Health Affairs

This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

M E D I C A R E
M E D I C A R E  D I S A D V A N T A G E D
W E B  E X C L U S I V E   

15 December 2004  
  

Medicare Disadvantaged 
And The Search For The 
Elusive ‘Level Playing Field’

What the changes to Medicare really mean
for competition and the future of the program.

By Robert A. Berenson

ABSTRACT:

The Medicare Prescription Drug, Improvement, and Modernization Act (MMA) raised payment levels for established Medicare Advantage (private) local plans and would-be regional preferred provider organizations (PPOs). Even though plans on average receive about 108 percent of what would have been spent for the same beneficiaries in traditional Medicare, the Centers for Medicare and Medicaid Services (CMS) added another 2.3 percent in 2004 and 4.0 percent in 2005 in its implementation of risk-adjusted payments. Although MMA gives a clear preference to private plans to start a fundamental restructuring of Medicare, the question remains whether Congress will maintain overpayments to private plans when faced with the pressure to reduce budget deficits.

Title I of the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003 establishes a new prescription drug benefit in Medicare. And while the adequacy and cost of the drug benefit has been a focus of controversy, Title II—Medicare Advantage (MA)—raises the issue of whether MMA’s clear tilt toward private plans will undermine the traditional fee-for-service (FFS) program.

The Conference Report that accompanied MMA articulates that Title II “modernizes and revitalizes private plans under Medicare.”¹ Specific objectives to accomplish this overall purpose, as reflected in the report, include the following: (1) to provide predictable and sufficient payments to health plans, by relinking payments to plans to both local traditional FFS program spending and growth in that spending; (2) to increase choice for beneficiaries, particularly in rural areas, by establishing new plans, which have the attributes of preferred provider organizations (PPOs), to serve large geographic regions; and (3) to encourage greater efficiency in the Medicare program, producing savings for beneficiaries and taxpayers, initially by moving from a payment system that uses administrative pricing to one that adopts a bidding process for plans, and, in a demonstration, introducing a model to promote price competition between private plans and the traditional Medicare program.

These very broad objectives do not explicitly articulate a goal of fundamental program restructuring. Indeed, advocates of Medicare overhaul have criticized the legislation because it does not provide a formal structure for price competition among plans and traditional Medicare.² Yet even though MMA relegated premium-support proposals, under which traditional Medicare would operate as a competing plan and require beneficiaries to pay more for choosing to stay in the program in many parts of the country, to a demonstration, the legislation sowed the seeds for massive program restructuring.

Whether that restructuring will occur depends to a great extent on decisions still to be made, in some cases under the discretionary authority of the Centers for Medicare and Medicaid Services (CMS), that will determine how much health plans are paid under the new bidding approach, whether payments are appropriately risk-adjusted for the health status of enrollees, and whether there actually will be new PPOs to substitute for traditional Medicare. These decisions will largely determine the future of the program.

The MMA legislation and the CMS’s initial implementation have tilted the playing field in favor of private plans. The purpose of this paper is to examine the policy elements that produced this tilt. After reviewing certain legislative specifications, I review the topics of payment adequacy, the implementation of risk adjustment, the new bidding approach for determining plan payments, and the special role envisioned for PPOs.

Private Health Plans In Medicare

Health maintenance organizations (HMOs) and other types of private health plans have long participated in Medicare, beginning with private health plan contracts in the 1970s and the Section 1876 Medicare Risk Program that was passed in the Tax Equity and Fiscal Responsibility Act (TEFRA) amendments in 1982. In 1997 Congress passed the Balanced Budget Act (BBA), which replaced the TEFRA program with Medicare+Choice (M+C).

Risk plans under Section 1876 and M+C were paid a capitated amount set at 95 percent of the level of per capita spending in the traditional FFS program as calculated at the county level (known as the average annual per capita cost, or AAPCC). For the capitated risk payment, plans were obligated to provide statutory Part A and Part B benefits to beneficiaries who elected to receive their care in a private health plan. If there were additional funds available after administrative costs and a circumscribed level of profit was retained by plans, plans typically chose to use the excess to provide additional benefits, to attract enrollment.

As has been well documented, from 1993 to 1999 private plans thrived, with enrollment increasing from 1.8 million to nearly 7 million.³ During this period health plans were receiving large annual payment increases because of the direct linkage to inflationary spending in the traditional Medicare program. At this time, studies documented that health plans were experiencing favorable selection—that is, a healthier-than-average cohort of beneficiaries were enrolling, whereas payment was based on a population of average health, adjusted only for demographic factors, such as age and sex.⁴ Because of this, the government was actually losing money through the risk-contracting program, even with the 5 percent reduction from what it would have paid plans for the average FFS beneficiary.⁵

The BBA instituted a new method for calculating payment rates in the newly constituted M+C program, to decrease the large variations in payment to plans that result from the geographic variations in spending in the FFS program, as manifested in the AAPCC rates.⁶ Under the BBA and subsequent legislation, the plans received the highest of three calculations or prongs: flat-dollar-amount guarantees or “floors,” one for urban counties and another for rural, that provided a guaranteed minimum payment; a minimum guaranteed increase of 2 percent from the prior-year payment; and a blend of national and local rates, subject to a budget-neutrality provision to limit overall spending on M+C. The BBA also mandated that the CMS administer a system for adjusting the payments to plans based on enrollees’ health status. By 2007 all payments to plans will be fully risk-adjusted.

Because of the new payment formula, after the BBA most large, urban M+C plans typically received only the guaranteed minimum update of 2 percent, and the new M+C program immediately experienced serious problems, with plans either leaving the program or greatly cutting back on additional benefits offered to enrollees. Enrollment in all private plans dropped from 7 million in 1999 (18 percent of total beneficiaries) to 5.3 million by the end of 2003 (13 percent).⁷

Little appreciated is that most of the plan withdrawals and benefit cutbacks actually occurred between 1999 and 2001. Before MMA was passed, plan submissions to the CMS suggested that only about 41,000 enrollees would have lost their health plans in 2004, compared with the peak of 934,000 beneficiaries affected in 2001.⁸ Further, the number of beneficiaries enrolled in a zero-premium plan decreased from 80 percent in 1999 to 46 percent in 2001 but declined only to 38 percent in 2003. In 1999, 84 percent of enrollees had drug coverage; this decreased to 70 percent in 2001 but remained at 69 percent in 2003.⁹ Although some plans likely adopted a status quo position in anticipation of possible legislation that would raise payment rates.¹⁰ Still, the program was no longer hemorrhaging, even before MMA was passed.

It seems clear that many years of 2 percent payment increases to most plans was the main cause of M+C retrenchment by plans.¹¹ However, plans actually benefited from the BBA formula changes. For 1998 through 2001, the years of greatest program instability, the 2 percent guarantee and a legislated special payment for 2000 actually permitted nonfloor county health plans to receive about 8.5 percent more during the three years than they would have received under strict county-level FFS spending equivalence, the pre-BBA payment method.¹²

The problem for plans was that most health plans’ underlying health inflation trends consistently approached or exceeded double digits, which resulted in a complete turnaround from the pre-BBA days, when private plans received payment increases that were higher than their own underlying cost trends. In short, tying payments to increases in FFS program spending and not to plans’ underlying costs has repeatedly led to either plan “overpayments” or “underpayments.” MMA has largely reaffirmed this linkage.

The Medicare Advantage Program

MMA creates two types of M+C, renamed Medicare Advantage (MA), plans—regional and local—with the initiation of the regional plans scheduled to begin in January 2006, at the same time the Title I drug benefit becomes effective. A major MMA initiative is the creation of regional PPO plans, which must have a network of contracted providers but must also reimburse for all covered benefits regardless of whether such benefits are provided within the network. MA regional plans must serve all parts of a geographic region, whereas MA local plans can continue to select individual counties to serve.¹³ To provide additional support to foster regional PPOs, MMA provisions include (1) inflating benchmarks by having plan bids affect calculation of benchmark amounts, (2) risk sharing with plans for the first two years, (3) a $10 billion “stabilization fund” to encourage plan participation, and (4) a “network adequacy fund” to assist regional plans in contracting with rural hospitals.

Payments to MA plans will be based on a new administered pricing formula for 2004 and 2005. Beginning in the 2006 contract year, payment will change to a competitive bidding system, a major departure in how Medicare pays providers and plans. Under MMA bidding, the secretary of health and human services (HHS) compares the plan’s bid for Parts A and B with a benchmark amount, which is determined through the same administrative pricing formula used in 2005. If the risk-adjusted bid is lower than the risk-adjusted benchmark, a rebate of 75 percent of the difference is available to the plan to provide supplemental benefits or a reduction in the prescription drug or Part B premiums. The government retains 25 percent of bids below the benchmark. On the other hand, if the plan’s bid for Part A and B benefits is higher than the statutory benchmark for a standard beneficiary, the plan’s basic premium that a beneficiary must pay to enroll is determined by this difference. Importantly, beneficiaries who remain in the traditional program do not have to pay differentially based on plan bids, which would be the case in the premium support demonstration scheduled for 2010.

Options for beneficiaries. The MA provisions in Title II interact in important ways with the provisions setting out the new Part D prescription drug benefit in Title I. From the perspective of Medicare beneficiaries, it becomes clear that the interaction of Titles I and II will greatly expand their options and the complexity of those options.

Both regional and local MA organizations must provide at least one plan with qualifying Part D prescription drug coverage, but they may also offer plans with no drug coverage to those who choose not to enroll in Part D. In any given geographic area, a beneficiary not on Medicaid or in qualifying retiree coverage may be able to select between traditional Medicare and any of six types of MA plans, including local HMOs and PPOs, regional PPOs, private FFS plans, medical savings accounts, and local specialized plans.¹⁴ Plans can offer options that include no drug benefits, standard or basic drug benefits, or enhanced drug benefits.

The options for someone wishing to select traditional Medicare are complicated by the new restrictions on the availability of Medigap (Medicare supplemental) insurance policies. Beginning 1 January 2006, Medigap insurers may renew a policy including prescription drugs only for enrollees who decline to enroll in a Medicare Prescription Drug Plan (PDP); these insurers may not renew or issue prescription drug policies to beneficiaries who enroll in Part D. Combining all of the permutations of traditional Medicare and private plan options, there will be at least fifteen categories of possible plan offerings, presumably with a number of different companies offering various options within the categories.

At the very least, the array of products and organizations offering these diverse products will add bewildering complexity. However, private plans will have a simplicity advantage because this panoply of options tilts in favor of private health plans. A beneficiary wishing to remain in traditional Medicare will have to purchase two additional supplemental policies—a PDP and a non-drug Medigap policy—to obtain comprehensive coverage, whereas a private plan will be offering “one-stop shopping” for comprehensive benefits.

On the other hand, beneficiaries previously were attracted to M+C plans offering prescription drug benefits—often more generous and lower cost than those available in the Medigap market.¹⁵ With even a relatively skimpy Part D benefit available starting in 2006 in traditional Medicare, this particular appeal of private plans might diminish, especially given the legacy of program withdrawals in recent years. Neither the CMS Office of the Actuary nor the Congressional Budget Office (CBO) has estimated sizable increases in enrollment in local MA plans. Their analyses have modeled whether plans would be able financially to offer extra benefits to attract beneficiaries, not the possibility that beneficiaries would seek simplicity in one-stop shopping.

Plan payments exceed FFS spending for 2004 and 2005. MMA altered the current M+C payment formula in several ways, effective 1 March 2004. A fourth calculation, or prong, entered the payment determination. The fourth prong assures that plans are paid no less than 100 percent of the county’s per capita FFS program spending.¹⁶ Importantly, the minimum update is now the greater of either 2 percent or the projected growth in total Medicare spending (for 2004, 6.3 percent for aged beneficiaries).¹⁷ The new calculation for the annual minimum update, built on a base that exceeds traditional Medicare spending, will ensure persistence of excess spending for private plans, or, more precisely, persistence of inflated benchmarks against which plans will bid.

The net effect of these various provisions has been a marked increase in plan payments.¹⁸ Various analyses have found that risk plans in 2004 have received on average an increase of 10.6–10.9 percent over 2003 rates, representing about a 7.4 percent point rise over the increase originally announced in 2003 for 2004.¹⁹ Three analyses found that on average, for 2004, MA plans will be paid 107–109 percent of what it would have cost to care for these beneficiaries had they remained in FFS Medicare, before consideration of favorable risk selection into health plans.²⁰ The excess payments approximate $2.75 billion on a base of spending of about $35 billion per year.²¹ Similarly, estimates find that in 2005, plans will receive nearly 108 percent of traditional Medicare payment levels, again before consideration of favorable selection.²²

The search for a level playing field. In recent years the Medicare Payment Advisory Commission (MedPAC), which advises Congress on Medicare policy, has taken a strong position that Medicare payment policy should be neutral as to whether beneficiaries enroll in traditional Medicare or in M+C (now MA) plans. Because health care is delivered in local markets, MedPAC believes that payment-neutrality needs to be pursued at the local level, which for practical purposes means at the county level. Divergence from payment equivalence, it argues, would give one sector—either MA or traditional FFS—an unfair advantage, essentially tilting the playing field and distorting beneficiaries’ choices.

Commenting on MMA, MedPAC now finds that the law tilted the playing field in favor of private plans because plan payments exceed local per capita FFS program spending in all counties—in many counties, by sizable amounts.²³ Nevertheless, it agrees with MMA that per capita, traditional program spending at the county level should remain the core determinant of administrative payments to plans and, after 2005, the benchmark amount against which MA plans should bid. However, studies have documented that county-level health plan per capita costs correlate weakly with per capita FFS program costs. For example, the CBO recently estimated that plans’ actual costs range from at least 40 percent below per capita FFS county spending to 40 percent above it.²⁴

Although there is a lack of empirical data documenting the reasons for the discrepancy between plan costs and traditional Medicare spending, plausible explanations abound. For example, plans in high-cost Medicare areas confronted with excess supply of hospital beds and physicians and high use patterns for discretionary services are well positioned to use contracting leverage to obtain price discounts and to apply utilization management tools for reducing excess use. Conversely, plans in low-cost Medicare areas may lack contracting leverage and thereby have to pay network providers higher prices than Medicare pays.²⁵

Thus, even MedPAC’s version of a level playing field would not be level, at least if underlying plan costs are considered relevant. In essence, FFS Medicare pays too much in many counties because of demonstrably nonuseful spending and then compounds the problem by paying private plans that much and more to private plans serving those counties.²⁶ These extra payments produce sizable extra benefits for beneficiaries who happen to live in these high-spending areas. But it is hard to see how paying plans based on inflationary FFS program spending can reduce spending. As discussed below, bidding theoretically could alter plan overpayments, but there are reasons to doubt whether it will have the desired effect.

Risk Adjustment

Few health policy initiatives are as broadly recommended and as little understood as risk-adjusting payments to health plans. Congress has largely deferred to the CMS the implementation of risk adjustment.²⁷ Although risk-adjusting payments involves handling reams of plan-reported data, which suggests technical complexity, the CMS must in fact make many operational policy decisions that determine risk adjustment’s financial impact on plans, the Medicare trust funds, and, ultimately, beneficiaries. For example, in calculating payments to health plans, the CMS has to decide what cohort of plans have submitted acceptable data on which to base risk scores; whether to adjust payment rates for trends in coding practices in traditional Medicare (so-called FFS normalization); and whether to use lagged or nonlagged data to calculate risk scores. To some extent, these operational policy decisions turn up in the rate calculations but have not been explained in the annual forty-five-day advance notice of rates and methodological changes or elsewhere.

In brief, data submitted by health plans for January–December 2003 showed that health plans had an aggregate risk score of 0.92 percent. That is, health plans in the most recent reporting period served beneficiaries for whom predicted costs were 8 percent lower than the costs for the average FFS beneficiary, demonstrating that these plans continue to experience favorable selection.²⁸

However, in 2002 the CMS decided not to take the small savings from the phased-in implementation of risk adjustment in 2003 and has now decided to extend that policy for at least 2004 and 2005, when program savings from risk adjustment would have been more substantial.²⁹ Rather, the CMS has applied what it calls a “budget neutrality” adjustment, using the term in this case as a policy that assures that risk adjustment does not reduce the aggregate amount of payments to MA organizations, but rather simply adjusts within the MA plan sector for relative health risks. In other words, the CMS has canceled for 2003–2005 any savings from paying plans more accurately for their enrolled populations.

Operationally, the adjustment is the estimate of the difference between aggregate MA payments that would be made using only demographic factors to adjust payments compared with the aggregate payments that would be made using health status or “risk” factors to adjust payments. Under the CMS’s policy, each health plan would still have its payments adjusted for the relative risk of its own enrollees, but all plans also would be provided the same budget-neutrality adjustment as an add-on to their payments.

The decision to apply this adjustment was originally made well before the large MMA payment increases were enacted, but the policy was reaffirmed even after MMA provided plans with major payment increases for 2004 and 2005. The CMS determined that the difference in payments between the demographic- and health risk–based payments was about 8 percent for 2004, and it used that difference to calculate the add-on given to plans.³⁰ Since risk adjustment was to apply to 30 percent of payments, the forgone savings represented about 2.3 percent (or $800 million). In 2005, when risk adjustment will apply to 50 percent of plan payments, a similar 8 percent giveback will result in additional payments of 4 percent (or $1.5 billion) to plans.³¹

The CMS’s implementation of risk adjustment, then, should more appropriately pay plans that try to attract beneficiaries with chronic conditions. That helps create a level playing field within the MA program. On the other hand, the budget-neutrality policy used now for three years undermines a main purpose for adopting risk adjustment in the first place: to reduce the overpayments that result from favorable selection into MA plans. The Bush administration’s most recent budget baseline assumes that there will be savings from risk adjustment, beginning in 2006.³² Such a projection assumes the end of the budget-neutrality policy, a step that, after three years in use, would surely face political opposition from plans and, possibly, legal challenge.

Plan Bidding

In the late 1990s the CMS and then Congress tried to test a bidding model for health plan payments by scheduling demonstrations of “competitive pricing” in Baltimore, Denver, Kansas City, and Phoenix. These demonstrations were killed by politicians from the various communities.³³ The bidding model in MMA adopts a new method of payment for the entire MA program that has not been tested.

Bidding among competitors theoretically can deliver to beneficiaries more generous benefits at lower premiums, with no additional cost to the federal government.³⁴ Thus, competitive bidding could produce plan bids that are closer to the cost of actually providing the defined-benefit package than are payments based on spending in the traditional Medicare program. If it works as theory suggests, it would address the current problem of paying plans in high-FFS-spending counties far more than their costs.

By the design of Title II bidding, beneficiaries will face differential premiums in choosing among the various private plans and traditional Medicare, although beneficiaries choosing to remain in traditional Medicare will be protected—that is, beneficiaries can benefit by moving to a plan that bids low but does not directly pay more for staying with traditional Medicare. If market competition among private plans produces greater efficiencies, it could lower government payments because the government retains 25 percent of the difference of health plan bids below the formula-driven benchmarks.

However, certain design features in the bidding approach adopted in MMA raise doubts about the potential for low bids and government savings. In the various aborted competitive-pricing demonstrations, the benchmark that set the government contribution was designed to be a function of the bids themselves—for example, the median bid or the enrollee-weighted mean bid, rather than an external, predetermined benchmark.

Bidding against a fixed benchmark creates different dynamics than bidding against an unknown benchmark. Indeed, in their opposition to the competitive-pricing demonstrations, health plans expressed concern that the added uncertainty provided by lack of an external benchmark might lead them to make lower bids than they would otherwise make.³⁵ Thus, in the MMA model, with all plans bidding against a known benchmark, bids may cluster closer to the external benchmark than they would in the competitive-pricing model.

Also, in contrast to the competitive-pricing demonstration’s bidding model, government will share in bids below the benchmark. In effect, the government’s 25 percent take of low bids may serves as a “tax” on low bids and could reduce the likelihood of aggressively low bids.³⁶

The actual dynamics of bidding will depend on how the CMS implements the bidding process. Plans’ bids will be subjected to an actuarial review by the CMS. Additionally, in a provision modeled after the Federal Employees Health Benefits Program (FEHBP), the CMS may negotiate with plans based upon that review, and in its recently promulgated proposed rules implementing MMA, the CMS has suggested that it plans to negotiate aggressively.³⁷

Whether this process will constrain the likely tendency for plans to move their bids toward the government benchmark remains to be seen. Nevertheless, although bidding is new and potentially transforming, the reality that the benchmark remains a function of a complex administrative pricing formula appears to assure that there will remain a gap between plan costs and Medicare payments in many areas.

Preferred Provider Organizations

The set of special provisions designed to promote large regional or possibly national PPOs as a new choice for Medicare beneficiaries attempts to emulate the range of choices that some large employers provide in commercial insurance markets. The special payments and other protections are designed to confront the reality that plans that cannot pick and choose the areas they wish to serve but, rather, must offer community-rated products over relatively large geographic areas may start out with a disadvantage compared to traditional Medicare and local MA plans, including local PPOs.

The CMS Office of the Actuary and the CBO disagreed fundamentally on the likely success of regional PPOs in the Medicare market and, consequently, on the additional costs associated with promoting them.³⁸ The CBO estimated that there would be little PPO enrollment, whereas the Office of the Actuary estimated that PPO enrollment would reach 16 percent of Medicare beneficiaries by 2009 and later.³⁹ Basically, the two organizations used different analytic approaches to project the costs that regional PPOs would experience in Medicare.⁴⁰

A dominant argument promoting broad access to PPOs was that all beneficiaries, not just those where local MA plans operate, should have private plan choices. Further, many have argued that beneficiaries want to have the same health plan choices in Medicare that some of them experienced with their employment-based health benefits. However, it is not clear that beneficiaries value PPO options in Medicare. Recently the U.S. Government Accountability Office (GAO) found that less than 1 percent of ten million eligible beneficiaries enrolled in the PPO demonstration that the CMS actively encouraged.⁴¹

PPOs’ main function in commercial markets is the aggregation of diffused purchasing power to avoid having to pay providers their actual charges. Indeed, PPOs have been labeled “discount management firms.”⁴² Further, although PPOs may engage in limited utilization management and quality improvement activities in commercial markets, they do not influence care delivery much.⁴³ For example, in the PPO study that formed the basis for their cost estimates, the Office of the Actuary assumed little savings from PPO medical management activities.⁴⁴ In fact, much of the current appeal of PPOs, compared with HMOs, among employers is based on the PPO product design’s greater ability to accommodate transfer of costs from employer-sponsors to covered individuals in the form of higher deductibles and copayments. Yet this feature of benefit design flexibility is not one that Medicare would seek to take advantage of.

In short, employers are turning to the PPO model for reasons that do not apply to Medicare. The traditional Medicare program has enough market power to impose administrative prices on providers at rates that are generally lower than those of commercial PPOs. Medicare beneficiaries already enjoy broader freedom of choice, with limits on balance-billing, than in most PPOs. In other words, the main virtues of the PPO model in commercial markets are not applicable to Medicare, which itself functions in many ways like a PPO.

Former CMS administrator Tom Scully argued that one main purpose for creating an extensive network of PPOs in Medicare would be to decrease the market power of the traditional program yet to replace it with nongovernmental insurers, which themselves might have sizable market power. In his view, private monopsony payers would be unencumbered by the political interests and regulatory requirements that arguably restrict the flexibility of the traditional Medicare program to act decisively to reduce spending and to respond to market-specific factors related to quality and access.⁴⁵

Concluding Comments

Although Medicare restructuring through premium support has now been relegated to a demonstration, which may not happen as senators scramble to make sure that their states are exempt, it seems clear that many still envision a formally structured competition between private health plans of various types and the traditional Medicare program.⁴⁶ One way, then, to understand the overpayments for local MA plans and the special provisions for establishing and sustaining regional PPOs is for the purpose of building an alternative infrastructure for head-to-head competition with, and ultimately replacement for, traditional Medicare.⁴⁷ In addition, plan overpayments continue to provide a way for Congress to provide additional benefits to some beneficiaries without actually having to expand benefit entitlements legislatively.

However, in the face of evident plan overpayments resulting both from MMA changes to the administered pricing formula and the CMS’s discretionary decision to not take risk adjustment savings, Congress, facing growing budget deficits, next year will have to decide whether to reduce these sizable overpayments. Without the extra payments, the current “wait and see” attitude of health plans, which showed reluctance in 2004 to jump back into Medicare, will have proved prescient.

Work on this paper was supported by the Commonwealth Fund. The author thanks a number of people for helpful comments on earlier drafts of the paper, including Shawn Bishop, Tom Bradley, Barbara Cooper, Steve Zuckerman, and two anonymous reviewers. The author also thanks the staff at the CMS for their willingness to discuss implementation issues.

NOTES

1. U.S. House of Representatives, Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (To accompany H.R. 1), Rpt. 108-391 (Washington: U.S. Government Printing Office, 2003).
2. S.M. Butler and R.E. Moffitt, “Time to Rethink the Disastrous Medicare Legislation,“ Web Memo no. 370, 17 November 2003, www.heritage.org/Research/HealthCare/wm370.cfm (13 October 2004). Because the term “premium support” had become a lightning rod for criticism, the would-be 2010 demonstration testing the concept was renamed “comparative cost adjustment.”
3. Centers for Medicare and Medicaid Services, “Medicare Managed Care Contract (MMCC) Plans, Monthly Summary Report,” 2 November 2004, cms.hhs.gov/healthplans/statistics/mmcc (5 November 2004). Not all beneficiaries enrolled in Medicare+Choice (M+C) receive their services through coordinated care plans (CCPs) (for example, HMOs, PPOs, or provider-sponsored organizations). Approximately 10 percent receive their services through cost plans, health care prepayment plans, demonstrations, and private FFS plans. Enrollment figures for M+C typically have included enrollment in the full array of private plan options, and not just for CCPs bearing risk.
4. U.S. Government Accountability Office, Medicare+Choice: Payments Exceed Cost of Benefits in Fee-for-Service, Adding Billions to Spending, Pub. no. GAO/HEHS-00-161 (Washington: GAO, August 2000); R.S. Brown et al., The Medicare Risk Program for HMOs—Final Summary Report on Findings from the Evaluation, Contract no. 500-88-0006, prepared for the CMS (Princeton, N.J.: Mathematica Policy Research, 1993); and G. Riley et al., “Health Status of Medicare Enrollees in HMOs and Fee-for-Service in 1994,” Health Care Financing Review 17, no. 4 (1996): 65–75.
5. Until the BBA, payments were set at a straightforward 95 percent of per capita spending at the county level. However, a mistake in drafting the BBA prevented the CMS from readjusting the 1997 payment base for prior incorrect estimates. As a result, the BBA formula actually used as a base for M+C plan payments about 98 percent of FFS, rather than the intended 95 percent. Office of Inspector General, U.S. Department of Health and Human Services, Adequacy of Medicare’s Managed Care Payments after the Balanced Budget Act, Pub. no. A-14-00-00212 (Washington: DHHS, September 2000).
6. Physician Payment Review Commission, Annual Report to Congress 1997 (Washington: PPRC, 1997).
7. CMS, “Medicare Managed Care Contract (MMCC) Plans.”
8. Medicare Payment Advisory Commission, Report to Congress: Medicare Payment Policy (Washington: MedPAC, March 2001); and M. Gold, “Can Managed Care and Competition Control Medicare Costs?” Health Affairs, 2 April 2003, content.healthaffairs.org/cgi/content/abstract/hlthaff.w3.176 (21 June 2004).
9. M. Gold and L. Achman, “Shifting Medicare Choices, 1999–2003,” Fast Facts no. 8 (Washington: Mathematica Policy Research, December 2003).
10. W. Black and M. Gold, “MMA Attempts to Breath Life into Troubled Markets, 2004,” Monitoring Medicare+Choice: Operational Insights no. 13 (Princeton, N.J.: Mathematica Policy Research, July 2004).
11. Other factors have contributed to M+C plan reductions, including the competitiveness of markets; that is, more-competitive markets were associated with an increase in benefits and a reduction in premiums. S.D. Pizer and A.B. Frakt, “Payment Policy and Competition in the Medicare+Choice Program,” Health Care Financing Review 24, no. 1 (2002): 83–94. Factors related to plan size, for-profit status, and other business considerations also influenced plans’ decisions regarding Medicare participation. T. Lake and R. Brown, “Medicare+Choice Withdrawals: Understanding Key Factors” (Washington: Mathematica Policy Research for the Henry J. Kaiser Family Foundation, July 2002).
12. Centers for Medicare and Medicaid Services, Office of the Actuary, “Comparison of Yearly Payment Rate Growth in Floor Counties, Non-Floor Counties, All Counties, and FFS” (Baltimore: CMS Office of the Actuary, 3 December 2003).
13. By January 2005, HHS is required to identify the regions that regional MA plans must serve, a number of at least ten and no more than fifty, covering the fifty states and the District of Columbia.
14. “Specialized MA plans” would be allowed to exclusively enroll special-needs beneficiaries in MA plans that have targeted clinical programs for them.
15. K.E. Thorpe and A. Atherly, “Medicare+Choice: Current Role and Near-Term Prospects,” Health Affairs, 17 July 2002, content.healthaffairs.org/cgi/content/abstract/hlthaff.w2.242 (5 November 2004).
16. The county-level spending data will again include an estimate for indirect medical education (IME), which means that the payment rate in counties subject to the FFS equivalence prong will actually average 102.3 percent of traditional FFS program spending in 2004.
17. The MMA Conference Report inaccurately states that payment increases are tied to increases in the traditional FFS program. In fact, the statutory definition makes clear that this growth reflects total spending increases in MA and traditional Medicare.
18. CMS, “Note to: Medicare Advantage Organizations and Other Interested Parties; Revised Medicare Advantage Payment Rates for Calendar Year (CY) 2004,” 16 January 2004, cms.hhs.gov/healthplans/rates/2004ma/cover.asp (11 June 2004).
19. The rates for March–December 2004 were inflated modestly to make up for the inability to raise the January and February rates so close to the December signing of the law.
20. L. Achman and M. Gold, “Medicare Advantage 2004 Payment Increases Resulting from the Medicare Modernization Act,” 17 February 2004, www.mathematica-mpr.com/publications/PDFs/MedAvpayrates.pdf (11 June 2004); B. Biles, L.H. Nicholas, and B.S. Cooper, “The Cost of Privatization: Extra Payments to Medicare Advantage Plans,” Issue Brief (New York: Commonwealth Fund, May 2004); and MedPAC, Report to Congress: Medicare Payment Policy (Washington: MedPAC, March 2004).
21. Biles et al., “The Cost of Privatization.”
22. B. Biles, L.H. Nicholas, and B.S. Cooper, “The Cost of Privatization: Extra Payments to Medicare Advantage Plans—2005 Update” (New York: Commonwealth Fund, December 2004).
23. MedPAC, Report to Congress: Medicare Payment Policy (March 2004).
24. Congressional Budget Office, “Cost Estimate, H.R. 1 and S. 1 (Washington: CBO, 22 July 2003), 38, footnote 8. See also B. Dowd, R. Coulam, and R. Feldman, “A Tale of Four Cities: Medicare Reform and Competitive Pricing,” Health Affairs 19, no. 5 (2000): 9–27; MedPAC, Report to Congress: Improving Risk Adjustment in Medicare (Washington:MedPAC, November 2000); and CMS, “Preferred Provider Organizations (PPOs): A Model for Medicare that Controls Cost and Enhances Quality” (Baltimore: CMS, 30 May 2003).
25. R. Berenson,”Medicare+Choice: Doubling or Disappearing?” Health Affairs, 28 November 2001, content.healthaffairs.org/cgi/content/abstract/hlthaff.w1.65 (5 November 2004).
26. E.S. Fisher et al., “The Implications of Regional Variations in Medicare Spending, Parts 1 and 2,” Annals of Internal Medicine 138, no. 4 (2003): 273–298.
27. Beginning in 2000, the CMS started phasing in risk-adjusted payments, initially based on diagnoses from inpatient hospital stays using the principal inpatient diagnostic cost group (PIP-DCG) model of risk adjustment and, beginning in 2004, diagnoses as well from physicians’ offices and hospital outpatient departments, using the CMS-Hierarchical Condition Category model of risk adjustment. CMS, “Note to: Medicare Advantage Organizations and Other Interested Parties; Advance Notice of Methodological Changes for Calendar Year (CY) 2004 Medicare+Choice (M+C) Payment Rates,” 28 March 2003, cms.hhs.gov/healthplans/rates/2004/45day.pdf (11 June 2004).
28. Initial data from about half of the plans from July 2001 to June 2002 showed plans with an average risk score of 0.84, which suggests 16 percent favorable selection. However, with broader and improved plan reporting for July 2002 to June 2003, the overall risk score had increased to about 0.88, or 12 percent favorable selection. Moving to a “nonlagged” reporting period—that is, using January–December 2003 for payment in 2004—caused a further reduction of favorable selection to 0.92, or 8 percent favorable selection. Cynthia Tudor, CMS, personal correspondence, 29 September 2004.
29. The 2002 decision to give plans back 0.65 percent for 2003, thereby forgoing savings from the 10 percent phase-in of risk adjustment in 2003 payments, was never posted on the CMS Web site but rather was announced at a public meeting of health plans.
30. The CMS made another operational decision labeled “FFS normalization” to calibrate the risk-adjustment model to eliminate distortions from changes in coding practices over time. CMS, “Note to: Medicare+ Choice Organizations and Other Interested Parties; Announcement of Calendar Year (CY) 2004 Medicare+Choice Payment Rates,” 12 May 2003, cms.hhs.gov/healthplans/rates/2004/cover.asp (11 June 2004).
31. $3.32 billion was spent on MA plans in June 2004. However, CCP plans make up 87 percent of total MA enrollment. Thus, estimated 2004 spending affected by risk adjustment is $34.7 billion, assuming no enrollment change. For 2005, MA plans are projected to receive an overall 7.1 percent increase in Medicare payments, to $37.2 billion.
32. “Fiscal Year 2005, Mid-Session Review, Budget of the U.S. Government,” 30 July 2004, www.whitehouse.gov/omb/budget/fy2005/05msr.pdf (18 October 2004).
33. Dowd et al., “A Tale of Four Cities”; and K. Ignagni, “Putting Principles First: A Better Way to Carry Out a Demonstration,” Health Affairs 19, no. 5 (2000): 44–48.
34. Pizer and Frakt, “Payment Policy and Competition.”
35. Dowd et al., “A Tale of Four Cities.”
36. T. Feldman et al., Premium Rebates and the Quiet Consensus on Market Reform for Medicare, Report to the Centers for Medicare and Medicaid Services (Cambridge, Mass.: Abt Associates Inc., 30 August 2001). Beginning in 2002, M+C plans were permitted to provide premium rebates to enrollees of up to 125 percent of the Part B premium, with 80 percent of any premium reduction distributed to the enrollee and the rest to the government. As of February 2004, only six plans offered rebates, affecting 4 percent of MA enrollees.
37. 42 CFR, Parts 417 and 422; Medicare Program; Establishment of the Medicare Advantage Program, Proposed Rule, CMS-4069-P (RIN 0938-AN06).
38. The CMS Office of the Actuary assumed a small increase (to 13 percent) in the percentage of beneficiaries served by local MA plans. The CBO initially projected a continuing downward trend to only 8 percent. However, in the 2004 budget baseline, the CBO revised its estimate of MA plans to slightly over 13 percent also. Douglas Holtz-Eakin, director, Congressional Budget Office, Testimony before the House Committee on Ways and Means, 24 March 2004; and Congressional Budget Office, “CBO March 2004 Baseline: Medicare,” 3 March 2004, www.cbo.gov/factsheets/2004b/Medicare.pdf (11 June 2004).
39. Richard Foster, chief actuary, Centers for Medicare and Medicaid Services, Testimony before the House Committee on Ways and Means, “The Financial Outlook for Medicare under the Medicare Prescription Drug, Improvement, and Modernization Act of 2003,” 24 March 2004.
40. CMS, “Preferred Provider Organizations (PPOs)”; and Holtz-Eakin testimony.
41. GAO, Medicare Demonstration PPOs: Financial and Other Advantages for Plans, Few Advantages for Beneficiaries, Pub. no. GAO-04-960 (Washington: GAO, September 2004).
42. R.E. Hurley et al., “The Puzzling Popularity of the PPO,” Health Affairs 23, no. 2 (2004): 56–68.
43. Ibid. There are exceptional PPOs that do provide HMO-like interventions, such as disease management, for which some employers pay an additional fee. See R.E. Hurley et al., “Preferred Provider Organizations and Medicare: Is There an Advantage?” Issue Brief no. 81 (Washington: Center for Studying Health System Change, April 2004).
44. CMS, “Preferred Provider Organizations (PPOs).”
45. Tom Scully, administrator, CMS, Testimony before the Senate Finance Committee, “Strengthening and Improving the Medicare Program,” 6 June 2003; and “Why Three PPOs? Scully Speaks,” Medicine and Health 57, no. 26 (2003): 4–5.
46. A. Fagan, “Medicare Competition Pilot Program Hit,” Washington Times, 8 June 2004.
47. Rep. Bill Thomas, chair of the House Ways and Means Committee, recently suggested that the increase in payments to private plans was not attributable to congressional preference for the plans, but rather was intended to help them improve their infrastructures of information technology, disease and care management capabilities, wellness programs, and other innovations, to help them prepare for competition with traditional Medicare. “GOP: Consumerism Must Work or Single Payer’s on the Way,” Medicine and Health 58, no. 11 (2004): 2–3.


Bob Berenson (rberenson{at}ui.urban.org) is a senior fellow at the Urban Institute in Washington, D.C. He is a former staff member of the Centers for Medicare and Medicaid Services, the federal agency responsible for Medicare.

DOI: 10.1377/hlthaff.w4.572
©2004 Project HOPE–The People-to-People Health Foundation, Inc.