The Mental Health Parity Act of 1996

The parity bills now before Congress would build on the Mental Health Parity Act of 1996, which became effective January 1, 1998, and sunsets September 30, 2001. Often criticized for containing so many exceptions as to be ineffectual, the 1996 law prohibits employers and insurers from imposing annual or lifetime dollar limits on coverage for mental health benefits that are more restrictive than those applied to medical benefits.¹ A scaled-down version of a full parity bill sponsored by Sens. Pete Domenici (R-N.Mex.) and Paul Wellstone (D-Minn.), the law exempts the plans of employers with 50 or fewer employees. Also exempted are group plans that experience an increase in health benefit costs of 1 percent or more because of the requirements. Employers not exempted can avoid the cost impact of the parity legislation by redesigning benefit packages. For example, they may increase employee cost sharing or limits on inpatient or outpatient visits or they may implement managed care programs. The law does not require employers to offer mental health coverage. So they may drop mental health benefits (or their entire health benefit package, for that matter), if parity mandates are perceived as too onerous.

Tucked in an appropriations bill, the 1996 parity act utilizes a regulatory structure Congress had established a few months earlier in the Health Insurance Portability and Accountability Act (HIPAA).² Under this structure, states have the option of enforcing federal mandates directed at insurers and are not preempted from adding regulations that are consistent with the federal floor. Unless states pass enabling legislation and actively enforce the law, enforcement duties fall on DHHS’ Health Care Financing Administration. While 45 states have assumed enforcement responsibility for HIPAA’s main provisions, only about one-third have done so for the mental health parity requirements for insurers, in part because the requirements themselves are generally perceived as ineffectual, according to several state regulators.³

More than 20 states have passed mental health parity laws, ranging from measures that simply mirror the federal statute to those that impose far stricter requirements. So far this year, governors in six states have signed legislation expanding private-sector mental health coverage.⁴ These state laws cannot be applied to self-insured, private-sector health plans governed by the Employee Retirement Income Security Act of 1974 (ERISA) but may apply to insurers providing services to ERISA plans. There is a great deal of variability among the state parity laws. Many of the more recent measures establish coverage parity for mental conditions with a biological basis. Some state laws apply only to certain populations (state employees, for example).

Costs of Parity

While opponents of the 1996 Domenici-Wellstone full parity proposal claimed that it would cause health plan premium increases ranging from 3.0 percent to 11.4 percent, these estimates were generated on models assuming utilization in an environment with a low penetration of managed care. The Congressional Budget Office estimated that premiums would increase about 4 percent under the Domenici-Wellstone full parity bill and 0.16 percent under the limited parity measure that Congress eventually enacted.

Recent analyses show that the cost estimates of full parity may have been overblown because of both the ability of managed care to reduce mental health benefit costs and the penetration of managed care in the marketplace. Examining 1995 and 1996 claims data from 24 managed care carve-out plans offering “unlimited” mental health coverage with minimal co-payments (full parity), Roland Sturm of RAND found that overall costs increased only by about $1 per enrollee per year and concluded that assumptions used in the 1996 debates over parity overstated actual managed care costs by a factor of 4 to 8.⁵ In the plans studied, costs were reduced by lower rates of hospitalization, a relative shift to outpatient care, and lower payments per service.

Last year, a National Advisory Mental Health Council (NAMHC) work group simulated the costs of moving to full parity using a model developed by the Hay Group.⁶ Incorporating some of the more recent findings about the impact of managed care on costs, the model predicts that if an indemnity or PPO plan moved to full parity (absent a managed-care carve out) mental health costs would more than double, rising from $8 per member per month (PMPM) to $17 PMPM. Implementing full parity in a managed care carve-out would increase costs from $4 PMPM to $5 PMPM, or about 25 percent, according to the simulation. Because full parity would raise costs substantially in an unmanaged environment, such a mandate gives health plan sponsors and insurers a strong incentive to implement managed mental health care if they have not done so already. The model predicts that simultaneous implementation of full parity and managed care would reduce costs from $8 PMPM to $5 PMPM.

NAMHC reported that implementation of parity may accelerate the trend toward management of mental health services. Some employer and insurer groups argue that, by doing this, the parity mandates in effect remove from the marketplace a benefit option that many people may prefer. Employers and insurers are generally opposed to benefit mandates at either the state or federal level.

More Parity on the Way

In May, the Clinton administration announced plans to make health insurance coverage for mental illnesses and substance abuse comparable to that for physical ailments for more than 9.5 million federal employees, annuitants, and family members covered by the Federal Employees Health Benefits program (which is not subject to the 1996 parity law). The previous month, bills were introduced in both the House and Senate that would widen the 1996 parity requirements applying to other employers and to insurers.

Legislation introduced by Reps. Marge Roukema (RN.J.), Peter DeFazio (D-Ore.), and Bob Wise (D-W.Va.) would provide full parity for coverage of mental health and addiction services. The Mental Health and Substance Abuse Parity Amendments of 1999 (H.R. 1515) would prohibit group and individual health plans from imposing treatment limitations or financial requirements on the coverage of mental health benefits and on the coverage of substance abuse and chemical dependency benefits if similar limitations or requirements were not imposed on medical and surgical benefits. The bill would prohibit plans from imposing limits on the frequency of treatments, number of visits, or the scope and duration of treatment for mental health benefits if similar limitations were not imposed on medical and surgical benefits. It would also prohibit plans from imposing copayments, deductibles, out-of-network charges, out-of-pocket contributions or fees, annual limits, and lifetime aggregate limits for mental health benefits if similar requirements were not imposed on medical and surgical benefits. The bill, however, would not stop a group health plan from negotiating separate reimbursement rates, establishing different service delivery systems for different benefits, or from managing the provision of benefits through the use of pre-admission screening, prior authorization of services, and other such mechanisms.

While H.R. 1515’s provisions would apply to individuals diagnosed with any mental illness or substance abuse disorder, it would not require a group health plan to provide any specific mental health benefits. It would eliminate the sunset provision in the 1996 parity law and end the current exemption for employers who show that their health insurance premiums rose more than 1 percent as a result of complying with the parity act.

Domenici and Wellstone have introduced the Mental Health Equitable Treatment Act (S. 796), which would prohibit group health plans from setting arbitrary day and visit limits on services for all mental disorders. It would provide full parity, including equal co-payments, deductibles, and other out-of-pocket costs, but only for specified, severe, “biologically based” mental illnesses. It does not address parity for addictive disorders. As would H.R. 1515, S. 796 would remove the 2001 parity sunset provision and the 1 percent cost exemption. It would also expand the scope of the 1996 parity mandate by including small businesses with 25 or more employees (the 1996 law applies to firms with 50 or more workers).

Some mental health advocates have objected to the distinction that S. 796 makes between biologically based mental illnesses and those seen as having other causes.⁷ The bill would exclude many childhood disorders, anxiety disorders, and mild-to-moderate depressive disorders, as well as Tourette Syndrome and bulimia. Illnesses categorized as eligible for full parity would be schizophrenia, bipolar disorder, major depression, obsessive compulsive disorder, panic disorder, post-traumatic stress disorder, autism, and other severe and disabling biologically based mental disorders, such as anorexia nervosa and attention deficit/hyperactivity disorder. As of this writing the Congressional Budget Office had yet to develop cost estimates of the impact on premiums of either bill.

Economics behind Parity

While mental health parity is coming to legislative fruition in an age of widespread managed mental health care, it is a political response to an uneven benefit design that some economists argue was justified in the era of fee-for-service medicine financed by indemnity insurance. In markets for most types of goods and services, each consumer must pay the full cost of what he or she is buying. In the case of items financed by health insurance, individuals consume services without the same incentive to control their spending because a third party is paying the bills. (Political economists call this phenomenon “moral hazard.” The “excess” spending resulting from third-party payment is eventually shared by everyone in the insured group, thereby reducing everyone’s welfare.)

To control spending caused by the moral hazard phenomenon, health plan administrators and insurers often impose costs, such as co-payments or co-insurance, that are triggered by consumption. Because the moral hazard problem is greater for many types of mental health services than for many types of general medical care, an argument can be made that higher co-insurance levels are warranted for mental health care.⁸ (The RAND Health Insurance Experiment reported responses to reduced cost sharing were nearly twice as large for mental health care as those in general medical care.) Applying higher dollar or day limits to mental health benefits, however, is not as justifiable as a means to control moral hazard.

Another reason that an health insurer may restrict mental health benefits is to protect itself from adverse selection—that is, to prevent the plan from attracting a disproportionate number of sick people who will drive up premium costs, thereby hurting the firm’s ability to compete. An argument has been advanced that the threat of adverse selection creates an incentive to offer too little mental health coverage and that public policy-makers are justified in limiting cost-sharing techniques applied to mental health benefits. Richard Frank, Chris Koyanagi, and Thomas McGuire summarize the policy dilemma brought on by the underlying economics well: “A long-standing problem for policymakers has been to determine whether restricted benefits for mental health are there for a ‘good reason’ (moral hazard) or a ‘bad reason’ (adverse selection). There is amply evidence that both have been at work.”⁹

Advocates of parity argue that unequal benefit limits should be eliminated. There is evidence of a widening gap between coverage for physical and mental illness.¹⁰ According to the Hay Group, 57 percent of plans surveyed had day limits on inpatient psychiatric care in 1997, up from 38 percent in 1988. In 1997, 48 percent had outpatient visit limits, up from 26 percent in 1988. The Hay Group also estimated that behavioral health care benefits dropped from 6.1 percent of total benefit costs in 1988 to 3.1 percent in 1997.¹¹

Despite the detected increase in demand-side controls, much of the drop in mental health benefit spending is probably due to the impact of managed care. By 1998, managed behavioral health firms reported a total enrollment of up to 162 million Americans (30 percent more than in 1996).¹² The publication Open Minds surveyed the industry with regard to enrollment in five types of programs and found

• 36.6 million people in stand-alone behavioral health utilization review programs (costing about $1.50 PMPM).
• 31.4 million people in stand-alone employee assistance programs (EAPs) (costing about $2.50 PMPM).
• 16.9 million people in integrated managed behavioral health/EAP programs (costing about $4.00 PMPM).
• 32.2 million people in non-risk-based behavioral health network programs (costing about $1.00 PMPM).
• 45.1 million people in risk-based networks (costing about $5.85 PMPM).

While employers and insurers strive to cut mental health costs, a case can be made that financing treatment of depression and other mental disorders can increase worker productivity and improve corporate financial performance. Up to 10 percent of all adults experience clinical depression each year, costing the nation a total of $44 billion, according to estimates in a recent report by the Washington Business Group on Health.¹³ Of this total cost, 28 percent is associated with direct health expenditures, 55 percent with absenteeism and reduced productivity, and 17 percent with premature mortality costs. The report also concluded that clinical depression, which is concentrated among people aged 25 to 44, can be effectively treated in more than 80 percent of cases with medication and psychotherapy or a combination of the two approaches.

Estimates of the cost savings achieved by managed mental health programs are open to question in several ways. First, such programs typically are not responsible for managing drug costs. Data used to analyze such programs’ performance usually do not include the cost of drugs, and drug costs have been rising rapidly in recent years. One economist interviewed for this paper said that he recently examined a set of contracts showing capitation rates of about $3.50 PMPM for mental health services and $3.75 PMPM for psychotropic drugs. It is not known to what degree pharmaceutical spending or general medical spending for that matter has changed, if at all, in response to the shrinkage in benefit costs for inpatient and outpatient mental health services. In order to understand trends in total mental health costs, an analyst would have to examine at least three factors: mental health carve-out costs, drug costs, and the cost of mental health services rendered in a general medical setting.

Issue Questions

Among the issues to be discussed at this Forum meeting are the following:

• What has been the effect of requiring parity for mental health benefits? Have the passage of parity laws and the prospects of more parity legislation speeded up the shift to managed mental health care?
• What problems do parity laws solve and what problems do they create?
• What are the principal mechanisms that employers and insurers use to manage mental health costs in today’s marketplace? What is known about the ability of such strategies to control overall mental health and health care costs?
• In the era of managed mental health care, what new issues have emerged for public policymakers?
• Are managed care techniques applied in a different way or more intensely to mental health services than to other types of medical care?
• What are the prospects that Congress may pass a full parity bill this year? How might such legislation address the parity issue with respect to managed care?

Speakers

William Goldman, M.D., is senior vice president for behavioral health sciences at United Behavioral Health in San Francisco. He is also a clinical professor of psychiatry at the University of California, San Francisco. Dr. Goldman is currently a member (and immediate past chair) of the American Psychiatric Association Council on Economic Affairs. He also serves on the UCLA/RAND Research Center on Managed Care for Psychiatric Disorders Advisory Board; the NIMH National Worksite Program Research Advisory Group; and the NIMH Parity Workgroup. Previously, Dr. Goldman served as medical director for several managed care organizations as well as director of mental health, drug, and alcoholism services for the City and County of San Francisco and as Massachusetts Commissioner of Mental Health and Mental Retardation.

Richard G. Frank, Ph.D., is professor of health economics in the Department of Health Care Policy at Harvard University Medical School, which he joined in 1994. He also serves as a research associate for the National Bureau of Economic Research. Dr. Frank has served on many national and state governmental agencies. Currently, he is a member of the Substance Abuse and Mental Health Services Administration Advisory Council and the MacArthur Foundation’s Mental Health Policy Research Initiative. His primary areas of interest include the economics of health and mental health, the behavior of nonprofit institutions, pharmaceutical pricing, and the financing of health services for vulnerable populations. He is currently conducting studies on managed care design for mental health and substance abuse, risk adjustment of capitation rates, and financial incentives associated with medical group practices.