Medicare+Choice: Payment and Service Areas. Final Report. Introduction


The Balanced Budget Act of 1997 (BBA) included provisions that dramatically affect the way that Medicare and its beneficiaries are integrated into the nation's health care system. By expanding the types of private health plans that are able to contract with Medicare, the BBA aims to move the program away from its fee-for-service roots to take advantage of the innovations in managed care service arrangements and financing that have occurred in the past decade. Through Medicare+Choice, beneficiaries in different parts of the country may be able to enroll in a variety of different private plans, presumably reducing their out-of-pocket costs through reduced cost-sharing or expanded benefits or increasing the quality or convenience of access to a full spectrum of care through an integrated plan. Although the BBA changed a number of aspects of how private plans contract with Medicare and receive capitated payments, it left in place some aspects of Medicare's earlier risk program. For example, plans are still able to define the service area covered by a Medicare+Choice contract.

As Medicare+Choice regulations were developed and scheduled for implementation throughout 1998, a large number of plans already in the risk program either decided not to renew their contract or redrew their service area. As a result, the Health Care Financing Administration(now known as Centers for Medicare and Medicaid Services(CMS)) (HCFA(now known as CMS)) reports (, Feb 1, 1999):

These withdrawals and service-area reductions affect slightly more than 400,000 beneficiaries in Medicare risk plans, representing approximately 6.5% of the more than six million risk enrollees. However, slightly more than 50,000 beneficiaries, or less than 1 percent of beneficiaries currently enrolled in Medicare risk plans, will not have access to another managed care plan.

If private plans become an increasingly important way for beneficiaries to receive health care, the causes of these types of changes are important to understand, particularly as they relate to specific policy characteristics. Whenever possible, policy design should presumably reduce, not contribute to, the volume of annual changes in the plan offerings to beneficiaries. A number of different factors, including implementation of the BBA itself, are likely to have contributed to these changes in 1998. Although some degree of churning in plan offerings in each community may be inevitable each year, it is important to understand the factors that affect plan decisions so that policymakers can interpret and respond to them.

Medicare policies about service area design are likely to affect plan behavior each year. Unfortunately, there is little information in the research literature about how plans and beneficiaries react to such policy changes. This report therefore addresses two distinct but closely related questions:

  • What are the key factors that led plans to reconfigure their Medicare risk plan service areas in 1998 or to not renew their contracts with Medicare?
  • Should Medicare continue to allow plans to design their own Medicare service areas, or should there be predetermined Medicare+Choice contract areas in each market area?

Ultimately this analysis aims to understand what factors determine whether private plans choose to contract with Medicare under Medicare+Choice. Theoretically, a plan supply equation could be developed and estimated to isolate the effect of local market factors, payment policies, and other relevant factors on plan participation. In addition to the usual challenges underlying such an exercise, this approach has two important problems. First, the historical use of plan-designated service areas complicates the notion of a market-based analysis, since plans' service areas are endogenous to the market phenomenon and policy factors to be analyzed. As a result, it is not clear how (or whether) markets could be defined that are relevant to how plans view their options and ultimately design their service areas. This problem is easily illustrated by the mismatch between plans' reported commercial service areas and their Medicare service areas - which should be taken as defining their appropriate market? Second, it is not clear how (or whether) Medicare participation fits into plans' decisions to enter particular geographic markets at all. Medicare policy and populations may have important effects on the numbers and types of plans that exist in some markets, for example in south Florida, while Medicare participation may be a more simple strategic decision among existing plans in other markets, such as the upper Midwest. If the existence of particular plans in particular markets is due, in part, to an assessment of the potential Medicare market, then a daunting intellectual problem arises. Those plans that considered entering a geographic market and decided NOT to - possibly because of the perceived opportunities (or lack thereof) in the Medicare segment of the market - are unobservable. As a result, any Medicare+Choice plan supply model that simply takes as candidate entrants those plans currently operating in local commercial markets may miss important factors and misstate the effect of studied factors. These two issues combined have foiled previous attempts to develop a credible, robust market Medicare risk plan supply model.

Before undertaking the development of such a model, or modified versions similar to Welch (1996), it is helpful to describe the characteristics of plans that participated in Medicare+Choice, the counties where such plans were available, and then how the two appear to interact. Insights into the plan and county factors that affect participation can then be used to inform analysis of the second main question, whether Medicare should continue to allow plans to describe their own service areas or should adopt predetermined contract areas. Again, before undertaking such an analysis, it is helpful to first describe several candidate geographic constructs that might be used to set such areas.

This report begins with a description of key measures and data sources. The results presented next are descriptive only, but lay the groundwork for more sophisticated analysis. In addition to remaining opportunities for more advanced analyses of these same data, the recent release of information about 2000 plan contracts creates additional opportunities for investigating these issues.