Marilyn Moon, Len Nichols, Korbin Liu, Genevieve M. Kenney, Margaret B. Sulvetta, Stephen Zuckerman, Crystal Kintz
Medicare is now facing enormous pressures to reduce the rate of growth of spending, both because of the goal of the 104th Congress to seek to balance the federal budget within the next seven years and because of the financial imbalance in Medicare's Part A trust fund. Like the rest of our health care system, Medicare spending has grown rapidly over the last three decades. Between 1996 and 2002, the year that some proponents of a balanced budget seek to reduce the annual deficit to zero, Medicare's spending is projected to total $1.85 trillion dollars. It is not surprising then that Medicare would be a major part of the changes sought in the Congressional Budget Resolution for fiscal year 1996. The goal of that effort is set at $270 billion in savings from outlays for Medicare benefits. Improved efficiencies, moderate limits on payments to providers and small increases in beneficiary contributions could save considerable amounts of money, but reaching $270 billion will require formidable changes in the program.
This paper analyzes trends in Medicare expenditures and discusses options to modify the program in order to control its spending growth. Options included here are evaluated in terms of their likely impact on the integrity of the Medicare program. This paper represents an effort to put in context the types of changes being discussed in the ongoing budget debate and to raise cautions, where appropriate, about various strategies. In a program as large and complex as Medicare, there are many issues that need to be assessed before putting together a package of changes.
We examine five of the major service areas under Medicare which account for 88 percent of total spending. In addition, we look separately at prospects for increased beneficiary contributions, followed by an examination of savings that could be obtained from allowing beneficiaries to choose private insurance plans. We conclude with a discussion of some of the principles that should be considered in balancing the necessary pieces of a large package of Medicare reforms.
The Challenge of Finding Immediate Savings
Reaching the Medicare savings targets in the Budget Resolution will require reducing per capita program growth to 4.9 percent a year. This growth rate is only slightly higher than the lowest annual growth rate ever achieved in the private health care sector over the last 20 years (Moon and Zuckerman 1995). The Congressional Budget Office projects that Medicare spending under current policy will grow by $20.9 billion between fiscal years 1995 and 1996, representing a per capita growth rate of 9.9 percent. The 104th Congress's call for a growth reduction of $7 billion over the next year implies a per capita rate of 6.6 percent—an immediate growth rate reduction of over one-third.
But many of the most promising areas for increased savings—such as developing new payment methodologies for home health or skilled nursing facility care, or new ways to oversee service use—will take years of development and implementation before significant savings can be expected.
Savings from Providers of Care
Medicare's traditional way of achieving savings has been to reduce payments to providers of care and to change the incentives in the payment formulae in a way that alters the pattern of service use. Reforms implemented in the 1980s in the way Medicare paid hospitals, for example, led hospitals to reduce the length of hospital stays and improve their oversight of service use by patients while in the hospital.
Further payment reductions and still stronger incentives for controlling service use can be expected to achieve additional savings in the future. But limits on the extent of future savings should be recognized. Medicare payments are already considerably below private sector rates (ProPAC 1995; Zuckerman and Verrilli 1995). Providers, in particular hospitals, will continue to be under pressure to give private sector discounts at the same time as they may be facing significant reductions in Medicaid revenues. At some point the financial viability of a whole set of hospitals will be threatened, reducing their willingness to serve Medicare patients and, in consequence, threatening the access of all patients to mainstream medical care.
The truly innovative strategies for controlling costs that are found in the private sector—in particular, payment systems that modify the open-ended fee-for-service structure—show great promise. But they will take time to develop and implement in the Medicare context, and in the meantime are likely to raise rather than lower the need for administrative oversight. Skimping on administrative costs while introducing major policy changes is likely to be penny-wise but pound foolish, running the risk of increasing both waste and abuse—as happened in the case of the Savings and Loan crisis some years ago.
To understand the potential for savings in the area of provider payment and organization we need to look at each component, because each has idiosyncracies and special challenges that need to be considered in making realistic savings estimates. Some areas are easier targets than others for short-term savings because structural changes have already been made. Applying a set percentage of savings to each area and then looking for policies that would achieve it does not make sense. The five areas of spending we highlight account together for 88 percent of total Medicare spending; and specific options for change are listed in Summary Table 1.
This is Summary Table 1.
Inpatient Hospital Services. Six Specific policy changes in inpatient hospital services are examined in detail. The largest savings would result from the gradual elimination of disproportionate share payments (except for 160 hospitals with extremely high need). These are special payments made to hospitals that serve more than the average proportion of Medicare and Medicaid patients. Reduced indirect and direct graduate medical education payments and a one-time, one-year freeze on the inflation-adjustment in the hospital payments formula would also yield substantial savings.
Physician Payments. A second major source of potential savings among the provider options would be in physician payments. These could be reduced by linking physician fee growth to overall economic growth rather than to historical service volume and intensity trends, and by a one-time 10 percent cut in the physician fee schedule factor that converts the relative values of different physician services to a dollar amount.
Home Health Services. Home health services is the fastest growing component of Medicare spending. The options we consider could save $14.7 billion over seven years, including reductions in the maximum payment to a home health agency from 112 percent to 100 percent of the national mean. Movement toward a prospective payment system with substantial payment cuts would also produce substantial savings, but would take considerable time to implement.
We also discuss moving coverage for home health care from Part A to Part B. Part A was originally intended to cover only hospital inpatient services providing a historical precedent for a shift. It may also be reasonable from a policy perspective, since the cost of home health services would be shared by all beneficiaries through increases in the Part B premium. As discussed further below, this shift would raise the net Part B obligations by necessitating a Part B premium increase.
Skilled Nursing Home Facilities. Reducing maximum payments to skilled nursing facilities from 112 percent to 100 percent of the national mean would produce savings of $3.6 billion over seven years. Development of a prospective payment system would produce further savings in the longer run.
Outpatient Hospital Payments. Substantial changes in outpatient hospital payments with a comprehensive restructuring of the payment system could produce substantial savings. The net impact would be lower, however, if savings from restructuring outpatient hospital payments would go to reducing the high cost sharing now imposed on beneficiaries.
Increased Beneficiary Contributions
A second major way to obtain savings under Medicare would be to require beneficiaries to contribute more toward the costs of their own care through a combination of higher cost sharing and higher Part B premiums. One major argument made for raising beneficiary contributions is that older Americans are becoming increasingly better off and thus able to afford to pay more of the costs of the Medicare program themselves. However, among the noninstitutionalized elderly, over one-fourth (8.2 million people) still have incomes below 150 percent of the poverty level.[1]. Thus, large numbers of moderate to low-income beneficiaries could find even modest changes in Medicare premiums and cost sharing burdensome.[2].
Affordability is not the only policy issue, however. Higher beneficiary contributions are advocated in part because of the general notion that higher cost sharing encourages cost consciousness and more prudent use of services. This may not apply to the elderly, however, because many have supplemental protection from these costs. About 12 percent of beneficiaries have gaps in their Medicare coverage eased by the Medicaid program through either standard eligibility or the Qualified Medicare Beneficiary (QMB) program started in 1989 (Chulis et al. 1993).[3]. Another 42 percent—often those with high incomes—get supplemental coverage through their former employers. The remaining elderly beneficiaries are likely to have higher burdens. About one-third of elderly beneficiaries have "medigap supplemental policies," but since they have to buy such policies on their own, this adds to their total health care expenses. Only about 11 percent have no supplemental coverage at all.
While a broad range of beneficiary contributions is possible and many could be justified individually, it is crucial to combine the net amounts together to assess what constitutes a reasonable set of changes. Increases in cost sharing fall disproportionately on the sick and only affect those who remain in fee-for-service medicine. In contrast, premium increases can be shared more evenly across the beneficiary population and apply to those in HMOs as well as those in fee-for-service. Various cost-sharing expansions and premiums increases are shown in Summary Table 2.
This is Summary Table 2.
Cost Sharing Changes. The cost-sharing changes involve 20 percent coinsurance for each day in a skilled nursing facility, $5 a home health visit, and an increase in the Part B deductible to $150 with an inflation adjustment. Together these could save a projected $26 billion.
Premium Increases for Part B. One way to obtain further savings would be to increase the beneficiaries' share of the Part B premium and raise the share still more for higher income beneficiaries. (All Part B premiums would be 13.1 percent higher if home health services were shifted to Part B from Part A.) Setting the target percentage of future Part B expenses covered by premium revenues at 27 percent still would leave the average Medicare beneficiary paying less than 15 percent of the actuarial cost of all Medicare benefits (including both Parts A and B) in premiums.
QMB Protections. Because changes of this sort would create a substantial burden for beneficiaries with low incomes, we also discuss moving the QMB program from Medicaid to Medicare. This would ease burdens on Medicaid at the same time as increasing beneficiary participation in Medicare Part B. Such protections are important for the lowest income beneficiaries, although they would still leave substantial out-of-pocket burdens for the elderly near-poor.
Vouchers and other Private Plans under Medicare
Vouchers are fixed-dollar contributions that beneficiaries could apply toward private insurance premiums in the free market. Vouchers could fundamentally restructure the Medicare program. They could also change the nature of the federal promise to the elderly, depending on the structure of the voucher options chosen. Unconditional adoption of the voucher approach could transform Medicare from a fee-for-service plan that controls service prices and regulates provider conduct into a voucher distribution program that regulates only health insurance plan behavior. This transformation would be analogous to private employers switching their workers from self-insured indemnity plans to health care arrangements with third party insurers. Instead of paying for specific services delivered to enrollees, Medicare would pay premiums or expected average costs in advance. Health insurers and their providers would be responsible for delivering care and would keep any residual funds as profit. Less clear is who would bear the risks of health spending under such a system. If plans are required to take Medicare's payment as the full premium, then the plans bear the risk. If plans are allowed to charge beneficiaries an additional premium for basic benefits, then the beneficiaries bear the risk.
If all Medicare beneficiaries were equally costly, competitive bidding among plans would reveal the lowest possible average premium for Medicare to pay on its beneficiaries' behalf. But this is far from the case. Twenty percent of enrollees account for 93 percent of program expenditures; the other 80 percent cost Medicare almost nothing. Thus, if all Medicare beneficiaries were given the average premium amount as their voucher for use in the private insurance market, Medicare would substantially overpay for most beneficiaries' health care needs and severely underpay for the remaining few. And unless each plan replicates the overall Medicare cost distribution among its own enrollees, some plans will do well and others will not.
Considerable evidence indicates that enrollees who would voluntarily leave fee-for-service Medicare and take their voucher to managed care plans are likely to be healthier than those who would opt to remain in fee-for-service. Because of this self-selection problem and the highly unequal spending distribution among beneficiaries, voucher proposals present two fundamental financial risks to the Medicare program: First, Medicare would probably pay more to buy coverage for the healthy (who would take the vouchers) than those beneficiaries would have cost in standard fee-for-service Medicare. Second, Medicare would still be responsible for the sickest (who would remain in fee-for-service Medicare). Unless payment rules are carefully structured and appropriately adjusted for enrollee risk, a voucher program could end up adding to the costs of the Medicare program. This is exactly what has happened in the Medicare HMO risk contracting program currently operating.
Whether a voucher plan saves money, and if so how much, will depend on a number of interrelated factors—the year-to-year growth in the voucher amount, the proportion of enrollees in voucher and fee-for-service Medicare, the relative costs of those enrolled in the voucher sector versus fee-for-service Medicare, the cost savings that managed care can achieve, the administrative costs that private health plans serving voucher enrollees would incur, and the extra costs that imperfect risk adjustment would impose. Estimating the magnitude of each of these factors under reform conditions is much more art than science.
Under optimistic but still reasonable assumptions, voucher proposals that guarantee traditional fee-for-service Medicare as an option (by allowing beneficiaries to keep fee-for-service coverage with no extra financial penalties or incentives) could save between $11.9 and $27 billion between 1996-2002, depending on whether the voucher was determined by competitive bidding or set administratively to hit a predetermined growth rate. These savings estimates assume appropriate insurance reforms so that the voucher market will work well. But they do not take into account the effect of vouchers on the savings already estimated from the provider and beneficiary changes discussed above. As beneficiaries move out of the fee-for-service sector and use their vouchers, savings that would otherwise come from restricting payments to providers of medical services and imposing higher cost sharing would be reduced because fewer people would be left in the fee-for-service sector. For this reason, net savings from a voucher program would be negative for the 1996-2002 period. Even so, it makes sense for Medicare to pursue a carefully structured voucher program. If the pricing rules and insurance reforms that accompany it are implemented properly, vouchers could expand choice for seniors and achieve significant long-run savings without jeopardizing the quality of care.
Calculating Combined Savings
Designing a package of savings from Medicare is more complicated than simply choosing from among all of the options described above. First, there will be significant interactions among alternative proposals that mean that the whole is considerably less than the sum of the parts. For example, if physician payment changes lower spending on Part B of Medicare, savings from raising the Part B premium will generate fewer savings than if it were the sole change being made in the program.
Difficulties in estimating just how much managed care enrollment would actually rise makes estimating the combined savings from all the options considered taken together hazardous. Nonetheless, this exercise is extremely useful in understanding interactions between fee-for-service and managed care enrollment. Increases in private plan enrollment would lower projected fee-for-service savings and savings from greater beneficiary cost sharing, particularly after the turn of the century when more beneficiaries may enroll in such plans.
Beyond the technical measurement issues, choosing what would be a fair and reasonable approach to a savings package requires considerable thought about questions such as how much should come from one area of the program versus how much from another. These and other issues ought to be considered before choosing any package. The challenge is to find savings that are consistent with protecting the integrity of the program and with achieving a reasonable balance in sharing the burdens of tough policy choices.
This paper raises a number of concerns about various options and offers a number of examples of ways to slow the growth of Medicare spending. However, the options we have discussed here would not add to $270 billion. Moreover, even if you combined all the nonoverlapping options described here, the total would be less than the sum of all the components because of strong interaction effects that would lower savings by 10 to 30 percent.
Finally, one of our most important concerns is with the pace of change, particularly since many of these options represent a substantial break with the past. Given an additional year or two, some of the savings we examined would grow rapidly, so that a relaxation of the deadline of the budget target for another year or more would likely make savings goals easier to attain. Basic reform takes time and forcing savings to occur in a short time horizon can bring unintended consequences.
Notes 1. In 1995, the federal poverty level is $7,467 for a single person and $10,000 for a couple. Thus, the 150 percent level was about $11,200 for a single person and $15,000 for a couple.
2. Disabled persons and those in institutions who receive Medicare also have low incomes, but income data are much more difficult to obtain on such beneficiaries.
3. The QMB program was added to have Medicaid pay Medicare premiums and cost sharing for persons up to the poverty line who do not receive traditional Medicaid benefits. In addition, those with incomes between 100 and 120 percent of the poverty line are eligible under a related program for assistance with Part B premiums.
This publication was funded by The Henry H. Kaiser Family Foundation. To order, call Kaiser's Publications Center at 1-800-656-4533 and ask for document #1109.
View the original article here
