Medicare is the third largest program in the federal government after defense spending and Social Security. It will soon become the largest program, absorbing an ever-increasing share of the budget and national income. While the basic facts about Medicare are not new, what is new and encouraging is the growing recognition that Medicare is unaffordable in its current form and must be fundamentally reformed.

The Medicare trustees report that Medicare presents the nation with an $85.6 trillion financial hole.[1] This is the present value of Medicare's projected excess costs, [2] which reflect current and future subsidies provided under current law to Medicare beneficiaries for health insurance. For example, in 2007, the average Medicare enrollee received a benefit valued at $10,460, which included a subsidy of $4,053.

Medicare is a vital part of a federal social safety net, and it should be preserved, kept affordable for lower-income seniors, and be available to all seniors. However, this does not justify taxing workers and families to subsidize the health insurance premiums of higher-income seniors.

This observation suggests a policy of phasing out Medicare subsidies for upper-income beneficiaries. Such a phaseout would be good policy even if Medicare were fiscally sound, but it is even more important given Medicare's fiscal plight. On fiscal grounds and on fairness grounds, phasing out the subsidy for upper-income seniors should be among the first steps toward comprehensive Medicare reform.

There is no theoretical basis for determining the correct levels of income at which to begin and end the subsidy phaseout. Current law phases down the subsidy to the Part B premium through a process called "income relating." Specifically, the subsidy is phased down from 75 percent to 20 percent (or the premiums are phased up) for married couples with an adjusted gross income (AGI) between $170,000 and $426,000 in 2009.[3] Thus, an obvious option is to begin phasing out the subsidy at $170,000, applying the Part B phaseout rate to Parts A and D. Analysis suggests that phasing out the Medicare subsidy for upper-income seniors in this way would reduce Medicare's projected excess costs by almost half, [4] to just over $44 trillion.[5] Phasing out the subsidy to upper-income beneficiaries would be a first big step toward Medicare's long-run sustainability.

Defining the Fiscal Target for Reform

The projected excess costs of $85.6 trillion are enormous, but the trustees have acknowledged that the actual figure is significantly higher because of a faulty assumption about Medicare Part B outlays. Under current law, Part B physician fees are slated for dramatic cuts. As in previous years, Congress will almost certainly protect doctors' fees by passing "docfix" legislation. However, it is impossible to predict whether and to what extent Congress will allow some sustained increases in doctors' fees every year. To address the issue, the Medicare actuaries offer two illustrative alternative versions of docfix legislation: a freeze in payment rates and a more generous steady increase in payment rates. The two versions would increase Medicare's projected excess costs by $3.0 trillion and $5.9 trillion, respectively.[6]

Of the many aspects of Medicare needing significant reform, addressing the aggregate fiscal dimension is the most compelling, but Medicare reformers need a useful and meaningful target. Eliminating Medicare's projected excess costs may be ideal, but this is more than what is needed to create a sustainable program. Medicare could be made sustainable by establishing an acceptable and comparably more modest long-term ceiling for excess costs.

For example, Medicare's excess costs of $179 billion in 2007 were paid from the U.S. Treasury's general revenue. This large drain on the Treasury, while problematic, appears to have been manageable. It certainly elicited little apparent excitement in Congress or the public to reduce it. These observations suggest setting the 2007 level of excess costs and general fund support, measured as a share of gross domestic product (GDP), as the ceiling.[7]

Analysis indicates that adopting the 2007 level of general fund support as a ceiling for Medicare would lower the target for reform by about $22.2 trillion. The average of the two alternatives to compensate for the faulty assumption regarding physician payment rates suggests that Medicare's projected excess costs are about $4.4 trillion higher than the official estimates. The trustees' projected excess costs of $85.6 trillion plus the $4.4 trillion estimate for docfix legislation minus the implications of general fund support maintained at the 2007 level ($22.2 trillion) yields an intermediate sustainability target of about $67.8 trillion as of 2007.

The $67.8 trillion target would apply if Medicare reform had been achieved in 2007. Of course, the federal government failed to address Medicare's financing issues in 2007 or 2008 and the $67.8 trillion target grows each year, much as the outstanding principal on a financial note increases when no interest payments are made. Thus, this figure was somewhat higher in 2008, will be higher again in 2009, and higher still in 2010.

In effect, this suggests a target for reform, explains why the target is appropriate, and proposes a methodology for calculating the target. However, the exact value of the target will need to be re-estimated when Congress and the President finally accept responsibility for Medicare's financial dilemma and enact sufficient reforms to restore it to sustainability.

Cause and Consequence of Medicare's Predicament

Total Medicare income in 2007 reached $461.9 billion, while total expenditures were $431.5 billion (the difference between income and expenditures representing increases in the Part A and Part B trust fund balances).[8] This level of outlays absorbed 3.2 percent of the economy.

As Table 1 suggests, Medicare is, in part, a self-contained program with dedicated revenues funding defined benefits. The exception-an exception that is a growing problem-is that Medicare's excess costs are funded by drawing large and increasing amounts from the general fund of the U.S. Treasury. Essentially, Medicare claims a rising share of federal individual income, corporate income, and other non-payroll tax collections.

From another perspective, Medicare's draw on the general fund represents resources that are unavailable for other spending purposes-whether national defense, education, or infrastructure-or tax relief for America's families and American workers and businesses. Budgeting is about tradeoffs. By claiming such a large and increasing share of federal tax receipts, Medicare is forcing the federal government to spend less in other areas and impose higher taxes.

It has not always been this way. In 1985, Medicare drew only 0.4 percent of GDP ($17.9 billion) from the general fund to cover its excess costs. Ten years later, this had increased slightly to 0.5 percent of GDP ($37.0 billion). By 2007, Medicare's draw had more than doubled to 1.3 percent of GDP ($179 billion).[9] Nor is Medicare's draw rate stable. Going forward, Medicare's claim on general revenues is projected to grow rapidly, steadily increasing the burden on taxpayers and on federal spending choices.

Medicare demands increasing general revenue support because of rapid growth in outlays, not because projected revenue sources are withering. For example, payroll tax receipts for Part A are running just below 1.4 percent of GDP and are forecast to decline slightly to 1.3 percent of GDP by 2050. This decline is more than offset by premium income and state transfers[10] to Parts B and D, which are projected to increase from 0.44 percent of GDP to 1.33 percent of GDP by 2050 and to continue rising. In contrast to the projected slight decline in Part A income and the modest growth in Parts B and D income, outlays under all three parts of Medicare will grow profoundly as shown in Chart 1.

Such a mismatch between outlays and income and the commensurate drain on general revenues or increase in federal borrowing is simply unsustainable, as is made abundantly clear in Congressional Budget Office (CBO) correspondence to Representative Paul Ryan (R-WI), Ranking Member of the House Budget Committee.[11] Conceptually, Congress could maintain Medicare (and Social Security and Medicaid) on autopilot, financing the resulting budget deficits by issuing additional debt. As the CBO letter explained, this autopilot policy would have disastrous consequences.

According to CBO, if the entitlement growth was funded by issuing debt, the debt-to-GDP ratio would rise from about 37 percent today to more than 290 percent in 2050. That would be "a large figure by any standard." According to the CBO analysis using a textbook growth model, such an increase would cause per capita income to stagnate until the late 2040s when it would begin to contract.[12] Furthermore, "beyond 2060, projected deficits would become so large and unsustainable that the model cannot calculate their effects." Additionally, the CBO believes that "such estimates greatly understate the potential loss to economic growth under this scenario."[13] In short, issuing debt to cover the explosion in entitlement spending would be economically catastrophic.

Alternatively, Congress could theoretically increase taxes to fund growing entitlement spending, of which Medicare is the greatest part. However, such a steady increase in the tax share would weaken the U.S. economy, diminish wage growth, and undermine U.S. international competitiveness, while diminishing the growth in Medicare's revenue base.

However, according to the CBO, raising taxes to close the gap with entitlement spending is not a realistic solution. CBO analysis indicates that income tax rates would nearly double from current levels. For example, the 10 percent rate would need to increase to 19 percent, and the top current rates on individuals and corporations of 35 percent would need to increase to 66 percent. Yet those rates only apply if the higher taxes do not slow economic growth. In contrast, the CBO estimates that applying such tax rates would cause per capita income to drop by 5 percent to 20 percent. Such a decrease in income would mean that even higher tax rates would be needed to finance Medicare's shortfall, further shrinking the economy. The inevitable conclusion is that raising taxes to solve the entitlement funding problem is also out of the question.[14]

Medicare spending has often been reduced by cutting payments to health care providers (for example, doctors, hospitals, skilled nursing facilities, and long-term care facilities). While Medicare, like any private health insurance company, needs to be run well to avoid paying for unnecessary services and overpaying for necessary services, how much Medicare can save by eliminating "waste, fraud, and abuse" is severely limited relative to the savings needed for sustainability. Medicare should be run well for the benefit of its enrollees and as efficiently as possible, but Medicare's excess costs cannot be eliminated simply by tinkering with payment formulae. Medicare's excess cost problem is structural, not operational.

Keeping Medicare as is and financing the growing Medicare deficits with debt or higher taxes is not a viable option. Nor is simply slashing payments to health care providers. Policymakers need to find another way to make Medicare financially sustainable.

Medicare's Excess Costs

To better understand the nature of Medicare's structural excess cost problem, it is useful to step back from all of the complexities and consider basic structures. Medicare's original elements-Hospital Insurance (Part A) and Supplemental Medical Insurance (Part B)-were enacted into law in 1965. The Medicare drug benefit (Part D) was added as part of the Supplemental Medical Insurance program in 2004.

Part A: Hospital Insurance. Part A primarily provides health insurance coverage for inpatient care in hospitals, inpatient stays in skilled nursing facilities, and related services. It is largely funded by a 2.9 percent tax on wages, salaries, and related compensation to employees, half collected directly from wages and half subtracted from wages at the employer level.

By tradition, workers pay a portion of their labor earnings to Part A, giving the illusion that they are in some sense pre-funding their Part A retirement benefits. In fact, the taxes are not and never were set aside to cover the costs of future Medicare benefits, but instead offset the Part A costs incurred by current beneficiaries. In past years, some income would remain after paying these costs. This remaining income was then credited to the Medicare trust fund as a matter of bookkeeping, but in practice the extra income was added to general revenues and spent on other programs.[15]

Part A total outlays will begin to exceed total income around 2011 according to the Medicare trustees' report, and the gap is projected to grow steadily. Part A will have exhausted its trust fund by 2019, demonstrating that the Medicare payroll tax has long been inadequate to cover the costs of promised benefits.

Congress will need to legislate to keep Part A solvent past 2019. It could break with past practice and allow Part A to draw on the Treasury's general fund, into which corporate and individual income tax receipts and other revenue sources are deposited, or it could reduce Part A benefits to align costs with income. If Congress fails to act, Medicare will be forced to reduce outlays administratively to match revenues. For ease of discussion, the rest of this paper assumes that Congress will allow Medicare Part A to make up any shortfall by tapping the general fund, meaning that either other programs will be cut, taxes will be raised, or more government debt will be issued.

Parts B and D: Supplemental Medical Insurance. The Supplemental Medicare Insurance element of Medicare includes Parts B and D. Part B primarily covers physicians' fees and outpatient care. Current beneficiaries offset about one-fourth of Part B's costs through premiums, which for most beneficiaries are $96.40 per month in 2009. The balance is subsidized by drawing on the Treasury's general fund and miscellaneous sources, including interest on investments held in the Part B account.

Medicare Part D also provides a large subsidy when seniors purchase a drug benefit plan from Medicare-approved private companies. Part D premiums offset about one-fourth of the total cost, with the federal government subsidizing the balance by drawing on the Treasury's general fund. For 2007, the average Part D monthly premium was $27.39.[16]

Medicare's Total Excess Costs. Medicare's projected annual excess costs are the sum of the projected shortfall in Part A, which begins after 2019, plus Part B and Part D's ongoing draw on the general fund. In 2007, Medicare drew $179 billion from the general fund primarily to cover costs associated with Parts B and D. By 2017, this figure is expected to reach $353.3 billion.[17] In effect, seniors pay a portion of Medicare's total costs through their premiums, workers pay a portion of Medicare's costs through the payroll tax, and a large and growing subsidy is drawn from the pool of all other taxes paid to the Treasury.

The Subsidy in Seniors' Medicare

The Medicare trustees estimate that the average Medicare beneficiary received a benefit in 2007 valued at $10,460. Of this amount, $6,406 came from dedicated and organic income, including $4,346 from payroll taxes, $1,212 from premiums paid by the beneficiaries, and $849 from miscellaneous other sources specific to Medicare. The remaining $4,053-the excess costs per beneficiary-were financed from general revenues. In other words, participating Medicare beneficiaries received an average subsidy of $4,053 in 2007.

Medicare's dire financial conditions under current law reflect the projected rapid growth in per beneficiary subsidies paid to seniors: The average annual subsidy is projected to rise from $4,053 in 2007 to $6,067 in 2020 and to $19,512 in 2050.[18] At the same time, the number of seniors qualifying for Medicare will balloon with the coming retirement of the baby-boom generation.

America's seniors rely on Medicare to provide comprehensive, affordable health insurance coverage. However, Medicare is in dire need of reform and modernization, and its large and growing excess costs will eventually compel Congress to enact fundamental reforms.

The primary source of the pressure Medicare places on the federal budget and of Medicare's long-term unsustainability is the large and growing subsidy given to America's seniors through Medicare benefits. This subsidy is essential and appropriate for low-income and middle-income seniors, past and present. As a matter of principle, the subsidy will be necessary in the future for low-income seniors as Medicare remains an integral part of America's basic social safety net. In contrast, competing principles will come into play as the nation considers whether and how much to subsidize health insurance for middle-income seniors in the future.

The subsidy to Medicare beneficiaries is financed by general revenues and government debt, imposing further costs on future Americans. Therefore, this subsidy has significant, substantive consequences: Either working Americans will face higher taxes or less tax revenue will be available to finance other spending priorities. Faced with the realities of budget tradeoffs, it is an open question whether and to what extent this subsidy will remain appropriate for middle-income seniors. Congress will need to deal with this issue in future years as it further refines Medicare, informed in large part by Medicare's sustainability and the willingness of taxpayers to subsidize middle-income seniors' health insurance.

In sharp contrast, subsidizing health insurance for the wealthiest Americans is not now, has never been, and will never be appropriate, even if Medicare's finances are otherwise sound. Nothing justifies the government adding federal dollars to the wealth of the wealthy, whether as health care subsidies for seniors or for non-seniors, as agricultural subsidies, or some other form of quasi-welfare payment arising elsewhere in the vast array of federal programs. Those who can clearly afford to purchase their own health insurance ought to do so. No qualified senior should be denied the opportunity to purchase health insurance through Medicare. However, subsidizing the health insurance of upper-income seniors is no more appropriate than subsidizing their housing costs, fuel costs, food costs, or vacations.

Income-Based Phaseout of the Medicare Subsidy

With the passage of the Medicare Modernization Act, Congress took two noteworthy steps toward Medicare reform in 2004. The first step was the enactment of a new drug benefit (Part D). This reform reflected the growing importance of drug treatments in health care delivery. Regrettably, in an act of stunning irresponsibility Congress passed and the President signed this important new benefit into law without offsetting its costs with reductions in other Medicare benefits or otherwise paying for the new benefit. As a result, Medicare faces an additional $17.2 trillion in cumulative excess costs.[19]

The second noteworthy reform in 2004 was the partial phasedown of the subsidy in Medicare Part B. Specifically, in an important and encouraging break from the past, the Part B subsidy was tied to the beneficiaries' incomes as reported for federal income tax purposes. (This is commonly called "income relating.") Medicare beneficiaries who file their taxes as a married couple will continue to receive a 75 percent subsidy if their income is below a certain threshold. The subsidy percentage then declines in four steps as the beneficiaries' income rises, down to the minimum subsidy of 20 percent. Table 2 shows the income levels and corresponding subsidy levels for 2009.

In his fiscal year 2008 budget, President George W. Bush proposed a partial redress to the shortcoming in Part D financing. This proposal, which was also included in his 2009 budget, would partially phase down the Part D subsidy like the Part B subsidy was phased down. The CBO scored this proposal as saving $30 billion over 10 years when combined with a second proposal to eliminate inflation indexing of the income thresholds for the Part B phasedown.[20]

There is no theoretically correct income threshold to begin phasing down or, more reasonably, phasing out the Medicare subsidy for upper-income seniors. One consideration is to ensure that the phaseout affects only those who are truly sufficiently well-off financially to bear the additional expense. Medicare is a crucial part of the nation's social safety net. Reforms need to be carefully crafted so that they do not increase the financial burdens on low-income seniors.

Phasing out the Medicare subsidy for upper-income seniors would reduce the total projected excess cost in the program. For example, a baseline case for phasing out the Medicare subsidy is to begin with the current Part B phasedown rules. However, these rules only reduce the subsidy for upper-income seniors to 20 percent, so an obvious additional reform is to extend the phasedown until the subsidy is eliminated for the most well-off seniors and to apply it to all of Medicare. In 2009, this would have meant beginning to phase out the subsidy for couples with an adjusted gross income (AGI) above about $170,000 and eliminating the subsidy for couples with an AGI above $658,000. Under this baseline case, phasing out the subsidy for upper-income seniors would reduce the present value of excess costs from $85.6 trillion to $44.1 trillion-a reduction of $41.5 trillion.

A phaseout of the Medicare subsidy could be a serious financial burden for some middle-income seniors, but this baseline case for phasing out the subsidy would not affect middle-income seniors in a material way. Married seniors with incomes up to $320,000 would still receive a 50 percent subsidy rate, and couples would still receive some subsidy until their incomes exceeded $500,000 per year. These are very high income levels for seniors to be receiving subsidies.

As discussed above, a defensible intermediate target for Medicare sustainability is to hold the level of excess costs as a share of GDP to some fixed level, such as the 1.3 percent share in 2007. After adjusting for expected docfix legislation to avoid slashing doctors' fees, an intermediate target for reducing excess costs would then be $67.8 trillion. Paring excess costs by $41.5 trillion by phasing out the Medicare subsidy by extending current law would, therefore, eliminate more than 60 percent of the program's excess costs, leaving the balance to be eliminated through other policy changes.

This baseline phaseout is by no means the obvious policy choice. For example, a case could be made for beginning the phaseout at a somewhat lower level of income. A much stronger case could be made for phasing out the subsidy more rapidly. For example, there appears to be no justification for subsidizing Medicare benefits for seniors with annual incomes above $250,000. Changing the baseline-starting and ending the phaseout at somewhat lower income levels-would eliminate significantly more of Medicare's projected excess costs, leaving that much less to be achieved through other reforms.

Why Is a Phaseout So Effective?

Phasing out the Medicare subsidy for upper-income seniors can be a surprisingly powerful tool for returning Medicare's finances to long-term sustainability. The short explanation for this is in two parts.

First, the rapid growth in Medicare's excess costs occurs some years out as shown in Chart 2.

Second, the income levels at which the subsidy phaseout begins and ends are indexed to inflation in the estimations presented here, whereas AGI generally grows over time with nominal income.[21]

Consequently, for both reasons, the effect of the subsidy phaseout is initially quite modest when the budget pressures are relatively manageable, but grows steadily as excess costs would otherwise explode. In other words, the cost containment achieved by phasing out the subsidy becomes most effective just as the problem would otherwise become most unmanageable. This feature of the policy is especially important because it will give today's and tomorrow's workers more time to adjust their retirement plans accordingly.

Conclusion

Projected growth in Medicare spending shows the program is clearly unaffordable as currently structured. Yet it is a vital component of the nation's social safety net. Reforms to strengthen Medicare to return it to fiscal sustainability are therefore essential. A key feature of Medicare is that it provides heavily subsidized insurance to seniors, with an average subsidy of $4,053 per beneficiary in 2007. Such a subsidy is warranted for low-income seniors and questionable for middle-income seniors, but indefensible for upper-income seniors. These subsidies are financed by taxes on non-seniors-workers and their families, small businesses, and others- putting upward pressure on revenue needs and limiting the resources available for other governmental purposes. There is no economic or moral justification for wealth transfers to wealthy seniors.

There are many ways to phase out the subsidy for upper-income seniors, and no single income level for starting or ending the phaseout is theoretically superior. The details will need to be considered in the context of overall reform. However, using the existing phasedown thresholds of the Part B subsidy as a baseline is illustrative. Expanding this phasedown to all of Medicare would reduce Medicare's excess costs in present value terms by almost half ($41.5 trillion). In dollar terms, such a policy would achieve over 60 percent of the reforms needed to meet the proposed reasonable target for Medicare sustainability. Adjusting the phaseout parameters, such as starting and ending the phaseout at lower levels of senior incomes, would move Medicare even closer to sustainability. Phasing out Medicare subsidies for upper-income seniors is not only good policy in the abstract, it would be a significant step toward returning Medicare to fiscal sustainability.

J. D. Foster, Ph.D., is Norman B. Ture Senior Fellow in the Economics of Fiscal Policy in the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation. See also Medicare Path to Sustainability.

Appendix Methodology

Estimates of Medicare savings from phasing out the subsidy to seniors as their incomes rise derive from a model that combines projections of Medicare's draw on the Treasury's general fund and projections of the growth in the number of Medicare beneficiaries and their incomes. All relevant data and assumptions regarding Medicare itself are from the Medicare trustees' and Social Security trustees' annual reports for 2008. The data on the distribution of seniors by income class derive from the Census Bureau's March 2007 Current Population Survey for calendar year 2006.

Projections of total Medicare expenditures as a share of GDP are available in Table III.A2 of the Medicare trustees' report.[22] These data are available at five-year increments, while some other data are provided in 10-year increments. In both cases, values for the intervening years are derived using simple interpolation. Projections of nominal GDP are available in Table IV.A2.[23] Combining the two sets of projections yields projections of total nominal Medicare outlays. Once the Medicare Part A trust fund is exhausted, Medicare becomes fully a cash-flow system, but for a small amount needed to manage cash-flow anomalies from year to year. Consequently, total outlays for each year can be estimated using total revenues.

Table III.A4 provides intermediate estimates for Medicare's annual draw on the Treasury's general fund as a percent of total income.[24] These estimates, combined with the projections of total income, yield the total nominal general revenue support to Medicare each year.

A critical feature in present value analysis is determining the discount rates to use. The Social Security trustees' report provides the discount rates used for Social Security.[25] However, the Medicare trustees' report indicates that Medicare uses a slightly different stream of discount rates. These Medicare-specific discount rates can be inferred using the GDP projections and the Medicare trustees' estimates for the present value of GDP over the 75-year and infinite horizons.[26]

The combination of Medicare-specific discount rates and derived projections of Medicare's nominal draws on the general fund permit calculation of the present value of Medicare's projected excess costs. Comparing this calculation with the 75-year and infinite-horizon estimates in the Medicare trustees' report provides an important cross-check on this part of the methodology.

The Current Population Survey provides a distribution of households with a Medicare beneficiary by income class. This distribution must then be projected into future years. Projections of GDP equal those for gross domestic income (GDI), assuming either a constant, declining, or zero expected-value statistical error term. The issue then becomes whether seniors' share of GDI will rise or fall over time and whether the distribution of income accruing to seniors itself changes.

As the baby-boom generation retires, the average age of the U.S. population will increase significantly and permanently, strongly suggesting that the share of the nation's income and wealth held by seniors will likely also increase. This shift in income and wealth to seniors could also shift the distribution of seniors' income and wealth up the income scale. However, neither of these economic forces is the subject of this paper, so this paper assumes no change in seniors' share of GDI or the distribution of income accruing to seniors. This is a very conservative assumption. A more realistic assumption would certainly increase the size of the effects reported here.

Using these assumptions, it is then possible to calculate the average subsidy per beneficiary under current law for future years and the subsidy percentage that seniors would receive at different income levels in different years under a variety of subsidy phaseout regimes. The present value of these savings can then be summed to yield a total savings figure analogous to the present value of excess costs under current law.