The Creation of Medicare and Medicaid: The Emergence of Insurance and Markets for Mental Health Services

The Medicare program was created under title 18 of the Social Security Act. Elderly and some disabled beneficiaries (those enrolled in Social Security Disability Insurance) are automatically enrolled in part A of Medicare, which pays for hospital care. Enrollment in part B, which pays for professional and other services, is optional and requires that a premium be paid. Nearly all part A beneficiaries enroll in part B.

For the most part, Medicare coverage for inpatient mental health has been generous compared with that of most private insurance programs. Psychiatric care delivered in general hospitals is covered on the same terms as all medical care. The beneficiary is responsible for a deductible. Medicare then pays 80 percent of covered costs for up to 90 days during a benefit period; the benefit period starts with an admission and ends after a 60-day period with no hospital admissions. Most Medicare beneficiaries pay little out of pocket because they are covered by supplemental polices that pay their copayments. Medicaid pays for copayments directly for individuals who are eligible for both programs.

Since its inception, Medicare has placed a 190-day lifetime limit on coverage for care in private and public psychiatric hospitals. The purpose of this restriction was to limit Medicare's financial responsibility for the provision of long-term custodial care for persons with mental disorders, a role traditionally played by state mental hospitals. The ideology of community psychiatry called for active treatment in short-term hospitals closer to the community. The designers of the Medicare program had this view in mind (4).

For most of its first 25 years, the Medicare program's part B coverage of outpatient psychiatric care consisted of a 50 percent copayment up to $500 in allowable charges. Eligible providers were limited to physicians. Because most supplemental policies did not cover outpatient psychiatric care, Medicare's coverage in this area reflected the low end of private health insurance policies. During the late 1980s, in part as a result of pressure from organized mental health interests, Congress instructed Medicare to make four changes in its mental health coverage: increase the limit on covered ambulatory care and then eliminate it altogether, cover medical management visits at parity with other ambulatory medical care, include psychologists and social workers among the covered providers in the program, and cover partial hospitalization (5).

Until 1982 Medicare reimbursed hospitals on the basis of allowed costs. Medicare now pays for inpatient psychiatric care in two ways. In psychiatric hospitals and most specialty units of general hospitals, Medicare pays hospitals a per-discharge amount set for each facility under the rules of the Tax Equity and Fiscal Responsibility Act (TEFRA) of 1982. TEFRA payments are based on 1981 costs trended forward. Subsequent changes in TEFRA have introduced some sharing of financial risk between hospitals and the federal government. Discharges from general hospital scatter-beds and a few specialty units are paid prospectively under diagnosis-related groups (DRGs). The mental health community in this case actively opposed treating psychiatric care in the same manner as all medical care because of concerns that the DRG payments would result in unfair payments to hospitals and possible undertreatment of hospitalized psychiatric patients (6,7).

Professional fees were initially set on the basis of Medicare's definition of "customary and usual or prevailing fees." Because of constraints on permissible price rises, professional fees were in essence based on a local fee schedule. In recent years professional fees have been based on a relative value scale. Psychologists and social workers are paid on a percentage markdown from psychiatrists.

The Medicaid program has been much more complicated since the passage of the Amendments to the Social Security Act in 1965. Medicaid is a joint federal and state program that pays for long-term care in nursing homes and other settings as well as acute care. The federal government sets its matching rate according to state per capita income. Federal matching rates have tended to vary between 50 percent for the wealthiest states and about 70 percent for the poorest.

Federal rules have traditionally contained a variety of mandated benefits and program requirements for eligibility and operations. Beyond the minimum benefit standards, states had the freedom to choose a number of optional benefits. Mandated benefits included general hospital inpatient care, physician services, outpatient services in general hospitals, nursing home care, and prescription drugs. Optional services have evolved over time. Those most relevant to the delivery of mental health care include nonphysician services, services provided in freestanding outpatient clinics, case management, rehabilitation, and home health care. Except for the case of prescription drugs, no copayments from beneficiaries are required.

Although the benefit package for Medicaid appears to be broad, states have often limited costs and coverage through their reimbursement policies. Most states impose per-year or per-admission limits on the number of covered inpatient days, or they pay on the basis of prospective rates. Fees to professionals are commonly set well below market levels of reimbursement, which limits the supply of physician services to Medicaid beneficiaries. These measures are partly due to the fact that Medicaid rules do not allow costs to be contained via patient cost-sharing provisions.

From the start of the Medicaid program, the federal government has prohibited payments for care of nonelderly adults in institutions for mental disease (IMDs). An IMD has been defined as any facility with more than 16 beds that specializes in psychiatric care. State mental hospitals, private psychiatric hospitals, and residential programs for mental and addictive disorders are examples of IMDs. Nursing homes may also be defined as IMDs if they are viewed as specializing in the treatment of mental disorders. The criteria for specialization in mental disorders and thereby being classified as an IMD include having 50 percent or more residents with mental disease.

The IMD exclusion had its origins in the 1950 Amendments to the Social Security Act. Those amendments prohibited payments for IMD patients and for patients with diagnoses of psychosis in any medical facility. The 1965 amendments in effect weakened that exclusion. The IMD rule was motivated by concerns similar to those underlying Medicare's 190-day lifetime limit. That is, the aim was for the federal government to avoid assuming the costs of long-term custodial care for psychiatric patients. This IMD policy encourages inpatient care in general hospitals where the federal government bears at least half the costs.

The creation of the Medicare and Medicaid programs served to put into motion a set of forces that would encourage the emergence of markets for mental health care and the privatization of mental health service delivery. It would also, in the coming years, repeatedly raise the issue of exceptionalism for mental health. That is, should policies made for general medical care be applied directly to the patients and the institutions of the mental health delivery system? Broadly speaking, parity for mental health care would be debated on many fronts in the context of Medicare and Medicaid.

Beginning with the passage of the Community Mental Health Centers Act in 1963, the decade of the 1960s marked a break from historical patterns in the role of the federal government in mental health care. The enactment of Medicare and Medicaid assigned the federal government a role as payer, insurer, and regulator of what would eventually account for nearly 35 percent of all mental health care services (8). By 1996 Medicare funds accounted for 14.4 percent of all mental health expenditures. As noted above, Medicare followed the design of private insurance plans of the day, which meant that patients and their physicians made choices about treatment subject to the terms of their insurance coverage. This approach placed elderly and disabled beneficiaries in the role of consumer—they could choose among a number of inpatient providers of mental health care as well as physicians providing outpatient care.

Thus patients were provided with opportunities to behave as consumers, which forced providers to take on the role of seller, with many of its commercial connotations. The terms of Medicare coverage were not neutral with respect to the types of care that were encouraged. For example, general hospital psychiatric care was encouraged at the expense of care in specialty psychiatric hospitals. One implication was that private nonprofit hospitals, the dominant form of general hospital, were preferred over public psychiatric hospitals. This shift in setting was one of several steps toward privatization of psychiatric inpatient care (9). It also served to direct Medicare psychiatric patients toward more "medicalized" treatment settings.

Medicaid as it was originally designed could be viewed as an open-ended voucher program (10). Patients were covered by the program for a defined set of services, for which they paid nothing or a nominal fee out of pocket. The program has evolved into one that covers a broad array of mental health services: from acute care in general hospitals, to physician and psychologist office visits, to case management services. The expansion of Medicaid to cover a full continuum of acute care and rehabilitative services has been the primary vehicle used by states to develop and finance community-based treatment programs for people with severe and persistent mental disorders.

Medicaid patients have traditionally been free to choose their mental health providers, although in practice care is not entirely free because low payment rates, as noted above, serve to limit the willingness of providers, especially those who are office based, to serve Medicaid beneficiaries. In theory, under voucher-type programs such as Medicaid, consumers choose providers on the basis of quality and convenience, because the out-of-pocket costs of care are zero. It was also expected that Medicaid beneficiaries would rely on providers that serve privately insured people. These features again drove the mental health system toward markets and private provision of care. In practice, low payment rates and geographic and informational constraints on the ability of Medicaid beneficiaries to comparison shop limited the potential for quality competition in Medicaid. In the end, the effort to include Medicaid beneficiaries in the same system of providers that more well-to-do citizens use was incomplete.

As a joint federal-state program with attractive matching provisions, the Medicaid program created strong incentives for states to reduce their role as direct providers of specialty mental health care and to expand their functions as payer and regulator. The program's design also sought to limit cost shifting by the states onto the federal budget. The consequences were that states increasingly designed public mental health programs so as to maximize the inflow of federal funds. The shift of tens of thousands of elderly people with mental disorders from public mental hospitals to nursing homes, noted above, is one of the clearest responses to the combination of generous matching arrangements and the IMD policy (11).

The number of state mental hospital residents over the age of 65 fell from 153,309 in 1962 to 78,479 in 1972. This reduction was proportional to the overall reduction in state mental hospital residents. During the same period the number of elderly people with mental disorders residing in nursing homes grew from 187,675 to 367,586, which was somewhat greater in proportion than the increase in the total number of people with mental disorders in nursing homes (12). It was particularly ironic that by 1969 the number of mentally ill people institutionalized in nursing homes and public mental hospitals had actually increased over 1963 levels (13). The modern nursing home industry comprises roughly 66 percent private for-profit facilities, which represents a clear shift from public provision to market-driven private health care.

Parity today is often argued in terms of equalization of the rules used to cover and pay for general medical and mental health services. In practice the application of the idea of parity often involves a more complicated set of criteria. The development of the Medicare program illustrates this point. Since the 1950s, when private health insurance began to be a broad-based phenomenon, the mental health community has argued against what was viewed as discrimination in the terms of coverage for mental health care. The design of the ambulatory mental health benefit under Medicare reflected the state of affairs under private health insurance in the early 1960s.

During the Medicare debate of 1965, many of the arguments made today against parity in benefit design under private insurance were first articulated. Among the points made against parity were the difficulty in defining mental illness, the lack of evidence on effective treatments, the high cost of covering mental health care, and the uncertainty in making actuarial estimates of costs. All of these points were invoked as reasons to limit ambulatory mental health coverage. Even today, the echoes of these arguments are frequently heard before Congress and state legislatures.

In the years that followed, issues of parity between general medical care and mental health care took other forms. For example, in 1983 Congress directed Medicare to pay for hospital care using per-case prospective payment based on DRGs. The mental health community was uniformly opposed to implementing the same payment policy for psychiatric cases as for all other medical services. The reasons were several. First, DRGs were shown to be a less useful method for classifying psychiatric patients than for classifying medical and surgical patients. Second, inpatient psychiatric providers are differentiated in their functions. Public mental hospitals care for the sickest patients; both private specialty hospitals and general hospital psychiatric units care for patients who tend to be sicker and costlier than patients treated in scatter-beds.

Because of weaknesses in the DRGs, these differences would not have been recognized by the payment system. Thus implementing prospective payment for providers of inpatient psychiatric care would have created an unfair outcome of financial winners and losers for reasons unrelated to their efficiency in providing high-quality care.

Finally, it was feared that the response to the strong incentives in prospective payment methods would especially disadvantage psychiatric cases. Specifically, a concern was raised that psychiatric patients might be prematurely discharged from hospitals, resulting in undertreatment. In the end, concerns with parity were dominated by fears that patients and providers would be unfairly hurt by the proposed policy.

The mental health community understood that the creation of Medicare and Medicaid would be important for psychiatry and the larger society. What was less well understood was the impact that these policies would have on the structure of the delivery system and the role of government in mental health care. Medicare and Medicaid set the foundation for the expansion of markets to the mental health sector and the accompanying privatization of many mental health services. Looking back, it turns out that Medicaid exerted enormous influence on the pace of depopulation of public mental hospitals, privatization of delivery, and the final location in the delivery system of people with severe mental disorders.

Today, as "managed competition" becomes the de facto model of health care delivery in the United States, the mental health system is evolving in a parallel fashion. Once again issues of privatization and the degree to which general approaches to managed health care ought to be applied to the mental health care sector are central policy concerns. State Medicaid programs are increasingly delegating management of the delivery system to private managed care companies with specialized approaches to managing mental health care that recognize unique aspects of the delivery of mental health care. The immediate effect is that private for-profit organizations are taking a central role in the restructuring of the Medicaid program and often the larger system of mental health care for people supported by public programs.

The most immediate results of the use of so-called managed behavioral health care carve-out programs have been a sharp reduction in inpatient care and expanded use of the continuum of community-based services. The mental health outcomes of these changes remain uncertain (14). As a result of these developments, the mental health field continues to wrestle with balancing disparity for mental health care and the need to address many of the special circumstances associated with mental health treatment. It must do so in the context of a wholesale reexamination of the social contract with elderly and disabled people regarding health care.

The author gratefully acknowledges financial support from the John D. and Catherine T. MacArthur Foundation. He thanks Nicole Pignatiello for excellent research assistance and Haiden Huskamp, Ph.D., for helpful comments.

Dr. Frank is the Margaret T. Morris professor of health economics at Harvard University, 180 Longwood Avenue, Boston, Massachusetts 02115-5899 (e-mail, [email protected]).