Commentary--The Fortune 500 Model for Health Care: Is Now the Time to Change?

James Maxwell and Peter Temin's findings fit well with my observations. And they are consistent with other sources (Henry J. Kaiser Family Foundation 2000; Marquis and Long 1999).

Managed competition was not intended to be a "theory" of employer behavior in health care purchasing (Enthoven 1980, 1986, 1993). Nor was it a prediction of what employers would do. It was a prescription for how purchasers could create incentives for consumers and providers to make economical choices so that overall resource allocation would improve. It was a critique of employer policies. Its ultimate goal remains a reformed health care delivery system of organized units, each accepting responsibility for the quality and cost of the care it delivers and each responding to powerful incentives to innovate in order to improve outcomes, patient satisfaction, and value for money.

The prescription was based on two observations. First, comparatively economical delivery systems exist, mainly prepaid group practices such as Kaiser Permanente and Group Health Cooperative of Puget Sound (Luft 1981; Manning et al. 1984). And second, experience in the Federal Employees' Health Benefits Program (FEHBP), the California Public Employees' Retirement System (CalPERS), and other groups showed that, where such delivery systems existed, given a choice and an opportunity to keep the savings, many people would migrate voluntarily to what they saw as value for money. Government and employers could then tie their financial contributions to the price of the lowest-priced plan and save money. And the whole system would be transformed, gradually and voluntarily, into one that delivered greater value for money.

What Is Managed Competition?

Maxwell and Temin offer a good working definition of managed competition for the purposes of their study, that is, the sponsor offers choices, fixed-dollar contributions, and information to consumers on quality, satisfaction, and price. But the actual prescription of managed competition is richer, as befits a complex problem.

First, it includes risk adjustment of premiums so that delivery systems that attract a disproportionate share of patients with predictably higher costs are not penalized. Without risk adjustment, there are incentives to avoid enrollment of chronically ill people and therefore to avoid excellence in treating them.

Second, for the strategy to achieve a reformed delivery system, a critical mass in each market must participate in managed competition. My first proposal, Consumer-Choice Health Plan, was for universal health insurance based on managed competition (Enthoven 1978a, 1978b). This could be approximated without the need for a single health insurance model if most insured people got their coverage through exchanges offering multiple choice of plan (Singer, Enthoven, and Garber 2001). (An exchange is an entity that links employers, employees, and health plans and arranges beneficiary multiple choice of plan. FEHBP and CalPERS, for example, are exchanges.)

Third, the goal is the reorganization of the delivery system. It is not merely carriers competing. Ten carriers each offering practically all the providers in an area is not "competition" in this sense. The competition that is being sought is competition among delivery systems where most of the value is added.

Has Managed Competition Been Tried?

Tolerably good, if not perfect, models of managed competition have been tried successfully, on a limited scale, in many employment settings: the FEHBP, CalPERS, the state employees of Minnesota and Massachusetts, California, Harvard, and Stanford Universities, and the Buyers' Health Care Action Group (BHCAG), a Minnesota employers' coalition. Also, Pacific Health Advantage (formerly the Health Insurance Plan of California or HIPC), an exchange model for small employers in California, and Benefits Alliance, a broker model aimed at the midsize employer market (www.pbgh.org, www.benefitsalliance.com). Some of these use some risk adjustment; some do not. These models usually represent large aggregations of employees and concentrations in individual markets. Carriers are more willing to bid for a piece of an...