Behavioral Health Parity

Consequences for Federal Employee Health Plans

Fact Sheet

In recent years, the push for behavioral health parity — health coverage for mental health and substance abuse care that equals coverage for general health care — has gained momentum. Concerns persist, however, that parity could invite overuse of services and drive up costs. To date, there has been no national assessment of how parity affects utilization and costs, and little is known about how insurers implement parity rules.

In 2001, the U.S. government required the Federal Employees Health Benefits (FEHB) Program to provide behavioral health benefits equal to its general health benefits. The FEHB Program insures 8.5 million federal workers, retirees, spouses, and dependents. Because it affects such a large group nationwide, the FEHB experience provides an important test case for understanding the effects of parity. A team led by researchers from RAND and Harvard Medical School assessed the effects of FEHB’s parity requirement on insurers and evaluated consumers’ use of services and spending.

How Did Insurers Respond?

The study analyzed the experience of 156 health plans, as well as the governmentwide BlueCross and BlueShield Service Benefit Plan:

  • Nearly all insurers complied with parity requirements by dropping cost-constraining benefit limits on behavioral health services.
  • Parity did not dramatically change most health plan operations, because the majority of health maintenance organizations (HMOs) and preferred provider organizations (PPOs) were already using “carve-outs” to manage behavioral care. (The term carve-out refers to an arrangement in which behavioral health benefits are managed separately from the general medical benefit.)
  • However, the move to carve-outs affected more than half of all beneficiaries, because they belonged to the BlueCross and BlueShield service plans, which switched to carve-outs in higher numbers.
  • A smaller number of plans also increased the use of specific managed-care mechanisms, such as utilization review. The parity effect in this area was smaller than expected because many plans were already using these mechanisms.

Effects on Service Use and Spending

The research team examined utilization and spending patterns in seven of the largest FEHB plans and compared them with patterns under a matched set of health plans serving large, privately insured populations but without parity benefits. Results showed that

  • use of mental health and substance abuse services increased at rates similar to those observed in comparison plans
  • total spending for services for beneficiaries decreased significantly for three plans and did not change significantly for four of the seven plans relative to the nonparity plans
  • out-of-pocket spending by beneficiaries also decreased under five of the seven plans, again relative to the nonparity plans.

Conclusion

These studies suggest that parity and managed care in combination can work to improve behavioral health care protection without increasing total costs.


This fact sheet is based on:

M. Susan Ridgely, M. Audrey Burnam, Colleen L. Barry, Howard H. Goldman, and Kevin D. Hennessy, “Health Plans Respond to Parity: Managing Behavioral Health Care in the Federal Employees Health Benefits Program,” The Milbank Quarterly, Vol. 84, No. 1, 2006, pp. 201-218; and Howard H. Goldman, Richard G. Frank, M. Audrey Burnam, Haiden A. Huskamp, M. Susan Ridgely, Sharon-Lise T. Normand, Alexander S. Young, Colleen L. Barry, Vanessa Azzone, Alisa B. Busch, Susan T. Azrin, Garrett Moran, Carolyn Lichtenstein, and Margaret Blasinsky, “Behavioral Health Insurance Parity for Federal Employees,” New England Journal of Medicine, Vol. 354, No. 13, 2006, pp. 1378-1386.


Copyright © 2007 RAND Corporation

The RAND Corporation is a nonprofit research organization providing objective analysis and effective solutions that address the challenges facing the public and private sectors around the world. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.

RB-9254 (2007)