The requirements for achieving parity coverage for behavioral health services are similar to the requirements for a successful string quartet—a score, or guiding plan, and four carefully coordinated players. We are partway there.

The first requirement is public support for behavioral health services and decreased stigma. Strong consumer and family advocacy combined with impressive scientific progress have filled this first seat of the quartet well.

The second requirement is addressing the actuarial fear that behavioral care is a financial black hole and that parity, however desirable, is unaffordable. With help from the large behavioral health care companies, especially the multiple studies using the United Behavioral Health database, we are close to proving that behavioral care costs under parity are predictable and manageable. Parity is affordable. The second seat in the quartet is about to be filled.

The third requirement is legislative and regulatory enforcement. The valuable study by Salkever and his colleagues in this issue tells us why we cannot get to parity without legislation and regulation. The investigators tested the hypothesis that "professional judgment and sound care-management protocols" would allow plans to dispense with arbitrary benefit limits and consumer cost-sharing. Unfortunately, in a survey of 577 plans, they found that health maintenance organizations and carve-out programs were about as likely to use these techniques as "unmanaged" insurance programs were. The Mental Health Parity Act of 1996 and President Clinton's directive that the Federal Employees Health Benefits Program build in parity by 2001 create a strong push for parity, as does legislation moving rapidly in many states. If the strong economy persists, the third requirement for parity may be met within a few years.

The fourth requirement is evenhanded application of "medical necessity" criteria. Removing discriminatory cost-sharing and benefit restrictions will not create parity if criteria for deciding what constitutes "necessary" or "appropriate" care are set in a discriminatory manner. In the April 1997 issue of this journal, Hollingsworth and Sweeney showed that 40 percent of solid, state-of-the-art care for patients with severe mental illness would be excluded under typical standards of "medical necessity." Filling this fourth seat in the parity quartet will be as challenging as the first three. The technocentric bias of health care funding accords lower prestige to rehabilitative and "social" components of care. Divided funding streams invite buck-passing and finger-pointing between health, housing, and vocational sectors.

But we have made substantial progress in this decade. We can enter the 21st century with optimism.