The Senate's budget rules last year put an absolute limit of $400 billion on how much Congress could spend in adding a new prescription drug benefit and making other changes in the Medicare program. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003, signed into law last December, came in just under that limit. According to the Congressional Budget Office (CBO), the new law (since dubbed "MMA" for Medicare Modernization Act) would cost $395 billion from 2004 to 2013. The drug benefit itself would cost $410 billion over the coming decade, with the extra spending offset by cuts elsewhere in Medicare.
Less than two months after the President signed the MMA into law, the White House released a new, higher estimate: $534 billion from 2004 to 2013. The drug benefit suddenly costs one-third more, even before the program has begun. This revelation stirred demands from conservatives in Congress to establish stronger cost-containment measures to hold additional spending to the original $400 billion limit.
There is no way to determine which number is more accurate, but no one should be surprised if the additional amount actually spent by Medicare over the next decade exceeds even the Administration's new estimate. Since the drug benefit is an entitlement, Medicare will pay its share of seniors' drug costs no matter how much is ultimately spent. Moreover, the drug benefit is permanent and does not sunset in 10 years even though the cost estimates span only a decade.
Douglas Holtz-Eakin, director of the CBO, noted in December that new spending for the drug benefit could reach $2 trillion between 2014 and 2023, and that assumes the lower $395 billion cost through 2013.1 That estimate could substantially understate the program's ultimate cost if a future Congress makes drug coverage even more generous, perhaps by filling in the "doughnut hole"2 in the current benefit structure.
By any estimate, the new drug benefit has added greatly to the cost pressures facing Medicare. The first wave of baby boomers will reach age 65 beginning in 2011. Over the succeeding two decades, there will be an unprecedented movement of people from jobs to retirement as some 76 million baby boomers enroll in Medicare. This will decrease our ability to pay for the needs of seniors just when the demand for health services and other support is at its highest. Prudent cost-containment measures are essential if Medicare is to meet this challenge.
The new MMA law contains important design elements that could help constrain Medicare spending in the long term. The drug benefit will be delivered through competing private plans, which will have a strong incentive to negotiate price discounts that will help constrain costs for both seniors and taxpayers. The law attempts to reinvigorate competition among health plans in Medicare, and bidding mechanisms will eventually mean that cost growth could be moderated. Other initiatives include efforts to improve the quality and appropriateness of care, disease management and related mechanisms to optimize treatment for the sickest patients, and greater use of information technology in Medicare. Such steps are promising but unlikely to take Medicare off its spending binge in the immediate future.
In addition to building those cost-containment features into Medicare's payment and delivery systems, the MMA includes an early warning system intended to prompt legislative action if program spending outpaces dedicated program revenues. The effectiveness of such a system is only as good as the willingness of policymakers to take what could be unpopular actions to limit Medicare spending. Stronger actions could be required if the warning system is triggered, but automatic policy responses carry their own risks.
In an era of rising federal budget deficits, policymakers may look for ways that can take effect almost immediately to halt the runaway growth in Medicare spending. Quick fixes to reduce program spending often do not work and might slow the development of reforms that would have a longer-lasting impact on incentives facing health plans, providers, and patients.
There is, however, another approach that can work. A payment formula similar to that used in the Federal Employees Health Benefits Program (FEHBP) would constrain Medicare spending and protect beneficiary interests. But that would require Congress to take on the difficult task of Medicare reform once again.
Title VIII of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 establishes for policymakers an early warning system that reflects the financial activities of all parts of Medicare. That new system improves upon the financial analysis offered in the annual report of the Medicare trustees. Until recently, that report focused attention on the solvency of the Hospital Insurance, or Part A, trust fund without integrating information about the flow of funds through the Supplemental Medical Insurance, or Part B, component of Medicare, which is almost as large as Part A and is growing more rapidly.
Part A is funded by the Medicare payroll tax, a tax on Social Security benefits, and other revenue sources that are specifically dedicated to the Hospital Insurance trust fund. Consequently, solvency is an issue for Part A. By contrast, beneficiary premiums pay 25 percent of Part B costs, and the rest is covered by general revenue, primarily from the personal income tax. The law provides unlimited infusions of general revenue sufficient to cover Part B spending. Part B cannot become insolvent, but the fiscal impact of Part B spending is the same as that of Part A spending.
The introduction of the new Part D prescription drug benefit makes a more comprehensive annual financial analysis of Medicare necessary. Part D is a major expansion of the federal Medicare entitlement, and all of the new program costs are paid out of general tax revenue--primarily the personal income tax--rather than from new funds specifically earmarked for the program.
Spending under Parts B and D combined will exceed Part A spending in 2006, the first year that the full drug benefit will be available.3 The proportion of Medicare spending accounted for by Parts B and D will grow over time, and traditional approaches to assessing the financial status of the program will be decreasingly informative as a result. The new indicator established by the MMA is the percentage of total Medicare spending that is financed by general revenue.
The new law establishes a trigger for subsequent legislative action if Medicare's draw on general revenue grows too large. General revenues may not exceed 45 percent of total Medicare outlays over a seven-year period in two consecutive annual reports. (By comparison, general revenue covers about 35 percent of Medicare spending today.4) If that condition is met, the President must submit legislation that addresses Medicare financing.
Expedited procedures for consideration of such legislation by the House and Senate are included in the new law. The law does not require passage of that legislation and does not preclude the enactment of provisions that would increase Medicare spending despite the excess general revenue condition.
In addition to the new financial reporting mechanism, Title VIII includes a change in the premiums paid for Part B by high-income beneficiaries. Beginning in 2007, Medicare beneficiaries with incomes over $80,000 for an individual or $160,000 for a married couple will pay a higher premium if they choose to participate in Part B. Beneficiaries below those income levels will continue to pay Part B premiums equal to 25 percent of the cost of the benefit. Premiums will rise with incomes, and beneficiaries with incomes over $200,000 (or couples with incomes over $400,000) will pay 80 percent of the average cost of Part B. Those higher premiums would phase in over five years.
The Social Security Administration will determine the incomes of all 40 million Medicare beneficiaries, using data from the Internal Revenue Service and other sources. These data may reflect incomes two years earlier than the year of a premium increase. Social Security must take changes in family circumstances into account when determining whether a beneficiary is required to pay the higher premium.
The new drug benefit moves Medicare very rapidly to the 45 percent general revenue zone, which would trigger legislative consideration of cost-containment action. In 2003, before passage of the MMA, the Medicare trustees estimated that general revenue accounted for about 31 percent of program spending.5 If the MMA had not been enacted, that figure would have grown a few percentage points by 2013, still comfortably below the 45 percent trigger level.
That picture has changed dramatically. According to the CBO, the drug benefit and other changes in the MMA will drive general revenue to 45 percent of Medicare spending in 2013.6 And even that might be optimistic. Spending for prescription drugs might grow more rapidly than projected, and Congress might make the benefit more generous or make changes elsewhere in the program that raise spending without adding to earmarked revenue. In that event, the general revenue provision could be triggered in only a few years.
It is not clear how much of an impact the financial early warning system will have on the public or policymakers. Regardless of how poorly the concept is understood, reports of Part A's impending insolvency have attracted widespread attention and appropriate concern. Such reports may occasionally have prompted cost-cutting moves by Congress. The 45 percent standard seems technocratic and is unlikely to have the same public impact, even though it is a more comprehensive measure of Medicare's financial status.
Policymakers may also be less responsive to the more comprehensive statistic. There is no compelling reason to set the standard at 45 percent instead of 44 or 46 percent, and there is no clear or immediate consequence to exceeding the standard by a percentage point or two. There is no requirement to enact program changes in response to an excess general revenue condition and little reason to think that the majority in Congress would take what could be unpopular actions simply because a statistic has exceeded an arbitrary level.
However, there is considerable value in requiring the President to propose legislation to address rapidly rising program costs. That, rather than the financial statistic, will provoke the public debate necessary to stimulate the enactment of future Medicare reforms.
Title VIII also requires high-income beneficiaries to pay a higher share of the cost of Part B, which makes more explicit the income redistribution that occurs through Medicare. High-wage workers pay more in payroll and income taxes than workers earning less, and high-income seniors were already contributing more to the operation of Medicare because of the tax on Social Security benefits. The higher Part B premium reduces further the subsidy given to high-income seniors, although most will continue to gain more than they have paid into the program over their lifetimes.
Policymakers were concerned that raising the direct cost of participating in Part B, which is voluntary, could cause high-income seniors to drop out of that program. Consequently, the Part B premium is capped at 80 percent of the average cost of covered outpatient services, ensuring that seniors affected by the higher premium would still receive a subsidy of at least 20 percent.
In addition, the premium increase is limited to a small fraction of the Medicare-eligible population. Even a substantial dropout rate among high-income seniors would result in the loss of only a few thousand people from the program.
Raising the Part B premium, even for a small subset of seniors, poses a challenge and an opportunity for Medicare: Consumers who pay more are likely to demand a health program that is more responsive to their needs. Congress has taken some initial steps toward reforming Medicare by enacting the MMA, but the program remains overly regulatory and rigid. Future initiatives can complete the job that has just begun by more thoroughly reshaping Medicare.
The Federal Employees Health Benefits Program has shown that government health programs can offer greater variety in insurance offerings, adapt to changes in medical technology, and constrain cost growth while ensuring that consumer and taxpayer interests are protected.7
Congress is facing high and rising federal budget deficits. There are pressing demands to increase spending for defense, homeland security, tax relief, and many other policy priorities. In that context, some approach that could constrain Medicare costs and reduce the chance of a fiscal surprise due to unexpected growth in Medicare spending would be welcome. One could consider strengthening the budgetary control provisions in Title VIII of the MMA, but cost controls that appeared to be effective might make matters worse.
True cost containment is possible only if the efficiency with which resources are used in Medicare is improved. That means establishing appropriate incentives for providers and beneficiaries, as well as promoting competition and more cost-effective use of health services.
Other titles of the MMA introduce reforms that begin to restructure Medicare and reduce unnecessary spending. Title VIII, however, does not add meaningful cost containment to Medicare. Instead, it adds a notification process that by itself does not improve the structure or functioning of the program. Serious cost containment will require further steps. Specifically:
The real value of the general revenue trigger as currently written is to initiate the policy debate over exploding Medicare costs. Since there is no requirement to enact cost-cutting legislation in response to that trigger, it makes little sense to delay the debate for as long as envisioned in Title VIII. Nothing happens the first time the Medicare trustees project excess general revenues, despite the fact that a significant financial imbalance is likely to occur within seven years. Instead, Congress and the President can wait for a second report a year later before starting to develop policies in response to the problem.
It is likely that the President will not wait for the second report to develop new proposals, but there is no reason why this could not be made a requirement. As the entry of the baby boomers into Medicare beginning in 2011 nears, the excess general revenue warning might be sounded annually until comprehensive reforms that build on the MMA are adopted. That will force the debate into the public eye and may promote broader understanding of the issues and the policy alternatives.
In addition, the President's charge could be strengthened. As currently written, the law requires the President to submit legislative proposals to Congress in response to the warning issued by the trustees, but there is no requirement that the proposals reduce program cost. Proposals that increase Medicare's dedicated revenue, such as increases in the payroll tax or Part B premiums, do not address the program's inefficiency and might delay more thorough reforms. The President's response could be required to contain a preponderance of cost-reducing proposals.
The general revenue provision in Title VIII could be given some teeth. For example, the law could be changed to require automatic reductions in Medicare payments to providers and health plans once general revenues are expected to exceed the 45 percent standard.
Automatic increases in payroll taxes and premiums for Parts B and D could also be imposed, which would raise program revenue (and help reduce the overall federal deficit) but would not reduce program cost. Alternatively, one could tie automatic cost-cutting policies to the new drug benefit itself. A proposal to keep Medicare drug spending to $400 billion was advanced last year but was ultimately dropped from consideration.
Such automatic provisions would alter the political climate of cost containment in Medicare. Once enacted, spending cuts that are triggered by some financial indicator could be overridden only by new legislation, which could prove to be difficult. Such a policy puts the burden of proof on policymakers who oppose spending cuts rather than on those who support cuts. This is the reverse of the current situation, which requires new legislation to address an excess spending problem in Medicare.
There are, however, drawbacks to putting cost containment on autopilot. Mechanisms that automatically trigger cuts in Medicare may be inflexible and not responsive to new circumstances that call for changes in cost-containment policies. Cost controls will not foster sustainable improvements in Medicare, and automatic across-the-board cuts might inadvertently undo reforms that have already been accomplished.
Targeting automatic cuts on specific health services may not be an improvement. In particular, price controls on drugs offer short-term savings, but controls also would retard the research and development necessary to create new treatments for diseases common to the elderly.
Automatic payment cuts are blunt tools and may not be appropriate in future years. Market conditions facing providers, economic conditions affecting beneficiaries, or the balance of political power among competing interests in subsequent years may create a need to alter previous policy, which would be difficult.
Congress is familiar with this problem. In 1997, lawmakers created the sustainable growth rate (SGR) formula to constrain the growth of Medicare physician payments. The formula created a trigger for automatically cutting physician fees paid by Medicare.
That seemed like a good idea until 2002, when fees were cut by 5.4 percent. Seniors in certain parts of the country found it increasingly difficult to make appointments with some doctors, who chose to close their practices to new Medicare patients in reaction to the lower payment rates, and the physician community was up in arms over such a substantial cut.
The fee reductions do not seem to have been effective in slowing Medicare spending, however. Although some physicians were paring back their Medicare business, others reacted to the fee cut by increasing the volume of services paid by the program. Medicare physician payments increased by nearly $3 billion over the year even though the average fee for each service was reduced.
Since then, Congress has tried several times to reverse the policy, succeeding for short periods of time but at some political cost. The biggest obstacle to changing Medicare's physician payment policy is the loss of billions of dollars of budget savings if the SGR formula were permanently set aside. The advantage of the SGR policy--cuts in fees that would not require new legislation--has become the barrier to making Medicare payments to physicians more realistic or more politically acceptable.
Proposals to cap the cost of the new drug benefit or automatically limit Medicare spending rely on government price controls rather than market competition. Such approaches can cut costs in the short term by limiting the supply of health care services to seniors. They do nothing to improve the efficiency of health care delivery in Medicare, and may result in continued subsidies to inefficient providers or continued waste in the use of services. Controlling Medicare costs through automatic across-the-board reductions in payments threatens to undo the reforms already accomplished by the MMA.
The Medicare+Choice (M+C) program illustrates this point. Congress tried to expand the range of private health plan choices available to seniors by creating M+C in 1997. Rather than enhancing choice and competition, new payment rules that were expected to keep Medicare costs under control contributed to an exodus of private plans from Medicare. Those rules did not allow payment growth to keep pace with the rising cost of services. Consequently, one of the essential elements of competitive Medicare reform--the presence of plan competitors--was dealt a blow by excessively tight price controls.
It would be ill-advised to repeat the M+C experience. Proposals to apply new price controls to provider payments in the name of limiting Medicare spending and reducing the federal deficit could have the same unwelcome consequence for Medicare Advantage.
Savings, Long-Term Costs
An across-the-board reduction in Medicare payments to providers is a blunt instrument that is unlikely to gain political acceptance. Cuts also could be targeted on specific health services, although the program's experience with the SGR policy used to determine physician payments should give anyone pause.
Policymakers of all political stripes have proposed various forms of price controls for prescription drugs, ranging from direct federal price negotiation with pharmaceutical manufacturers to importation of pharmaceuticals from Canada or other countries that have their own price controls. It is a short step from government price setting to an automatic mechanism to cut those prices once some measure of excess spending is triggered.
Such a strategy could reduce program outlays without immediately reducing the supply of pharmaceuticals available to seniors. Manufacturers typically set prices that are substantially above the cost of producing an additional dose of a drug, but those high margins are needed to cover the costs of research and development for that same drug and the large number of others that fail to reach the market. If prices remain above the cost of production and distribution, manufacturers will continue to sell existing products, and shortages are unlikely.
Over the long term, however, a federal price-cutting mechanism will discourage private funding for research and development. Fewer new drugs will enter the market, particularly drugs that are used primarily by seniors and thus are subject to price controls. Total Medicare spending might be even greater in the long term with controls on pharmaceutical prices if new cost-saving therapies for diseases like cancer or heart disease never come on the market.
Various schemes to impose top-down cost containment on Medicare are quick fixes that merely fine-tune a flawed program. Without other changes in Medicare's structure, patients and providers continue to face incentives that promote overuse of services. The growth of program spending can be reduced over the long term if we adopt measures that promote a better functioning market in Medicare--one that enhances consumer choice and rewards the delivery of effective, high-quality care.
The MMA includes provisions that promote competition among private plans in Medicare, and it introduces some new initiatives to improve the quality and effectiveness of care delivered to seniors. Those first steps could be improved upon by moving to a payment structure modeled after the Federal Employees Health Benefits Program.
Rather than attempting to set the prices of thousands of individual health services delivered in every local market in the United States, the FEHBP uses a payment formula that sets the government's contribution to the premium charged by competing health plans. Currently, the government pays 72 percent of the weighted average premium of all health plans participating in the FEHBP. The government contribution can be no greater than 75 percent of any plan's premium.
The FEHBP's formula limits federal outlays but keeps the federal contribution in line with general health cost increases. It also allows enrollees to select the health plan that best meets their needs. But enrollees who select a plan with high premiums must pay the full amount over and above the maximum federal contribution. Since the federal contribution is capped, health plans have an incentive to price their products competitively. As a result, FEHBP costs have grown at about the same rate as those of Medicare even though the FEHBP has offered superior benefits (including prescription drugs).8
If Medicare adopted a "premium support" approach similar to the FEHBP payment method, adjustments would be necessary to ensure that beneficiaries with low incomes or chronic health needs were protected. The MMA introduced income-related subsidies for the new drug benefit and tied Part B premiums to income, setting the precedent for more explicit recognition that low-income beneficiaries need more financial support through Medicare. Payment risk adjusters have been developed for plans participating in M+C to account for enrollees with higher health costs, and those adjusters could be adapted to an FEHBP-style formula.
Premium support would change the inappropriate financial incentives built into traditional Medicare. Providers in traditional Medicare are paid on a fee-for-service basis and are guaranteed federal payment for all necessary services--a loose concept at best. Beneficiaries typically have additional coverage from private insurers or Medicaid that pays for most of their out-of-pocket costs.
Cost-containment methods that rely on price controls and on lowering payment rates are often ineffective because providers can expand their services while beneficiaries have little financial reason to refuse care that has only minimal health benefits. Where traditional cost-control methods are effective, patient access to care is often interrupted.
Premium support would create incentives for both providers and beneficiaries to seek appropriate and efficient care. Over the long term, such an approach can slow the growth of Medicare outlays, reduce unnecessary care, and promote customer service and quality improvements.
Concern over the skyrocketing cost of Medicare is well-founded. Medicare will spend $300 billion in 2004, and those outlays are expected to grow 9 percent a year for the next decade. That growth is fueled in part by the new drug benefit, the cost of which is even now being debated by policymakers and estimators. The one certainty is that taxpayers must pay whatever is demanded of a program that remains an open-ended entitlement.
The formal cost-containment provisions in Title VIII of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 do little to address the cost problem in Medicare. The provisions require the Medicare trustees to give a more comprehensive measure of Medicare's financial status, and they also require that the President submit legislative proposals to Congress if Medicare requires too large an infusion of general revenues.
Title VIII does not include stronger cost controls, such as spending cuts that are triggered in response to the new financial measure, but that is fortunate. Those types of cuts are at best palliative and do not address the underlying incentive structure that drives Medicare spending growth.
If we hope to control Medicare spending over the long term, Congress must build on the competitive reforms contained in the MMA. While political consensus on further reforms would be very difficult to muster, simpler approaches are unsustainable.
Instead of working on new and more complicated ways to set thousands of prices, Medicare should adopt a premium support framework modeled after the Federal Employees Health Benefits Program. Such a framework makes responsible cost containment possible. Beneficiaries can be offered realistic choices among health plans. Instead of imposing arbitrary payment cuts, the federal government can make contributions under premium support that can be limited while providing necessary protections to beneficiaries.
2. Beneficiaries would have 75 percent of their drug spending covered by the new Part D for the first $2,250, after satisfying a $250 deductible. Those spending more than $2,250 would receive no additional reimbursement until they have spent $3,600 out of pocket. The gap in coverage is called the "doughnut hole."
5. Author's calculations for fiscal year 2004 using data from Tables II.B5 and II.C5 of 2003 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, at http://www.cms.hhs.gov/publications/trusteesreport/default.asp.
7. See Walton Francis, "The FEHBP as a Model of Medicare Reform: Separating Fact from Fiction," Heritage Foundation Backgrounder No. 1674, August 7, 2003, and Joint Economic Committee, Health Insurance Spending Growth--How Does Medicare Compare? U.S. Congress, June 10, 2003.