How to Think About Medicare Reform

Report Health Care Reform

How to Think About Medicare Reform

November 22, 2011 16 min read Download Report
Stuart Butler

Abstract: With Medicare facing a bleak fiscal future, and because the program is a major factor in the enormous federal deficit and long-term debt, there needs to be structural reform. The way forward is to (1) require the program to operate under a real, long-term budget based on “premium support”; (2) add catastrophic protection; (3) convert fee-for-service Medicare into a premium-based program; and (4) systematically link premiums to income.

There is little argument that Medicare is in long-term financial trouble, even if the degree and nature of that trouble is debated. But there are two schools of thought about what to do about this problem. One school consists of what I would call the “Tweakers”—those who maintain that the essential structure of Medicare is sound and that the challenge is to find the right mixture of delivery system reforms, payment reforms, etc. that will get the program’s finances back on track without fundamentally altering the program.

And then there is the school of “Structural Reformers.” This school believes two things: first, that the scale of the financial problem, especially within the context of America’s structural financial imbalance, requires far more than tweaks, and second—and actually more important—that the very nature of the program and its promises are outdated and need to be put onto a different foundation, one that will make Medicare’s design and promises sounder and more appropriate than today’s aging, creaky program. Count me with this school of thought.

Medicare’s financial condition is quite staggering, as is its red-ink contribution to the country’s long-term fiscal picture. The Center for Medicare and Medicaid Services Office of the Actuary puts the program’s 75-year structural imbalance at $37 trillion. This means we would have to find and put aside, today and in cash, $37 trillion in an interest-bearing account to honor the promises Medicare makes to today’s and future retirees. Fixing that takes more than tweaks.

Medicare is also an entitlement and constitutes what Congress calls “mandatory” spending, meaning that the cost of the Medicare promise and other mandatory items has first claim on all revenues. Absent any reforms of that promise, money for “discretionary” programs, like defending the country or educating our children, comes out of whatever is left. Alarmingly, 58 per cent of the total federal budget now takes the form of this autopilot spending. Add in net interest and close to two-thirds of spending is now “uncontrollable”—a technical term now routinely used by budget analysts. Fixing that takes more than tweaks.

But attempts to deal with this structural financial problem often are greeted with almost incandescent rage by many seniors at town-hall meetings and in the ubiquitous AARP ads, which see reform proposals as tearing up the promise of Medicare. That’s perhaps understandable, given what I would call the Great Illusion that drives the politics of Medicare and which must be challenged and corrected as the prelude for charting out a new future for Medicare.

Among the myths that constitute the Great Illusion:

  • Myth: Seniors have paid for their Medicare thanks to their payroll taxes they paid.
    Reality: Medicare payroll taxes do not pay for all of Medicare, and they are not even intended to cover it fully. The payroll tax funds (inadequately) only Part A hospital coverage. Spending for that part of the program has exceeded income since 2008. Part B (principally physician costs) and Part D (the drug benefits) are simply voluntary insurance—heavily subsidized by general taxes—available to seniors. Nobody pays into these latter parts of Medicare with dedicated taxes during their working life, and reforming Parts B and D would therefore not alter something that has been “paid for.”
  • Myth: Medicare is financed by a real trust fund.
    Reality: Medicare is a pay-as-you-go system. Money comes in one door from various sources, including an increasing proportion from general tax revenue, and immediately goes out another door to pay for benefits coming due. No money has been put aside in anything resembling a savings account. In that sense, Medicare is really no different from any program that is financed by current taxpayers.
  • Myth: The payroll tax is a premium, and quite different from the income tax.
    Reality: Since the Medicare payroll tax, unlike the Social Security tax, does not have an income cap, it is actually indistinguishable from a regular income tax bracket. Moreover, as noted, it is quite unconnected to the total cost of Medicare. The Medicare tax on employees is thus actually nothing more than the first bracket in the federal income tax and applied to labor income.
  • Myth: Medicare is a classic social insurance program, and income-adjusting premiums or benefits would be a radical departure from the accepted vision of Medicare.
    Reality: Parts of Medicare are already income-adjusted, including the premiums for physician coverage (Part B) and drug coverage (Part D). Today, the standard premiums most seniors pay are set at 25 percent of the total premium cost. But for exactly the same benefits in Part B and Part D, upper-income seniors pay a range of premiums equal to 35 percent, 50 percent, 65 percent, or 80 percent of the total premium costs. So, for this year, the standard monthly Part B premium is $115.40, but among the wealthiest seniors (paying 80 percent of the premium), the monthly premium will amount to $369.10. For Part D, this year’s standard premium is estimated to be roughly $31.00; but for the highest income seniors, it will amount to $69.10. There are also special subsidies for low-income persons. The idea that Medicare is insurance related to seniors’ ability to pay for coverage is well established. It is nothing new.

The Way Forward

So as we think about how to get Medicare back on an affordable track, it is important to remember that the program today differs considerably from the image held by so many Americans. What we need to do now is to redesign it according to principles and strategies that actually improve the protections it provides seniors and at an affordable cost to them and to our children and grandchildren.

To do that involves four major steps.

1. Medicare should operate under a real budget, similar to other basic programs such as defense.

Needed: A level playing field among priorities. Medicare today assures a defined benefit consisting of particular services. It is also a “mandatory” item in the budget, so the cost of those services must be paid for by the government, whatever they amount to. For this reason, Medicare spending is merely projected, not budgeted in the sense that any family or business would understand the word “budget” as a limit or constraint. Spending on Medicare is the cumulative amount paid out in claims, and budgeteers guess what that will be for each year.

Contrast that with, say, defense or roads. For these “discretionary” programs, Congress votes a certain amount of money for a period, and the military or road builders have to work within that limit. True, they can come back and request more during the budgeted period, as the Department of Defense has done during the wars in Iraq and Afghanistan. But in these cases, Congress must make an explicit decision to change the spending limit that applies to defense, or for highways and bridges. The default is to stay within the budget. Not so with Medicare, where the default is to spend whatever the hospital, drug, and physician bills add up to.

This means that decisions among important national priorities are not made on the same level playing field in the current budget process. Entitlements, mandatory programs like Medicare and Social Security, have an automatic inside track over such priorities as education and defense. This is not the case in most other advanced democracies. In Britain, for instance, Parliament treats budgeting for the National Health Service essentially in the same way as other programs, and funding for the NHS must compete on the same level playing field as funds for other programs.

We need to do the same here.

Needed: Pressure to be more efficient. A real budget for government-funded health care is also necessary because budget pressure is needed to force serious action on the delivery system reforms that are broadly agreed as necessary. Without that budget pressure, it is difficult for doctors, hospitals, and health plans to make the tough decisions needed to increase efficiency. Even if tough decisions are made, without a budget, greater efficiency does not assure reduced total spending. The result may simply be more services, efficiently provided but of questionable additional value. From a medical point of view, that may be a marginal improvement, but it does not address the spending problem.

Constructing a real budget for Medicare, however, requires two particularly important considerations, each of which recognizes the nature of retirement health programs.

First, a real budget for Medicare needs to be established for a long period (perhaps 30 years) and reviewed every five years. Retirement requires careful long-term planning, and it is difficult for beneficiaries to make major adjustments when in retirement. Elements of some other programs also share this requirement of long-term planning, such as major highway infrastructure or defense weapons systems that may take many years to develop.

Recognizing this long-term planning feature of Medicare led me in 2008, as part of a bipartisan group of budget experts, to propose establishing a 30-year fixed budget for the program.[1] Such a budget could have specified dollar spending levels over the period. It could alternatively be indexed in some way in order to keep it in line with the price of some health-related benchmark, as discussed below. But the critical thing is that the budget for Medicare would in effect be the default setting over a long period, allowing seniors and the health sector to plan.

This long-term budget would replace the open-ended autopilot entitlement of today, in which the budget is little more than a guesstimate. To assure that the budget was on track to meet the long-term objectives of the program and yet continued to “compete” with other priorities on a level playing field, we recommended that while the 30-year budget should be reviewed every five years, the default setting could be adjusted with the agreement of the House, Senate, and President if conditions required a change.

Second, the budget should be distributed in the form of a defined contribution (or “premium support”). A limited budget has to be distributed in some way to pay for health services, and that is the second critical Medicare budget reform decision that has to be made. While the British NHS is budgeted, it takes the form of a “top-down” allocation of funds to hospitals and other groups of providers. This means that the critical decisions on who gets what care that are fundable under the budget limit are in the hands of the health industry in association with government officials responsible for allocating the budget. That process leads directly to explicit rationing and waiting lists.

The alternative approach to operating a budget is to allocate funds directly to Medicare beneficiaries so that the spending decisions are essentially “bottom up” and ultimately under the control of enrollees rather than health care or government officials. This is the methodology of the approach known loosely as “premium support.” In this model, a beneficiary would receive a certain level of financial support that is sufficient to defray all or part of the cost of a reasonable level of coverage, and the beneficiary would have the final control over how that money was spent.

For premium support in Medicare to operate successfully, the design would have to contain three critical elements:

A structured market. Enrollees in Medicare have a wide range of abilities to process technical health information and make informed decisions among alternative plans (and this is a particular concern about current fee-for-service Medicare). Thus a premium support approach is widely seen by its proponents as functioning within a structured market, with easily understandable information from reputable sources. There can be intense debate about what constitutes the best structure and information system. Some, like me, favor the relatively light regulation of the Federal Employees Health Benefits Program (FEHBP)—a form of premium support available to federal workers—with its standard comparative buying guide supplemented by a range of nonprofit and employee association guides. Others envision a more standardized market. But there is broad agreement that some degree of structured market is needed.

A risk-adjustment mechanism. There is also general agreement that a premium-support system must contain an adequately workable mechanism to reduce the incentive and ability of competing Medicare plans to select enrollees based on their medical risk—that is, “cherry picking.” This is far from an exact science, but it is a necessary feature of any system that seeks to provide choices among competing plans or providers. Medicare today, as well as the FEHBP and many employers who provide a defined contribution, already wrestles with this challenge. The steady improvements being made in risk adjustment are very important for premium support, as they are for any successful health system based on choice and competition.

An appropriate index for the government contribution. The third and generally more controversial feature of premium support approaches in recent years is how the amount of the government contribution should be allowed to grow over time—that is, the method used to index the government contribution.

Three Goals of Indexing. To think through the choice of index, it is import to recognize that an indexing mechanism for Medicare is intended to achieve some combination of three different goals—goals that are often in tension with each other:

  • Goal 1: Maintain a budget target for total spending. Premium support, as a publicly funded defined contribution system, can be a powerful tool to achieve a long-term budget goal.
  • Goal 2: Limit the financial risk to beneficiaries to combine protections with incentives to economize. A core objective of Medicare is to assure access to an adequate and affordable package of health services for seniors. For this reason, virtually all proposals include a mechanism to adjust the basic federal payment in some way to health risk factors, in addition to indexing it, in order that older and/or sicker individuals are not disproportionately burdened by out-of-pocket costs.

    But other financial risks also fit into this goal. We must also balance the financial risk for individuals with the financial risk to current and future taxpayers and also provide the incentive for beneficiaries to seek value for money. So the chosen method of indexing will affect the degree to which these combined risk-limitation goals are achieved.
  • Goal 3: Force a reasonable pace of adjustment by the health system. Premium support has long been seen as having a dynamic effect on the health care system. By limiting the financial support each beneficiary receives, it encourages Medicare enrollees to choose plans or seek services that are more likely to offer the best value for money and, in so doing, encourage delivery system changes that improve efficiency and help slow total costs.

    But the index rate must be in line with a feasible pace of change within the system. If the index is too low or tight, such that hospitals and other providers cannot improve their efficiency sufficiently quickly, there will be reductions of service and/or increased financial burdens on beneficiaries. If it is too generous, the spur to greater efficiency will be blunted.

In the current conversation about premium support, various advocates solve this three-part equation in different ways. Given the growing concern about federal deficits and the mounting unfunded obligations of the Medicare program, it is perhaps not surprising that Representative Paul Ryan (R–WI), chairman of the House Budget Committee, opts for a tight index (a growth rate of CPI). The Heritage Foundation’s Saving the American Dream proposal phases in a premium-support system earlier and more comprehensively than Ryan and uses a CPI+1 percent index,[2] while the Domenici–Rivlin Bipartisan Policy Center plan uses an index related to broad economic growth (GDP+1 percent) rather than an inflation-based index. Because under these proposals the government would make a per capita contribution, these indices encompass Medicare population growth.

The term “premium support” was actually coined in 1995 by Henry Aaron and Robert Reischauer, then both at the Brookings Institution.[3] They argued that the index should be anchored in some way to the actual cost of health care—in their version, the federal payment was indexed to the rate of increase of per capita spending on health care for the nonelderly. In a competing premium support–style proposal developed at the same time as the Aaron–Reischauer plan, Robert Moffit and I argued that the federal contribution should be adjusted to a weighted average of certain local Medicare plans.[4]

The idea of indexing the federal contribution to a benchmark cost of health care featured strongly in the early versions of premium support. That reflects the conditions of the mid-1990s, when concerns about the deficit picture and unfunded liabilities, while prominent, were not as acute as today. The emphasis of Medicare reform at that time was more on making the system operate more efficiently within a competitive health market. Hence, the goals of getting the right balance of risk and nudging the health system toward greater efficiency featured more prominently than the goal of spending control.

That said, advocates at the time recognized that budget control could be a factor in the design of an index. As Aaron and Reischauer noted in their plan, “If Congress found it necessary to reduce federal support for Medicare, it could slow payment increases, thus shifting costs to Medicare enrollees.”[5]

The deteriorating and alarming fiscal conditions since the 1990s are the reason for the shift we see in the balance between the three goals above in the indexes in current proposals (though, it should be noted, Aaron and Reischauer lament this rebalancing of their original premium-support model). While using any index based on CPI does implicitly reflect underlying health costs, an index of CPI or CPI+1—or, for that matter, GDP+1—would increase the federal contribution at a slower pace than the recent rise in health costs.

Yet, although such indexes would initially increase the financial risk for enrollees, their advocates make two arguments. One is that the fiscal situation is so dire that, in combination with other steps to protect more vulnerable seniors, older Americans with reasonable means should shoulder more costs in order to lighten the burden on younger Americans. The other is that delivery reform and efficiency improvement are proceeding more slowly than they could, and so greater budget pressure can and should be applied to speed the third goal above.

2. Medicare should be strengthened with catastrophic protection and thus become “real insurance.”

Not only has Medicare ceased to be traditional social insurance in any meaningful sense, but it also does not contain the central feature of real insurance—protection from financial catastrophe. While Medicare provides a wide range of benefits and parts of the program retain unrealistically low deductibles, it has no maximum stop-loss. Thus, seniors with Medicare can be bankrupted by medical costs unless they choose to buy additional Medigap or other supplemental coverage.

The reformed Medicare program of the future should fix this glaring defect in combination with reform of the program’s deductibles and copayments. This relatively small but important reform would help make the program the source of true insurance protection for seniors.

3. The traditional fee-for-service form of Medicare should be converted into a premium-based system.

Roughly three-quarters of today’s seniors are in the traditional fee-for-service (FFS) parts of Medicare as opposed to the Medicare Advantage integrated plans. This means that they pay premiums for parts of their coverage—the voluntary physician coverage (Part B) and drug coverage (part D)—while also receiving the fee-based hospital coverage in Part A. For the most part, the care is uncoordinated, and with such an open, unmanaged arrangement it is very difficult to spur the improvements in efficiency and coordination needed to achieve better value for money and a slower rise in total spending.

A reformed system should combine the disparate elements of FFS into a combined plan or plans with a single premium that reflects the full cost of services. Deductibles and copayments should be rationalized, and plan administrators should have greater discretion than today in establishing provider networks and more efficient fee structures for providers. As part of this element of reform, seniors would receive premium support to reduce the FFS premium, just as they would under premium support for managed care or other private plans.

In principle, the payments and service authorizations could continue to be centralized within the government, though that would be very unwise, or this part of Medicare could become a series of competing FFS plans under contract to the government but with greater organizational discretion, much as public charter schools operate within public school systems.[6]

4. Premium support should be systematically linked to the income of seniors.

Medicare already adjusts the contributions seniors make toward their benefits (separate from the income-rated Medicare payroll tax). But although this principle is now a central feature, its application is a mish-mash. There is no income adjustment for deductibles and copayments in Part A hospital coverage, so modest-income seniors in FFS can incur the same, yet potentially devastating, out-of-pocket costs as richer seniors (very low-income seniors on Medicaid do receive greater protection). There is some degree of income-related premiums for seniors who choose to obtain their Medicare benefits through integrated Medicare Advantage plans. Meanwhile, Part B is significantly income-related, as is Part D, but each with a quite different structure.

A reformed Medicare system should rationalize the income adjustment of premiums. Moreover, given the severity of the deficit and debt problem and the need to protect financially vulnerable seniors, the existing range of income adjustment in Medicare needs to be strengthened, with the most affluent seniors covering the full cost of their coverage.[7]


When Medicare was enacted in 1965, it was, like most major new programs, a political compromise. Its design features represented the requirements of that compromise. As the decades have gone past, the design weaknesses and internal contradictions of the program have become more evident. Patching the framework, or tweaking it at the edges, is not going to address the long-term weaknesses of the program or the enormous financial load it adds to the country’s structural financial problems. Basic reforms are needed, and these reforms can and should reflect the principal goals of the program as well as the principal goals of fiscal prudence.

Stuart M. Butler, Ph.D., is Director of the Center for Policy Innovation at The Heritage Foundation. This lecture is adapted from remarks delivered at the 2011 Medicare conference of America’s Health Insurance Plans in Washington, D.C.

[1] Joseph Antos et al., Taking Back Our Fiscal Future (Washington, D.C., The Heritage Foundation and The Brookings Institution, 2008), at (November 16, 2011).

[2] Stuart M. Butler, Alison Acosta Fraser, and William W. Beach, eds. Saving the American Dream: The Heritage Plan to Fix the Debt, Cut Spending, and Restore Prosperity, The Heritage Foundation, 2011, at (November 16, 2011).

[3] Henry J. Aaron and Robert D. Reischauer, “The Medicare Reform Debate: What Is the Next Step?” Health Affairs, Vol. 14, Issue 4 (1995), pp. 8–30, at (November 16, 2011).

[4] Stuart M. Butler and Robert E. Moffit, “The FEHBP as a Model for a New Medicare Program,” Health Affairs, Vol. 14, Issue 4 (1995), p. 52, at (November 16, 2011).

[5] Aaron and Reischauer, “Medicare Reform Debate,” p. 24.

[6] For a version of this model, see Saving the American Dream, p. 21.

[7] See ibid., p. 20.


Stuart Butler