|
MARKETWATCHHaving It All: National Benefit Equity And Local Payment Parity In Medicare
The Medicare Payment Advisory Commission (MedPAC) has identified two important problems with the Medicare+Choice (M+C) program: nationwide geographic inequity in government-financed benefits, and unequal government payments for M+C plans versus fee-for-service (FFS) Medicare in the same market area. MedPAC concludes that both problems cannot be solved simultaneously. We argue that both problems could be solved if Congress discontinued its policy of underwriting the cost of FFS Medicare. Instead, Congress should define a national entitlement benefit package and have all health plans submit bids on the package in each market area. The government’s premium contribution should be equal to the lowest bid submitted by a qualified health plan in each market area. The contribution could be adjusted for health risk, the special obligations of FFS Medicare, and welfare enhancements associated with FFS Medicare that are valued by both beneficiaries and taxpayers but unrelated to beneficiaries’ health status.
In March 2001 the Medicare Payment Advisory Commission (MedPAC) released its annual report on Medicare payment policy to Congress.1 Chapter 7 of the report, "Medicare+Choice Payments and Fee-for-Service Spending," describes a problem with the system that Congress adopted (and subsequently modified) to set capitation rates for private health plans contracting with Medicare. Prior to 1997, when these plans still were known as Tax Equity and Fiscal Responsibility Act (TEFRA)-risk health maintenance organizations (HMOs), capitation payments were set at 95 percent of fee-for-service (FFS) Medicare spending levels for "similar" beneficiaries in each county.2 The result was wide variation in HMO payments, which in turn resulted in wide variation in the availability of HMO plans and the benefits they offered. According to the MedPAC report, this outcome "seemed inequitable" (pp. 111–112). Taxpayers and beneficiaries nationwide pay into the program under the same set of rules, and under the original Medicare legislation, beneficiaries nationwide are entitled to the same benefit package. The 1997 Balanced Budget Act (BBA) and subsequent legislation addressed geographic variation in benefits by compressing the national variation in payments to private Medicare health plans, now known as Medicare+Choice (M+C) plans. That move reduced some benefit disparities, but government payments to M+C plans in many counties now differ by more than the original 5 percent "discount" from FFS Medicare. This situation is at odds with a second goal expressed by MedPAC: that "Medicare payment policy should be neutral as to whether beneficiaries enroll in traditional Medicare or in M+C plans" (p. 112). MedPAC thus identified two separate problems: nationwide geographic variation in government-financed M+C benefits, and unequal payments to M+C plans and FFS Medicare within local markets. MedPAC concludes, "No matter how payments are manipulated, both problems cannot be solved simultaneously as long as there is significant underlying variation in fee-for-service spending across market areas" (p. 111). After it released its March 2001 report, MedPAC recommended achieving local payment parity by setting M+C payments "substantially equal to local FFS costs."3 In other words, MedPAC chose single-minded pursuit of local payment parity over any degree of national benefit equity.
MedPAC’s characterization of the payment dilemma hinges on two assumptions, one stated explicitly, the other unstated. The stated assumption is that large variations in FFS Medicare spending across market areas will continue. The unstated one is that the federal government will continue to pay the full cost of FFS Medicare (minus the Part B premium). If the latter assumption is dropped, then national benefit equity and local payment parity can be accomplished simultaneously by adopting the following three-part proposal: (1) Congress determines a national entitlement benefit package (current benefits or an expanded package that includes outpatient prescription drugs); (2) private health plans and FFS Medicare submit bids on the entitlement benefits for a standardized enrollee in each market area (private plans choose their service areas; FFS Medicare serves all areas); and (3) the government premium contribution for all plans (including FFS Medicare) is set equal to the risk-adjusted bid of the lowest-price qualified health plan in each market area.4 Under our proposal, beneficiaries would have guaranteed access to all health plans in their market area through annual open enrollment. Medigap policies could be included in the bidding, in addition to M+C plans and FFS Medicare. FFS Medicare would be available in all market areas, but it would have to charge an additional premium if its bid exceeded the premium of the lowest-price plan. The special obligations of FFS Medicare, including universal availability and subsidies for medical education or facilities that treat a disproportionate share of low-income patients, would be acknowledged by explicit payments that reduce the FFS bid. If no M+C plans were available in a market area, then FFS Medicare would be the lowest-price plan, and it would be available for the Part B premium. The FFS "bid" would be the federal government’s best estimate of the cost of providing the entitlement benefit package to a standardized beneficiary in every U.S. county, a figure that was published each year until the 1997 BBA. If Congress expands the entitlement, Medicare would have to determine the price of the new benefit package, but that exercise would be required by any expansion of the entitlement, with or without bidding.
Our recommendations are designed to improve the efficiency and fairness of the Medicare program but not to change its original purpose, which was to contribute public funds to help beneficiaries purchase a national entitlement benefit package. Improving efficiency. Our proposal addresses two types of inefficiency that are caused by distorted prices in the Medicare program. First, the inability of M+C plans to give cash rebates forces them to provide supplementary benefits that are inefficient in that they are worth less to beneficiaries than they cost to produce. Second, basing decisions about entitlement benefits on FFS costs, rather than on the prices of benefits in the most efficient health plan in each market area, artificially inflates the price of entitlement benefits faced by taxpayers. As a result, taxpayers, through government, purchase an inefficiently low level of entitlement benefits. Both sources of inefficiency can be corrected simply by having beneficiaries face the marginal risk-adjusted prices of all health plans in their market area and ensuring that those prices reflect the plans’ costs. When beneficiaries pay the marginal risk-adjusted prices of the entitlement benefit in more costly health plans (either higher-price M+C plans or FFS Medicare), their choices reflect the value they attach to the competing plans. The threat of losing market share will force plans to reduce premiums to the cost of care and to offer efficient levels of amenities at prices that are close to their costs. The power of competitive bidding was demonstrated in the aborted 1997 demonstration of Medicare competitive pricing in Denver. The bids submitted by four M+C plans for the entitlement benefit package were 25–38 percent lower than the government’s payments to the plans at that time.5 Value of FFS Medicare to beneficiaries. Is unmanaged care in FFS Medicare an efficient benefit, that is, is it worth to beneficiaries what it costs to produce? Only a market test can answer this question. Thus far beneficiaries have expressed a clear preference for unmanaged care, but they do not face a cash alternative because M+C plans have not been allowed to offer cash rebates.6 Taxpayers under age sixty-five, who do face a cash alternative in the form of forgone wages, have expressed a clear preference for managed care.7 The best way to determine the value of FFS Medicare is to have beneficiaries purchase it on the margin with their own money.8 From an efficiency perspective, this test is a positive feature of our proposal. Drug coverage. If the government purchases entitlement benefits at the prices of the most efficient health plan in each market area, rather than at FFS Medicare prices, the entitlement benefit package is likely to expand. Kenneth Thorpe and Adam Atherly estimate that M+C plans can provide the current entitlement benefits for 20 percent less than FFS Medicare can.9 If that estimate also applies to additions to the entitlement, outpatient drug coverage for Medicare might be viewed in a much more favorable light. To put this another way, the additional cost to Medicare of providing an outpatient drug benefit to beneficiaries in Miami through FFS Medicare would be substantial, but the additional cost of providing it through the most efficient health plans in Miami would be zero. The government is now purchasing drug coverage for beneficiaries who enroll in the most efficient health plans in Miami. Knowledge about cost of care. Much of the inefficiency in the current M+C payment system (and in MedPAC’s local payment parity proposal) is due to the fact that the entity that knows the least about the cost of care in private health plans (Congress) announces payment rates to the entities that know the most about it (the plans). Congress lowers inflation-adjusted payments to M+C plans in high-payment areas until something bad happens. Congress raises payments in low-payment areas, hoping that something good will happen. However, there are no operational definitions of "something good" or "something bad." Would one M+C plan in a county that offers just enough additional benefits to entice a minimum number of beneficiaries to join be a good outcome or a bad outcome? There simply is no economic justification for Congress to operate Medicare as a public utility and to act as the administrative rate-setter. Public utilities are established in response to intractable market failure such as natural monopoly. Medicare is not subject to that type of market failure. A more logical approach to determine the price of the national entitlement benefit package would be to ask the health plans how much it costs. Their incentive to tell the truth in our proposal is higher out-of-pocket premiums for higher-price plans. Health plans’ flexibility. There are additional reasons why the government’s premium contribution should not be based solely on costs in FFS Medicare. M+C plans are free to adjust to market demand and supply conditions. If providers refuse to contract with M+C plans at one set of rates, M+C plans can pay them more. If providers are willing to contract at lower rates, plans will pay them less. If a given level of enrollee satisfaction or access to care cannot be met at one level of expenditure, M+C plans can spend more. Paying providers more (or less) in FFS Medicare literally requires an act of Congress. Capitated private health plans also are free to ignore inefficient practice patterns that can arise in local areas and that are perpetuated by the "professional silos" of FFS payments. Analyses by MedPAC and by Timothy McBride show that the supplementary benefits offered by M+C plans vary directly with FFS spending levels, which suggests that the cost of entitlement services is lower and less variable in M+C plans than in FFS Medicare.10 Because the higher mean and variance of cost in the FFS sector do not appear to improve enrollees’ health (see next section), the higher mean FFS cost is a wasteful basis for the government’s premium contribution, and the greater variance in FFS spending makes it appear capricious. Budgetary impact. Although the budgetary impact of any reform proposal is of interest to Congress and the administration, the goal of our reform proposal is not to reduce government spending, per se.11 The question we address is, "Can the efficiency and fairness of the program be improved?" Nevertheless, we note that under our proposal, government spending would fall immediately to the cost of the lowest-price plan in each county for all Medicare beneficiaries in the county. The savings per beneficiary would be the average difference between current government expenditures—FFS expenditures for beneficiaries in FFS Medicare, and M+C payments for M+C enrollees—and the new government premium contribution. Thorpe and Atherly’s estimate of the savings based on the current entitlement benefit package and current M+C payment levels is approximately $16 billion per year.12 Risk of unfavorable selection. An important efficiency-related concern under any payment system is that some plans may experience systematic selection by high- or low-risk enrollees over the long run. If plans receive only one "community-rated" premium for all enrollees in a market area, the plans experiencing unfavorable selection will enroll an inefficiently low number of beneficiaries. Low enrollment in the "unfavorable selection" plans is inefficient because some beneficiaries would pay their own marginal cost of joining the plan but not the average cost (premium) of all enrollees who are already in the plan.13 Also, some plans may be forced from the market, and plans may discriminate against high-risk enrollees at enrollment or disenrollment. Inefficiency and community rating. We emphasize that these outcomes result from risk selection and community rating, not competition per se. The inefficiency problems of community rating could be solved by individual experience rating, where each enrollee is charged his or her own marginal cost to join a plan. However, individual experience rating is perceived to be unfair to sicker beneficiaries, and Medicare has displayed no interest in it (although it is common among Medigap policies). We accept community rating as a given for Medicare, just as it appears to be for most firms that offer multiple health plans. The solution to the inefficiency arising from community rating is to subsidize plans or beneficiaries.14 The subsidies could be based on the current PIP-DCG system, a new system that is less sensitive to the treatment site where a diagnosis is made, or a simple premium subsidy for higher-price plans. For example, the government might set its premium contribution at the lowest bid in each market area but pay 10 percent more to the higher bidders. Many employers use a similar approach to subsidize higher-price plans.15 Simple percentage premium subsidies would be inexpensive to administer but would subsidize some inefficiency as well as differentials in risk. Improving fairness. The primary fairness problems of Medicare are geographic inequity in government-financed benefits and the overall regressivity of the program.16 Although Medicare is not the only government program that has these problems, that is no defense for perpetuating deficiencies that can be corrected without much harm. Our proposal solves the first problem by establishing a national entitlement benefit package. Our proposal also is likely to make Medicare less regressive because the demand for some types of unmanaged care appears to be positively related to income.17 Fairness and quality of care. Including FFS Medicare in a competitive pricing system raises the possibility that low-income beneficiaries might be "forced" into a low-price M+C plan, while wealthier beneficiaries continue to purchase unmanaged FFS. This concern overlooks both the findings on quality of care and the reality of current trends in Medicare. More is known about the quality of care in FFS Medicare versus M+C plans than in any other health insurance sector. In a comprehensive review, Robert Miller and Harold Luft found no systematic empirical support for the concern that managed care plans skimp on quality for Medicare beneficiaries.18 Eugene Nelson and colleagues did find that low-income, chronically ill beneficiaries fared worse in managed care plans, but Jon Christianson and colleagues reviewed all the studies of chronic care in managed care and FFS and found the results generally to be inconclusive.19 Jeanette Preston and Sheldon Retchin found that hypertensive enrollees in Medicare HMOs were more likely to receive appropriate medical histories and tests than FFS beneficiaries were.20 Dolores Clement and colleagues found that HMO patients with joint pain were more likely to have an outpatient visit and a prescribed medication but were less likely to be see specialists, have follow-up care recommended, and be monitored.21 Retchin and Barbara Brown found that HMO enrollees were "more likely to have a follow-up visit a week after discharge, and HMO providers were more likely to order serum potassium levels regularly, but no other differences were observed."22 In sum, there is no compelling evidence that quality-of-care concerns justify basing the government’s premium contribution on the cost of the FFS Medicare entitlement benefits. In addition, we note that M+C plans are monitored for enrollee satisfaction and access to care, whereas FFS Medicare is not. If further assurances of quality are needed in a competitive pricing system, the lowest-price plan in each market could be required to enroll some minimum percentage or number of enrollees before its bid is used to set the government’s premium contribution. New plans could be ineligible to be the "baseline" plan for their first year in the market. Current program trends. Concerns about low-income beneficiaries’ being forced into managed care also ignore current trends in Medicare. As more M+C plans charge premiums under the current administered pricing system and supplementary premiums continue to rise, it is likely that most low-income beneficiaries will be "forced" into FFS Medicare with no supplementary insurance, where they will face high cost sharing.23 In the past, Congress responded to that problem with the Qualified Medicare Beneficiary (QMB) and Specified Low-Income Medicare Beneficiary (SLMB) programs, which provide assistance for out-of-pocket premiums and cost sharing for low-income beneficiaries. Similar assistance could be provided under our proposal. However, again the literature offers no compelling evidence that low-income beneficiaries should avoid M+C plans, and low-income beneficiaries are now disproportionately more likely to join M+C plans. A final fairness concern is that the elderly may be unable to make informed choices. That concern must be placed in context. The elderly make choices about investments, pensions, presidential candidates, and other complex topics. While some beneficiaries may not be able to make informed choices, it would be a mistake to design an entire delivery system on that assumption, particularly as more beneficiaries enter the program from employers that offer multiple health plans. If Congress thought that the cognitive ability of the average beneficiary was a serious concern, one would expect to see far greater investment in systems to help the elderly choose among competing health plans, providers, and therapies.
Given the political power of the elderly, proposals that reduce beneficiaries’ welfare are unlikely to be politically acceptable. It is likely that current beneficiaries would have to be compensated for our proposal’s most likely adverse effect: higher out-of-pocket premiums for FFS Medicare in some markets. Medicare beneficiaries could be compensated for these premiums through an increase in the entitlement benefit package. As we suggested earlier, the 20 percent discount for benefits purchased through M+C plans rather than through FFS Medicare might be enough to encourage a major expansion of the entitlement. One implementation strategy would be to give beneficiaries a one-time opportunity to choose a combination of competitive pricing and expanded benefits, much like the onetime opportunity for purchasing outpatient drug coverage in many Medicare reform proposals. Beneficiaries who did not want the combination of competitive pricing and expanded benefits could remain in the current system. Another, perhaps complementary strategy would be to implement the reforms and expanded benefits for a new cohort of beneficiaries. The majority of newly eligible beneficiaries will be entering Medicare from some type of managed care plan, often chosen from a menu of competing plans at their place of employment, so the issue of compensation may be less relevant. Another political concern is that Congress or the administration might impose two constraints on FFS Medicare: It cannot change, and it cannot die. Under those constraints, it is unlikely that FFS Medicare could ever be included in a competitive bidding system. Political perpetuation of a plan that beneficiaries declined to purchase with their own money and that was outperformed by competing plans would validate Barbara Cooper and Bruce Vladek’s assessment of Medicare as a "provider entitlement program" and provide a strong argument for minimizing the government’s role in all health insurance markets.24
MedPAC has identified two important problems with the M+C program: geographic inequity in government-financed benefits, and unequal government payments in the same market area. MedPAC’s conclusion that both problems cannot be solved simultaneously holds only if Congress continues to underwrite the cost of FFS Medicare. Our reform proposal solves both problems by reestablishing Medicare as a defined-benefit program. In addition, it guarantees local payment equity by letting the market—not administrative calculations—determine the government’s premium contribution in each county. Absent these reforms, Medicare is approaching a crisis that will likely result in the wholesale departure of private plans. That outcome would be adverse for both beneficiaries and the program, leaving Medicare with an outdated, inflexible delivery system that is incapable of making simple efficiency improvements, unresponsive to market conditions, and not directly accountable for satisfying consumers or meeting basic access standards.
Bryan Dowd and Roger Feldman are professors in the Division of Health Services Research and Policy, University of Minnesota School of Public Health.
|
Top
How To Achieve Benefit...
Motivations For Reform
