cost concerns. For those who can be treated successfully outside an institutional setting, or within a short window of institutional care, managed-care organizations would play a “patient advocate” role similar to defense attorneys under the Olmstead-based legal regime.

Long-term psychiatric care will always be expensive. It may be necessary, were the IMD Exclusion repealed, to develop a funding program to assist public psychiatric hospitals similar to the “disproportionate share hospital payments” program for safety-net hospitals that states have in the past used to fund IMD care.[89] But the first and most important step toward public mental health reform to take is to eliminate the IMD Exclusion.

Defenders of the exclusion argue that it protects funding for community services that would otherwise be crowded out by increased spending on IMD-based care.[90] However, crowd-out dynamics might just as well work the opposite way: federal fiscal relief for inpatient services could free up state funds, and capacity, to devote to mentally ill individuals for whom community services are most appropriate. According to one assessment of a Vermont demonstration program that used Medicaid funds for IMD care: “In Vermont’s experience, providing institutional care for the most acute patients reserves community-based services for those who do not need institutional care.”[91] IMD investments would relieve pressure on community programs that are ill-prepared to deal with mentally ill people in a state of crisis. Research examining the impact of Medicaid expansion through the Affordable Care Act finds that an overall increase in Medicaid spending did not lead to reductions in spending on other non-Medicaid categories, such as education or transportation.[92]

A more realistic assessment of crowd-out or tradeoff-type dynamics would focus less on the tension between hospitals and community services and more on the tension between different modes of community services. Supporters of the IMD Exclusion charge that focusing on IMD-based treatment represents a shortsighted focus on crisis. For them, substantive mental health reform requires expanding the network of services and programs available to stabilize people before and after they’ve entered a state of psychiatric crisis.[93] But the real problem is that many publicly funded community services do not serve the seriously mentally ill in a crisis state.

That problem goes back a long time. During the early years of deinstitutionalization, psychiatrists whose educations were funded by taxpayers went into private practice, and Community Mental Health Clinics focused their attention on individuals who would never have been considered for civil commitment.[94] As the number of diagnoses has expanded—and the number of Americans diagnosed at some point in their lifetimes with a mental disorder has increased—the number of claimants on public mental health resources has increased.[95]

New York City’s ThriveNYC and California’s Mental Health Services Act are examples of extremely well-funded investments in community mental health services whose outcomes have been negligible because of a holistic approach as opposed to one targeted to the seriously mentally ill.[96] Only community programs that “focus exclusively on people with serious mental illness”[97] truly offer an alternative to hospitalization. Programs without that focus can’t be said to be preventing hospitalization or serving as a safety net to stabilize people postcrisis.

Also antiquated are arguments that general hospitals can suffice for the mental health-care system’s inpatient needs.[98] General hospitals are, and will almost certainly remain, the preeminent provider of inpatient psychiatric care in the nation.[99] There are 1,033 general hospitals with separate psychiatric units, compared with 214 public psychiatric hospitals.[100] But general hospitals cannot be relied on to the extent that they used to. In past decades, general hospitals added beds while state mental hospitals were cutting theirs, thus relieving pressure on the system.[101] But psychiatric beds in general hospitals have been declining since the 1990s.[102] Private general hospitals have been cutting psychiatric beds to make more system capacity for more remunerative services. Ninety percent of all general hospitals with a separate psychiatric ward are run by a private (nonprofit or for-profit) organization.[103] Bed reductions by nonprofit general hospitals have created pressures in other parts of the public mental health-care system,[104] and these pressures have increased during the Covid-19 pandemic.[105] More basically, as specialized institutions, state IMDs provide a greater range of psychiatric services appropriate for more longer-term commitments than are available in general hospitals.

As for Section 1115 waivers, state officials report that the process is cumbersome, the terms can change between administrations and even during the same administration, and budget neutrality requirements focus only on cost savings within the Medicaid program itself.[106] CMS has had a history of changing what is allowable in terms of how states can fund inpatient treatment. As noted above, while CMS released guidance in 2018 that encouraged states to seek Section 1115 waivers for behavioral health, it wasn’t until after that letter was sent to state Medicaid directors that CMS stipulated that a 60-day total limit for an individual (as opposed to the presumed average 30-day limit for all stays) was required for waiver approval. As with the Medicaid managed-care regulation, the expanded waiver authority has weakened some of the IMD Exclusion–related perverse incentives and recognized the need for greater access to inpatient treatment as part of a full continuum of psychiatric care. Overall, though, these modifications are insufficient.

Repealing the IMD Exclusion would undoubtedly increase the cost of Medicaid, which is already expected to exceed $1 trillion in 2027.[107] But the true cost is a matter of dispute. The Congressional Budget Office (CBO) estimated that a full repeal, when it was proposed in the 2015 Helping Families in Mental Health Crisis Act, would cost $40–$60 billion over 2016–2025.[108] But CBO allowed that its estimate was “highly uncertain.” It has since become even more unreliable.

CBO estimates how much it would cost the government to enact a new program or change an old one, relative to the existing “baseline” level of expenditure. However, Section 1115 waivers and the 15-days-of-care-per-month allotment for Medicaid managed care have changed the baseline of current Medicaid funding for IMD-based care. CBO’s estimate also does not take into consideration the savings that other service systems could realize. To the extent that IMD repeal would encourage more IMDbased treatment for people otherwise confined to jails and shelters, there would be cost offsets in those systems. When the Trump administration proposed an optional full repeal as part of its FY21 budget, it estimated the cost to be $5.4 billion over 10 years.[109]

The benefits of longer-term inpatient psychiatric care, measured in weeks or even months, include stabilizing difficult cases and keeping them and society safe. Under current law, Medicaid’s reimbursable care for specialized psychiatric facilities is generally limited to facilities with 16 or fewer beds. That is economically impractical for a hospital that needs to hire psychiatrists, nurses, social workers, security staff, and other support staff. The 16-bed limit also applies to any residential program that cares for the mentally ill, including those that don’t utilize locked wards to which people are committed involuntarily. States should be pursuing greater availability of longer-term inpatient psychiatric care for more severely mentally ill Americans. But intermediate-length and intermediate-level treatment would also be encouraged by IMD repeal.

IMDs serve as the safety net of the safety net. They care for and treat the mental illnesses of individuals who cannot be accommodated in general hospitals or community-based services.[110] Repealing the IMD Exclusion would neither result in mass reinstitutionalization nor disrupt the community orientation of public mental health care. It would, however, remove the fiscal disincentive against providing more inpatient care, forestall further bed cuts, ease boarding-related strains in the health-care system, encourage investment in new service models, reduce social discrimination against the seriously mentally ill, and facilitate long-term care for those who need it. The chief beneficiaries would be the cohort of vulnerable seriously mentally ill individuals who are at extreme risk of incarceration and homelessness by their inability to thrive in a community setting.


See endnotes in PDF


The authors would like to thank Scott Dziengelski, Dr. Tim Murphy, Dr. Steven Sharfstein, Elizabeth Sinclair Hancq, and Dominic Sisti for their insights.


Are you interested in supporting the Manhattan Institute’s public-interest research and journalism? As a 501(c)(3) nonprofit, donations in support of MI and its scholars’ work are fully tax-deductible as provided by law (EIN #13-2912529).