Skip Navigation
| Skip Navigation |
|
|
|
|
|
February 18, 1999
The Honorable John B. Breaux
United States Senate
Washington, D.C. 20510
Dear Senator:
I am pleased to respond to your letter of February 4. We do not have
specifics on many aspects of your proposal, so our response may be less
precise than you or others would prefer. However, I hope that what we say
is at least helpful and that we can continue to assist you as you refine
your proposal. I believe that the most important piece of the analysis
at this stage is to get the questions right and begin to suggest how your
proposal might change the Medicare program.
Summary
Under current law, health plans in the Medicare program compete on the basis of covered benefits and quality of service, not on price. Your proposal would foster greater competition among plans and greater choice for beneficiaries. We believe increased competition will reduce costs. As the attached paper indicates, the details that remain to be specified would determine the ultimate effectiveness of the proposal in slowing the growth of Medicare's costs. But the general direction of the proposal is clearly promising.
Reducing Medicare's costs should not be the only goal of reform. Costs could be reduced--without necessarily ensuring Medicare's long-term financial stability--by cutting payments to providers, reducing access to services, or making other changes that are likely to reduce the welfare of Medicare beneficiaries. An effective reform would introduce strong new incentives for efficiency. Other important goals of reform include ensuring an acceptable level of quality and access to services and allowing maximum flexibility for beneficiaries to choose a plan that meets their needs. Needless to say, proposals must also be feasible to implement. Designing a proposal that meets all of those goals is clearly a tall order!
Your proposal attempts to address those issues. Its ultimate success will depend on the details of its design and on the interaction of a restructured Medicare program with other programs.
The Congressional Budget Office (CBO) does not have the ability to assess
alternative policies with any precision once we move past the 10-year budget
window. Like the Medicare trustees, we have projections only over the long
term--projections that make assumptions about general changes in policy.
By contrast, long-term analyses require a baseline free of unreasonable
assumptions about the course of spending without major policy interventions.
Discussion
Although we cannot provide a cost estimate of your proposal, we can offer a preliminary analysis that is perhaps less satisfying but potentially more informative. We suggest a few principles by which to assess the potential for changes in policy to reform Medicare. Those principles are certainly related yet different enough to justify their separate consideration.
First, we believe that introducing competition into the Medicare program could help to reduce costs in both the short and the long run. A premium support system that resulted in effective price competition among plans would most likely lower Medicare costs.
Second, Medicare reforms should also enhance efficiency--the productive use of medical resources. If beneficiaries face choices among health plans, they tend to recognize more readily the trade-offs those choices entail. Allowing greater choice results in a more effective use of health care resources. Another issue related to efficiency is the considerable excess capacity that exists in the U.S. system for delivering health care. In 1997, for example, about 40 percent of all hospital beds went unoccupied on an average day, even though the number of beds had declined by 20,000 from the year before. Similarly, there is some evidence of an oversupply of physicians, at least in particular markets. Your proposal could help to reduce some of the costs associated with the inefficient use of health resources.
Third, reforms that improved efficiency could maintain the quality of health care while reducing its costs. The goal of any change in policy should be to at least maintain the system's quality, if not improve it. Unfortunately, there is little agreement about how to measure the quality of health care, particularly for the elderly. What is clear is that improving quality is not synonymous with increasing expenditures.
Your proposal would maintain the government's large contribution toward the care of Medicare beneficiaries. That contribution level is well in excess of the level in health insurance programs for federal employees, such as the Federal Employees Health Benefits Program (FEHBP). Expanding pharmaceutical coverage in private plans--to the extent the costs do not squeeze out other, more effective treatment--could improve the quality of care. Again, the specific design aspects of the reform proposal will have a critical bearing on the actual outcome of the policy.
Fourth, allowing beneficiaries to choose among multiple plans will help to modernize the Medicare program and allow the elderly to select benefits that are more closely aligned with their needs. As the commission knows, most Medicare beneficiaries are still enrolled in the traditional program formulated 35 years ago, which has significant gaps in coverage compared with the typical employer-sponsored plan of today.
Finally, it is obvious but true that any reform proposal must actually work--that is, it must create a system of rules under which the intended effects can actually occur. Of course, there are practical limits on how burdensome and intrusive such a system might be. Your proposal is modeled in part on the FEHBP, which could provide useful guidance for implementation. However, a restructured Medicare program would be considerably more complex than the FEHBP. The additional responsibilities of the proposed Medicare Board, the potential expansion of the number of competing plans, and the large number of Medicare beneficiaries make the implementation of reform a formidable challenge.
Medicare's many interactions with current programs will affect the ultimate
success of any reform, and two of those interactions merit particular mention.
Most fee-for-service enrollees have supplemental insurance coverage through
medigap policies, employer-sponsored insurance, or Medicaid. That additional
coverage increases Medicare spending by encouraging greater use of services.
To the extent reforms mitigate that incentive, Medicare spending could
be reduced. In addition, restructuring Medicare would establish a new,
complex relationship between the Medicare and Medicaid programs. That relationship could have important
implications for federal costs and the quality of care for dually eligible
beneficiaries.
Estimating Issues
Reforming programs such as Social Security and Medicare is challenging for many reasons, not least because of the need to assess the long-term effects of any change. Although the solvency of Medicare's Hospital Insurance Trust Fund has been the focus of much policy debate, we know that it is not an accurate measure of the fiscal health of the program.
We also know that the Medicare trustees' long-term projections of spending include assumptions about future, unspecified changes in behavior and policy. The trustees essentially assume that Medicare's increasing claim on the economy and the federal budget--following Herb Stein's dictum--"cannot go on forever" and that something will happen to slow the growth in spending.
They are clearly right in that assumption, but by itself, the assumption provides little help in assessing the impact of various policies. Indeed, it may well be your policy proposal that will produce their outcome. However, it is simply not legitimate to "score" or compare any proposal with the trustees' projections. For long-run comparisons, a baseline is needed that is free of unreasonable assumptions about the course of spending without major policy interventions.
Senator, I am sure this is both more and less than what you expected as a response. Issues of health care are unusually complex, but we can also get lost in the complexity and in the elegance of our analysis. I think it is important to keep in mind a set of principles for reform and to try to assess the desirability of any plan relative to those principles. We certainly have not cornered the market on defining such principles, or assessing the impacts, but I hope this response provides a useful template for further consideration.
If you have any question about CBO's analysis, please call me. If your staff has any questions, they may call Joseph Antos or Linda Bilheimer.
Sincerely,
Dan L. Crippen
Director
c: The Honorable William M. Thomas
Enclosure
OVERVIEW
The aging of the baby boomers will place unprecedented demands on the Medicare program. Between 2010 and 2030, the elderly population will grow at an annual rate of almost 3 percent, rising from 39 million to 69 million. Medicare costs are likely to grow considerably faster than program enrollment because costs per beneficiary are also likely to increase rapidly. To reduce the growing share of the nation's resources that the Medicare program would otherwise absorb, major policy changes are necessary to slow the rise in costs per beneficiary.
The Bipartisan Commission on Medicare Reform is considering a premium
support model as a basis for restructuring the Medicare program. That approach,
which adopts some of the attributes of the Federal Employees Health Benefits
Program (FEHBP), is intended to produce greater competition among health
plans serving the Medicare population and greater choice for beneficiaries.
A premium support system that resulted in effective price competition among
health plans would have the potential to lower Medicare's costs.
BACKGROUND
Under current law, Medicare beneficiaries may enroll in the traditional fee-for-service plan or in private health plans that serve Medicare beneficiaries in the Medicare+Choice (M+C) market. The large majority of enrollees have chosen to remain in the fee-for-service program, but the Congressional Budget Office (CBO) projects that the percentage of beneficiaries in private plans will double over the next 10 years, rising from 15 percent in 1999 to 31 percent in 2009. By contrast, more than 85 percent of workers with employer-sponsored health coverage are currently in some form of managed care plan.
Most beneficiaries in the traditional program have some form of supplemental coverage to pay for their deductibles and copayments. Almost one-third of those beneficiaries pay for private medigap insurance; a similar proportion obtains supplemental coverage as a retirement benefit from former employers. Supplemental coverage raises Medicare's costs because beneficiaries who do not face cost-sharing requirements use more of the services covered by the program. Medigap premiums are rising rapidly, however, and employers are becoming less willing to provide coverage for retirees. Those factors will contribute to growth in the proportion of beneficiaries enrolling in managed care plans that have low cost-sharing requirements and provide additional benefits, such as prescription drug coverage.
Before enactment of the Balanced Budget Act of 1997 (BBA), Medicare's payments to health plans were based on average fee-for-service costs in each county. That system resulted in wide variations in payments to plans and considerable volatility in payments from year to year. It also meant that plans had incentives to compete on the basis of the benefits they covered rather than on price.
The BBA introduced Medicare+Choice with the intent of reducing payment variation and volatility. In each county, the payment that health plans now receive is the highest of:
The annual growth in the components of the blended rate and in the floor amount is determined by the projected growth in per capita spending in the fee-for-service sector, less a statutory reduction for 1998 through 2002. Other payment changes in the BBA will also lower payments to health plans. Thus, before the act, Medicare paid plans about 95 percent of per capita costs in the fee-for-service sector, but that rate will drop to about 90 percent when the BBA provisions are fully phased in. Nonetheless, the rate of increase in payments to plans remains tied to growth in per capita spending in the fee-for-service sector. More fundamentally, the payments that plans receive are still unrelated to their performance.
Program rules foster competition among M+C plans on the basis of expanding
benefits rather than lowering premiums. If an M+C plan makes profits that
are higher than the Medicare rules allow, the excess must be returned to
enrollees as additional benefits. Plans may not offer rebates to enrollees.
(Excess profits could be returned in the form of a rebate to the federal
government, but all plans prefer to offer additional benefits because of
the obvious marketing advantage.) Beneficiaries pay a premium (in addition
to the Medicare Part B premium, which all beneficiaries pay) only if the
cost of the plan that they select is higher than Medicare's payment. However,
only a minority of health plans currently charge an extra premium.
THE PROPOSAL
The premium support approach would tie the government's contribution for each health plan, including traditional Medicare, to the national weighted average premium. Beneficiaries selecting lower-cost plans would have a larger share of their premium subsidized by Medicare than those selecting higher-cost plans, and the core benefits offered by plans could vary only within a limited range. Two options are under consideration; they differ only in the schedule of federal premium contributions.
This preliminary assessment of the proposal is based on the following assumptions, which CBO staff developed after discussions with commission staff and receipt of a letter dated February 4, 1999, from Senator Breaux.
KEY ISSUES REQUIRING CLARIFICATION
Those assumptions, and other design elements not listed above, would determine the effectiveness of the commission's premium support approach in slowing the growth of Medicare spending. Changing any key element of the proposal could have a profound impact on program costs. Some of the more important aspects of the proposal that need further clarification include:
In addition, it has been suggested that the premium support proposal
might include a provision that would require higher-income beneficiaries
to make larger premium contributions. The specifications that CBO analysts
discussed with commission staff did not include a provision for means-tested
premiums, and that issue is not discussed in this attachment. However,
such a provision could have a significant effect on Medicare costs under
a premium support system.
EFFECTS OF THE PROPOSAL ON MEDICARE'S COSTS IN THE SHORT TERM
As described above, the payments that M+C plans receive bear no relationship to their performance, and the plans have no incentives to compete on the basis of price. By contrast, under the premium support model, health plans would be given new flexibility to compete by reducing premiums or enhancing benefits. That additional element of price competition might result in beneficiaries having a broader array of plans from which to choose, thus enabling them to select a plan that meets their needs more appropriately than the choices currently available to them.
The interaction between beneficiaries' choices of health plans and decisions by plans about what benefits to offer and what premiums to charge would affect program costs in complex ways. Many beneficiaries would make decisions that would leave government costs unchanged. For example, beneficiaries who did not change plans would not generally increase government costs. (They could cost Medicare more, however, if their plans were not already receiving the maximum government contribution and chose to raise their premiums.) In addition, as is similar to the situation in M+C today, some beneficiaries enrolled in traditional Medicare who purchased medigap policies might find a competing plan that would be an attractive alternative. Switching to a private plan might lower their own costs because they would no longer be paying a separate medigap premium, but it would not necessarily change federal costs.
Some plans might seek to expand their enrollment by enhancing their benefits while still remaining competitive in terms of price. Some M+C plans, for example, have costs below those of the fee-for-service program and charge no additional premiums. Those plans could upgrade their benefits, raise their premiums to the level of the national average, and still compete with the fee-for-service plan. Plans currently offering benefits that cost between 90 percent and 100 percent of the national average, for instance, might find that opportunity quite attractive. Their enrollees would pay only 33 cents for every dollar of increased benefits, up to the national average. Such increases would boost the national average premium in the short term.
To capitalize on the demand for lower-cost coverage, other plans might decide to reduce their benefits and market themselves as low-cost alternatives. It is reasonable to assume that some beneficiaries would move from traditional Medicare--whose premiums would be close to the national weighted average in the short term--to a more preferable plan with premiums below the national average. Government costs would fall for beneficiaries who chose less expensive health plans only if they selected plans that would receive a lower government contribution than their current plan.
The ongoing shift from the traditional fee-for-service sector to managed care that is occurring under current law could accelerate under a premium support system. With premium support, costs in the fee-for-service program would largely determine the national average premium for several years, that is, until the majority of beneficiaries were enrolled in competing plans. If people moved from traditional Medicare into lower-cost plans--those with premiums below the national weighted average--the average premium would fall. That outcome would lower the government's total contribution for premiums. In addition, the traditional program would become an increasingly costly option for beneficiaries unless it could lower its premiums as well.
The adjustments that the Medicare Board made to premiums to reflect geographic differences in health care costs could also affect the government's costs. If the adjustments reflected only differences in input costs and did not incorporate the effects of differences in service utilization, plans operating in high-cost markets might face significantly lower payments than they currently receive and might have to reduce their benefits. Conversely, plans in low-cost markets would gain from such adjustments and have more flexibility to enhance their benefits and raise their premiums. How local plans might change their benefits is uncertain, as is the resulting net effect of those changes on the national average premium.
The premium adjustments would also influence the number of plans electing
to participate in different markets. The adjustments would, at best, only
approximate the underlying cost differentials among geographic areas. Consequently,
as they do today, plans would seek out markets in which their projected
per capita costs would be significantly lower than the adjusted per capita
payment--and avoid markets in which the converse was the case.
EFFECTS OF THE PROPOSAL ON MEDICARE'S COSTS IN THE LONG TERM
If the Medicare program became more competitive, with a much higher percentage of beneficiaries enrolled in private plans that competed on the basis of price and quality, the future growth of program spending would be more closely tied to trends in private health care markets. A major incentive for restructuring Medicare is to generate the same competitive forces within the program that the private sector experienced in the mid-1990s. Between 1993 and 1996, the growth of employer-sponsored health insurance premiums slowed dramatically as a result of the shift to managed care and increasing competition among health plans. By contrast, Medicare spending per enrollee continued to rise rapidly.
Whether recent experience in the private sector reflects longer-term spending trends is uncertain, however. Over the past year, premiums for employer-sponsored insurance have once again begun to grow more rapidly, as health plans that had held down premiums to capture a larger market share have sought to improve their profit margins. As a result, controversy has arisen about the long-term effects of managed care on prices and costs in the private health care market and whether slower cost growth associated with the shift to managed care is a one-time phenomenon.
Analysts generally agree that part of the recent slowdown in private health insurance premiums did, indeed, reflect a one-time change in the level of premiums, as employers switched their employees from higher-cost to lower-cost plans. But most analysts do not anticipate a return to the double-digit rates of growth in premiums that occurred before 1993. Both employers and health plans now function in a much more competitive health care environment than existed 10 years ago. Purchasers are likely to continue to be aggressive in pressuring plans to hold down premium growth, and plans will continue to seek innovative ways to control costs while constraining payments to providers. Moreover, persistent excess capacity in the health care system will continue to give plans leverage with providers.
If, however, the current trend toward consolidation among health plans continues, so that only a few plans operate in any market, the incentives for price competition among plans may be reduced. (The number of plans operating in a market does not necessarily predict how competitive that market will be.) But whether consolidation will continue in the long term or whether new patterns of market organization may emerge is still uncertain.
As in the private sector, analysts do not anticipate a return to double-digit growth in Medicare's per capita costs over the next decade. CBO projects that per capita spending growth in the program will be slower, on average, over the next 10 years than in the 1990s. But that projection primarily reflects payment policies affecting the traditional fee-for-service program. After 2010, the program will begin to experience the extraordinary demographic pressures associated with the retirement of the baby boomers. Addressing that boost in demand will require growth in per capita spending that is slower than the growth that will occur under current policies.
Whether a more competitive approach slowed Medicare spending in the
long term would depend in part on the competitive environment that existed
more generally in health care markets. It would also depend on how aggressive
the Medicare Board was in its negotiations with health plans and whether
the board would be allowed to negotiate with the traditional fee-for-service
program.
THE ROLE OF THE BOARD
Commission staff compare the Medicare Board's role to that of OPM in overseeing the FEHBP. But if the board had limited authority to negotiate with the traditional program, its task could be much more difficult than OPM's because the traditional program would be the market leader--at least in the early years of the program. OPM exerts considerable control over the national plans that offer services under the FEHBP, especially Blue Cross and Blue Shield, which is the market leader and accounts for more than 40 percent of federal enrollment. Within the FEHBP, the national plans are the major competition for local health plans, just as the fee-for-service program is the major competition for private health plans under Medicare.
OPM seems to use its market power in modest ways to extract favorable terms from local health plans. The plans are required to provide OPM with detailed information on their premiums, and how they were developed, for the two employer groups that are closest in size to their federal employees' group. OPM uses the lower of those two rates to establish the premium for the FEHBP. Whether the Medicare Board would be able to fully exploit its considerably greater market power is uncertain.
How effective the board was in limiting the expansion of covered benefits would be of critical importance for long-term spending growth. The rate of growth of the national average premium would be a function, in part, of the services that plans included in their premiums for core benefits. There would be tremendous pressure to continue to expand those benefits as a result of the rapid development of medical technology. That pressure exists today but is likely to increase in the future, especially considering that many future medical breakthroughs will probably be targeted toward the elderly market.
Under the proposal, the board's authority with respect to prescription drugs would apparently be limited, which could have a sizable effect on program costs. The proposal would allow private plans to include the costs of prescription drugs in their premiums for core benefits. Thus, a new service with rapidly rising costs would be built into the base for determining the government's contribution, potentially causing Medicare's long-term costs to grow more rapidly as well. Initially, the effects on the national average premium would be small because most beneficiaries are in the traditional program, which would not offer drug coverage. But over time, the effect could be compounded if more beneficiaries shifted to private plans that offered drug coverage, which in turn could cause prescription drugs to become an increasingly important component of the national average premium.
Pressure by beneficiaries to expand covered benefits is also likely
to grow over the next decade and beyond, regardless of any policy actions
taken to reform Medicare. When the baby boomers retire, they are going
to be wealthier, on average, than previous generations of retirees. They
are therefore likely to be more willing to pay for plans charging higher
premiums if those plans offer richer benefits or are judged to be of higher
quality. Under a premium support model, many of those plans would also
have higher federal contributions. If the demand for new benefits was strong
and was backed up by beneficiaries' willingness to pay for them, the board's
ability to limit "benefit creep" could be compromised.
THE ALTERNATIVE OPTION
The commission has developed a second option for consideration that differs from the first only in having a different structure of government subsidies for Medicare. Beneficiaries would pay:
The steepness of the schedule could discourage benefit creep somewhat because
beneficiaries would pay a larger share of the costs of additional benefits
than they would under Option I. But given the high percentage of the premium
that the government would pay--regardless of the plan a beneficiary chose--it
is unclear whether small changes in beneficiaries' contributions would
have much effect on their choice of health plans. The schedule might also
encourage plans to establish premiums that were about 85 percent
of the national average. Because such plans would probably have "lean"
benefits, however, it is unclear whether they would capture a significant
share of the market.
MEASUREMENT AND BASELINE ISSUES
Estimates of the long-term effects on costs of any proposal to restructure the Medicare program depend critically on the baseline against which the proposal is measured. Ideally, such a baseline would assume that current policies would continue without the introduction of significant program reforms. It is reasonable to assume that over the long term, without restructuring the Medicare program, the government would continue to adjust its administered prices, as it has in the past, in an attempt to slow the growth in outlays.
CBO does not currently have a baseline that extends beyond a 10-year window. The Medicare trustees make long-term projections for the program that might be considered for such a purpose, but those projections assume that growth in per capita spending will decline to the rate of growth of hourly wages by 2020. Such a reduction in the rate of growth is unlikely to occur in the absence of policy actions that go significantly beyond the adjustment of administered prices.