Medicare is facing severe financial strains that threaten its future viability. On a per capita basis, Medicare spending is increasing at twice the rate of the gross domestic product, and, according to Medicare's chief actuary, the program is facing a breathtaking funding shortfall of $62 trillion--nearly six times larger than the much-discussed shortfall in Social Security. The newly enacted Medicare prescription drug benefit could cost more than $700 billion over the next 10 years and will only add to the program's financial woes.
That new drug law would provide a sizable net benefit to retirees and older workers without existing coverage, even if Congress immediately funded it through higher Medicare payroll taxes. Workers born before 1965--baby boomers and current retirees--would receive a net gain of about $20,000 per capita. Younger workers and all future generations, however, would suffer net losses of between $2,500 and $4,000 per capita. Furthermore, failure to include meaningful Medicare reforms in the drug program may cause steeper cost escalations, diluting its benefits.
Congress should revisit the Medicare prescription drug program and insist on significant market-based reforms, not merely an ever expanding array of benefits.
Even more than Social Security, the Medicare program is in deep trouble, facing demographic and financial pressures that threaten the future viability of the program. The demand for health care benefits will rise steeply as the retiree population balloons and retirement life spans grow longer. An explosion of the population of those eligible for Medicare coverage--78 million baby boomers--will commence this decade.
Health care providers--physicians, hospitals, and others--have developed better diagnostic and treatment methods for many diseases. Technological development will likely continue in the future, adding to our arsenal of effective treatments. However, new treatments are likely to be expensive because of high development costs. Therefore, under current Medicare reimbursement rules, spending per Medicare enrollee will also continue to grow rapidly, adding to spending growth due to purely demographic changes.
That implies growing pressure on the federal budget to find dollars to continue feeding the Medicare beast. Medicare expenditures were 0.7 percent of gross domestic product in 1970 and grew to 2.6 percent in 2003, reflecting rapid increases in the factors affecting the growth of health care costs. That trend is likely to continue at an accelerated pace. To meet the rising demand for more and better quality health care and yet slow the pressure generated by Medicare on the budget, Medicare reforms need to be considered and implemented--the sooner the better.
Simply spending money does not necessarily lead to better health care. Despite the explosive growth in Medicare spending, the program still does not provide complete coverage of seniors' health care needs. Responding to one of Medicare's inadequacies, Congress passed the Medicare Modernization Act in 2003 to add prescription drug coverage.
The drug coverage attempts to restructure Medicare benefits in the face of changing health care needs of seniors and the disabled--who, in the future, are more likely to rely on prescription drugs than on traditional medical treatments and procedures. The drug coverage, however, adds to the financial challenge facing Medicare. In their 2004 report, the Boards of Trustees of the Federal Hospital and Medical Insurance Trust Funds estimated that the drug benefit's total cost is about $16.6 trillion in present value terms. That comes on top of the already massive financial shortfall in Medicare, some $45.0 trillion even without the new benefit. By enacting the drug benefit, Congress has increased the financial burden of Medicare by more than a third. In 2006, when the full phase-in of prescription drug coverage is completed, Medicare's projected expenditures will immediately jump to 3.4 percent of GDP.
Such rapid growth in Medicare costs is unsustainable, even for the largest economy in the world. We need prudent reforms that will encourage competition among health plans, give beneficiaries realistic choices, and begin to slow the overall growth of Medicare spending. Adding a drug benefit alone without thorough reforms sharply increases the cost of Medicare to future taxpayers and jeopardizes the financial viability of the program.
In this paper, we examine the likely cost of a Medicare prescription drug benefit over the next 10 years and in the decades to come. We make three points:
- Actual program cost for Medicare prescription drugs through 2015 will probably exceed the $720 billion that has been estimated as the cost of the benefit under current law.
- Constraints could be placed on spending for prescription drugs, but that approach would likely limit the availability of new and beneficial drugs. A more comprehensive strategy is needed to limit taxpayer cost while ensuring that beneficiaries receive appropriate services.
- The long-term cost of the new drug benefit is astronomical, and most of it will likely fall on our children and grandchildren. We provide estimates of how large those costs would be for the average person under official economic assumptions and demographic projections.
The Drug Benefit Will Cost More Than We Were Told
In November 2003, the Congressional Budget Office estimated that the federal cost of the Medicare drug benefit amounts to $409 billion through 2013. Many observers, including us, thought that estimate could greatly understate actual outlays under a Medicare drug benefit. First, the new program has many novel features, requiring the CBO to make numerous assumptions about how the complex policies proposed in the two bills would actually work. The resulting estimate is highly uncertain, and there is nothing in the legislation that would hold spending to $400 billion. Second, the CBO estimate cannot account for future legislation that could substantially increase program outlays.
Estimates from the Centers for Medicare and Medicaid Services are much higher than those from the CBO. The president's fiscal year 2005 budget, released in February 2004, indicated that the MMA would cost federal taxpayers $534 billion through 2013. The following month, the Medicare trustees clarified that report. Medicare Part D (the drug subsidy program) would require general revenue infusions of $690 billion over the next decade. In September 2004 the president's budget office increased the spending estimate by another $42 billion, bringing the total (federal and state) cost to $732 billion through 2014. Thus, within one year of passage, the estimated cost of drug coverage grew by 86 percent. The latest budget for fiscal year 2006 escalates the spending on prescription drug coverage to $1.2 trillion--an increase of $470 billion, simply from moving the budget window ahead by 1 year! Now the 10-year net cost through 2015 is estimated at $720 billion. This indicates the hazard of making a long-term policy on the basis of a short-term evaluation of its cost.
Clearly, the new drug subsidy law poses an unusually difficult challenge in making cost estimates. It is a major departure from the way Medicare has operated in the past. Many of the features of the new benefit are unique, and the empirical evidence used to develop the cost estimate is more limited than usual. Even the $720 billion estimate may severely underpredict the actual net cost of the program. The new drug plan may be less effective at containing costs than is assumed in making the estimates; more employers than expected may drop retiree drug coverage under their plans; newer drugs may be more expensive than assumed; more retirees may enroll in the drug program than assumed; enrollees may demand more drug treatments than assumed; physicians may begin prescribing more drugs to seniors than assumed; consolidation in the pharmaceutical industry may lead to less competition, greater monopoly power, and higher drug prices. In all of those cases, federal spending would come in substantially larger than estimated.
Actual outlays may also be driven well above current estimates by future legislative changes that cannot be included in official cost projections. When the full drug subsidy becomes effective in 2006, many seniors will find that the benefit does not cover a substantial portion of their drug costs. They are likely to continue pressing for a more generous subsidy--and policymakers are likely to accommodate such demands. Sen. Edward Kennedy (D-MA) is already on record that the current drug subsidy represents a mere down payment on adequate drug coverage for seniors.
Some Democrats in Congress have proposed a more generous program that would increase outlays by $1 trillion in total over the next decade. If policymakers fill some of the significant gaps in coverage in the current subsidy over the next few years, the drug benefit enacted in 2003 could conceivably cost twice its latest cost estimate of $720 billion.
The history of the gap between budget estimates and actual Medicare costs does not leave one sanguine about likely future outcomes. Actual Medicare spending has usually come in much higher than initial projections, reflecting changes that could not be anticipated in the behavior of patients and providers, changes in the practice of medicine, and changes in legislation. For example, in 1965 the actuaries estimated that hospital insurance under Medicare would cost about $9 billion annually by 1990. Actual spending was about $67 billion in that year.
History will most likely repeat itself. Congress has awarded a drug benefit that today appears to cost $720 billion on net over 10 years. The actual cost over the next 10 years could easily be substantially higher.
Medicare Drug Spending Must Be Controlled
House Republicans proposed to limit the overall cost of the Medicare drug benefit to prevent spending from skyrocketing, but that provision was ultimately dropped from the final bill. Although spending could be arbitrarily limited, such a measure could have unintended consequences unless considerable flexibility and consumer choice are built into the program. Poorly designed spending limits could cause beneficiaries to shift from prescription drugs to other, less-restricted forms of treatment, which could raise overall Medicare costs and worsen health outcomes.
Perhaps the most direct way to limit program cost is to cap total spending for prescription drugs in each year. Spending in excess of the cap in one year could be "recouped" the following year by reducing the federal prescription drug subsidy. Plans would pass that reduction on to beneficiaries by raising premiums or reducing the benefit (through increased cost-sharing requirements and tighter drug formularies). Some people have suggested that cost sharing could be automatically increased for middle- and upper-income beneficiaries, while those with low incomes were protected from cost increases.
That general approach could be effective in reining in the government's prescription drug costs. But the exclusive focus on prescription drugs overlooks the broader fiscal problem facing Medicare. Even without a new drug benefit, Medicare spending is rising at an unsustainably high rate. The Medicare program has been growing substantially faster than the economy. Ignoring the early years of Medicare's existence (when rapid growth of outlays is only natural), Medicare outlays have grown more than twice as fast as the nation's GDP in real terms. Calculations based on the CBO's historical data suggest that since 1980 Medicare outlays have grown about 2.2 percentage points faster than GDP. The baby boomers, who begin to turn age 65 in seven years, will place unprecedented new demands on Medicare and will likely cause the difference in the two growth rates to widen during coming decades.
Moreover, if the new drug benefit were singled out as a target for spending controls, we would run the risk of biasing treatment decisions. Medical practitioners and patients would opt for other, less-regulated treatment procedures that would frustrate attempts to slow Medicare spending without harming patient welfare. Rather than trying to control the costs of individual health services, as Medicare has tried unsuccessfully to do over the past four decades, we should integrate prescription drugs with all other benefits. Our proper concern should be the overall growth in health spending. Micromanaging the cost of individual services has proven to be both ineffective and inefficient.
The traditional Medicare program is an uncapped entitlement to payment for health care providers. The payment incentives foster ever-expanding spending for services that, at the margin, are not worth what they cost. A new approach, modeled after the premium support system of the Federal Employees Health Benefits Program, is needed if we are to control costs sensibly. By offering beneficiaries real choice among health plans, we can promote competition that over time can slow the growth of spending and improve the value of what we purchase.
The new law takes a step in that direction. MMA introduces a new approach to competition that places the traditional Medicare program and the private plans on a more even footing, at least on an experimental basis. Beginning in 2010 in selected market areas, all plans (including traditional Medicare) would bid against the same benchmark. Beneficiaries would pay lower premiums if they enrolled in plans with below-average costs. Those choosing traditional Medicare might face higher or lower premiums, depending on whether the traditional program is effective at containing cost.
That premium support provision in MMA is authorized as a six-year demonstration project. If the project is implemented, it will begin to move us toward premium support and fair competition among all the health plans in Medicare. However, the project is not likely to be undertaken. A similar approach was attempted during the late 1990s and failed because of grassroots opposition from local health providers and politicians. Their concern was that competition might, in fact, be effective in limiting federal reimbursements. The MMA provision has already met resistance of the same sort from several members of Congress, who have declared "not in my back yard" six years before the start of the demonstration.
Despite the opposition, the issue of premium support is not dead. Medicare's deteriorating financial condition will reopen the question of whether market competition or government control is more effective in bringing the rising cost of the program in line with our willingness and ability to pay.
Medicare's Long-Term Funding Shortfall Is Enormous
Before the drug law was passed, the political debate focused on the $400 billion taxpayer cost projected by the CBO for the next 10 years. Few policymakers at that time anticipated that the estimated cost of the new benefit would escalate so soon after the law's enactment. The latest 10-year net cost estimate published by the president's budget office is $720 billion--80 percent larger than the initial estimate--and there is no reason to think that is the last word on the subject.
The drug subsidy makes a permanent commitment to a new Medicare benefit, but there is no explicit commitment of resources to pay the cost. That means reliance on general revenue transfers--in other words, increasing the federal budget deficit.
There are substantial political and economic risks in legislating a permanent expansion of Medicare benefits without good estimates of its future cost. The option of cutting back coverage may not be available if we come up short on resources. That would mean cutting other federal programs or imposing huge and self-defeating tax burdens on future generations. As the baby boomers retire and enroll in Medicare, that dire scenario is likely to play out. If effective cost containment measures are not added, future generations will face massive bills to support the more generous Medicare entitlement.
The Medicare trustees estimate that the present value of the government's unfunded obligation attributable to the drug benefit is $16.6 trillion. That figure understates the problem since it is based on an optimistically low assumption about future cost growth. The estimate assumes that the per beneficiary growth in drug spending remains only 1 percent greater than the growth of GDP per capita through 2080, with spending growth assumed to slow thereafter.
Medicare spending has grown at substantially higher rates over its 40-year history. Program spending on a per capita basis has grown about twice as fast as GDP per capita since 1980. Using a growth rate in line with historical experience through 2040 and assuming much slower growth in spending subsequently, the government's unfunded liability amounts to $21.9 trillion. That is about double the size of the U.S. economy this year.
As large as those figures are, they do not capture the full magnitude of Medicare's financial crisis. In addition to the new drug benefit, Medicare Parts A and B are also underfunded. According to the trustees, the entire Medicare program falls short by some $61.6 trillion in today's dollars. By comparison, Social Security is a model of fiscal rectitude, with unfunded obligations of about $10.4 trillion in present value.
Those figures demonstrate the seriousness of our financing challenges for Medicare based on the current structure of that program. If a future Congress enacts effective reforms, especially stronger incentives for competition, the funding shortfall could be substantially reduced.
Huge Long-Term Cost Will Be Borne by Younger Generations
Although the MMA does not address how the drug benefit will be financed, ultimately someone will have to pay the bills. Congress will no doubt try to reduce overall Medicare spending in future budget deals, once outlays grow so large that they can no longer be financed through additional debt creation. Then the choice will be to scale back the program or tolerate wider-ranging cuts in other programs. Or Congress could raise taxes to prevent the political fallout from burgeoning deficits. However, that is sure to cause slower economic growth as workers scale back their market labor supply.
Tax hikes appear to be the most probable outcome, simply because once the baby boomers retire, they will constitute a potent political force seeking to preserve and expand their entitlement to Medicare benefits.
Suppose we were to enact a new permanent tax on wages sufficient to pay for the additional benefits. How large would it have to be to fully fund the projected outlay increase due to the new drug benefit? Let us assume that such a tax is levied immediately. Today's seniors--who are already out of the labor force--would pay little of the new taxes. Measured in today's dollars, the value of drug benefits net of taxes per capita for those aged 59 and older would be $22,500 ($23,600 in additional benefits and $1,000 in additional taxes). Baby boomers--76 million Americans aged between 40 and 58--would shoulder a little more of the cost, but they would come out significantly ahead. The average boomer would receive $35,500 in benefits (measured in today's dollars) and would pay about $17,300 in new taxes--a net gain of $18,200 in present value.
The losers are 162 million young people and all future generations. Those born between 1966 and 1990 would lose, on net, $2,498 each. Those born between 1991 and 2004 would lose $4,190 each. Thus, children, who obviously do not get to vote, and very young workers, who tend to not vote in large numbers, would pay the most.
Congress is unlikely to consider ways to pay for the new Medicare outlays until there is a fiscal crisis brought on by rising budget deficits. The longer Congress waits to pay its bills, the more those costs will be shifted from older generations to younger ones. Delaying the tax increase by 10 years, for example, reduces the tax burden on boomers by about $10,800 per person. Very young workers and today's children could also gain because they would escape higher taxes for 10 years. However, their gain would be offset by the need to raise taxes even higher subsequently to finance today's drug coverage costs that are paid for by borrowing. Future generations would unambiguously lose by a larger amount because a 10-year delay in action on reforming Medicare compels higher future tax increases.
The Medicare prescription drug benefit is likely to cost much more and do less good than previously anticipated. Federal outlays will almost certainly grow faster over the next decade than current estimates indicate, and the added liability that will be incurred over the long term is enormous. Despite all of that spending, Congress will hear plenty of complaints over the coming years that the benefit is not good enough.
Rationalizing Medicare benefits to include prescription drugs may well have made sense, but Congress erred in failing to address how that benefit would be financed. As a result, financial pressures on the program have been increased significantly, moving us that much closer to crisis.
The Medicare Modernization Act included a number of reform steps intended to promote competition among health plans and improve the efficiency with which health care is delivered to America's seniors. But those measures are unlikely to change the basic character of the program, which remains firmly rooted in the insurance market of 1965 and top-down control by Congress. If Medicare is to survive the onslaught of the baby boomers, more thorough reform is needed.
A fiscal crisis in Medicare and, by implication, for the federal government, can be averted, but not without a great deal of effort. The policy objective should be to moderate the growth in health costs without putting an undue burden on the recipients. The Federal Employees Health Benefits Program, which has served federal employees and retirees well for 45 years, presents a practical model that could be adapted to the special circumstances of the Medicare population.
There are risks in undertaking such a reform. But one fact is clear: we do not have the option of the status quo. We can either take prudent actions to make Medicare sustainable, or we can accept the program's inevitable decline at the hands of political inaction--a decline that may not be limited to Medicare and may adversely affect other economic sectors as well.
1. See 2004 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, March 23, 2004, Table II.C22, p. 108. Hereafter, 2004 Trustees' Report. This is the "infinite horizon" estimate, which computes the present value of the funding shortfall assuming that Part D continues in perpetuity.
2. See ibid., Table II.B11, p. 60, and Table II.C16, p. 99. Part A is underfunded by $21.8 trillion and Part B is underfunded by $23.2 trillion.
3. See Douglas Holtz-Eakin, Letter to Chairman Thomas on the cost estimate for H.R. 1 (Medicare Prescription Drug, Improvement and Modernization Act of 2003), November 20, 2003, http:// www.cbo.gov/showdoc.cfm?index=4808&sequence =0; see also Congressional Budget Office, The Budget and Economic Outlook: Fiscal Years 2005 to 2014 (Washington: CBO, January 2004), pp. 12–13. The CBO estimates that the Medicare drug benefit will increase federal outlays by $409 billion between 2004 and 2013; MMA, which also includes some provisions to reduce Medicare spending, will increase federal outlays by $395 billion.
4. See Budget of the United States Government, Fiscal Year 2005, and FY06, Summary Table S-13, http:// www.whitehouse.gov/omb/budget/fy2006/bud get.html. The drug benefit alone was estimated to cost $649 billion, according to the 2004 Trustees' Report, Table IV.A1, p. 157.
5. See ibid., Table IV.F5, p. 190. The $690 billion figure is the sum over 10 years of the general revenue transfers column in that table.
6. The federal component is $576 billion. Part of the remainder is state spending on people covered for prescription drugs under Medicaid. However, a part would be new spending on the same population of dual eligibles if Medicare coverage turns out to be more generous than their existing coverage.
7. Ceci Connolly and Mike Allen, "Medicare Drug Benefit May Cost $1.2 Trillion; Estimate Dwarfs Bush's Original Price Tag," Washington Post, February 9, 2005, p. A01.
8. Section 257(b) of the Deficit Control Act of 1985 prescribes that for federal revenues and mandatory spending, baseline projections by official budget reporting agencies must assume that current laws will continue without change. Because the laws that govern revenues and mandatory spending are, for the most part, permanent, the baseline projections reflect only anticipated changes in the economy, demographics, and other relevant factors that affect the implementation of those laws. See testimony of former CBO director Dan L. Crippen before the House Committee on the Budget, May 2, 2002, http://www.cbo.gov/showdoc.cfm?index=3384&sequence=0.
9. Quotes from Sen. Edward Kennedy's interview on CNN's "Inside Politics," June 18, 2003: 7 "There's enormous need for prescription drugs, a need for the senior citizens of this country. Costs are too high. This is only $400 billion. The seniors are going to spend $1.7 trillion over the next 10 years. We're only providing $400 billion. That's only 22 percent. I'd like to do much better. . . . But this is going to be a down payment. And one thing is going to be for sure. When we get this as a down payment, we're going to come back again and again and again and fight to make sure that we have a good program."
10. Robert J. Myers, Actuarial Cost Estimates and Summary of Provisions of the Old-Age, Survivors, and Disability Insurance System as Modified by the Social Security Amendments of 1965 and Actuarial Cost Estimates and Summary of Provisions of the Hospital Insurance and Supplementary Medical Insurance Systems as Established by such Act, U.S. House of Representatives, Committee on Ways and Means, Committee Print, July 30, 1965, Table 11, p. 33.
11. Medicare outlays came in at $34 billion in 1980. In 2003 they amounted to $274.2 billion. GDP figures in those two years were $2,732 billion and $10,829 billion, respectively. Using continuous compounding to calculate the implied annualized growth rates, Medicare spending grew at 6.5 percent whereas GDP grew at 4.3 percent. The annual inflation rate during the same period was 2.4 percent. That means that real GDP grew at 1.9 percent per year whereas real Medicare outlays grew at 4.1 percent per year--more than twice as fast in real terms. This statement remains true even if we deflate Medicare outlays with an index of inflation in health care goods and services instead of general inflation between 1980 and 2003. The budget data are available at http://www.cbo.gov/showdoc.cfm?index=1821&sequence=0, tables 9 and 11. Inflation indices are taken from the Bureau of Labor Statistics' consumer price index for urban wage and clerical workers.
12. The trustees assume that the 1 percentage point growth rate wedge will be eliminated linearly over 20 years starting in 2080--that is, reduced by 5 percent each year so that Medicare expenditures per capita grow no faster than GDP per capita after the year 2100. Note that total Medicare expenditures could grow faster as a result of faster growth of the retiree population relative to growth of the working-aged population. However, this source of faster growth in total Medicare outlays is eliminated in the very long run as the age distribution of the U.S. population is projected to stabilize after 2100.
13. We assume that growth in Medicare expenditures will continue at historical rates. Because this rapid growth rate generates more intense pressure on the federal budget, we assume that Congress will enact effective spending restrictions in 2040 rather than 2080, as assumed by the trustees. Thereafter, Medicare expenditures per capita are assumed to grow at the same rate as GDP per capita. Nevertheless, the initial faster growth of Medicare expenditures yields an even higher estimate of the infinite horizon financial shortfall, as reported in the text.
14. See 2004 Trustees' Report, Table II.B11, p. 60, and Table II.C16, p. 99. Part A is underfunded by $21.8 trillion, and Part B is underfunded by $23.2 trillion.
15. See 2004 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds, Table IV.B8, p. 60.
Joseph Antos is the Wilson H. Taylor Scholar in Health Care and Retirement Policy at AEI. Jagadeesh Gokhale is a senior fellow at the Cato Institute. An earlier version of this analysis was presented in testimony before the House Subcommittee on Human Rights and Wellness on July 17, 2003.