Cato Policy Analysis No. 305                                       April 29, 1998

palogo16.gif (10981 bytes)

by Peter J. Ferrara

Peter J. Ferrara is general counsel and chief economist with Americans for
Tax Reform and an associate policy analyst at the Cato Institute.

Executive Summary

The Balanced Budget Act of 1997 made extensive Medicare reforms to delay the impending financial collapse of the system for a few years. But those changes do not sufficiently address the long-term problems of the program. In fact, Medicare's long-term problems remain deep and intractable. Even after the 1997 reforms, by the time today's young workers retire, Medicare's current sources of funding will likely be sufficient to finance only 50 percent or less of the promised benefits.

Such an enormous financial gap cannot be closed by raising taxes and cutting benefits, which would greatly harm working people and retirees. The only solution is to reform the system by taking advantage of the efficiencies, incentives, competition, and productivity of the private sector.

Every retiree should be free to use his share of Medicare funds to purchase private health insurance. In addition, every worker should be free to put his Medicare taxes into a personal health savings account rather than the Medicare program, using the accumulated funds to pay for health care and health insurance during retirement. Because of the high returns that result from investing in private capital markets, those privately invested accounts would be able to pay significantly higher benefits than Medicare.

Essentially, these proposed reforms would shift control over Medicare from Washington to individual retirees and workers across the country. That Medicare reform would ensure that workers and retirees achieve more freedom and prosperity than are possible under the current system.


Those politicians who have not enjoyed the Medicare debate over the last few years should consider this: the fun has just begun. The problem is not just that Medicare's funding will be insufficient to pay promised benefits in a few years. Over the long run, the program's financial gaps become truly enormous. By the time today's young workers retire, Medicare's current sources of funding will likely be sufficient to finance only about 50 percent or less of the promised benefits. Consequently, without fundamental reform of the system, U.S. politics over the next two generations will be dominated by battles over draconian tax increases or draconian benefit cuts for Medicare.

Attempting to salvage the current system by raising taxes and cutting benefits will greatly harm workers and retirees. Quite simply, the program cannot be saved by those means. The solution lies in reforming the system to take advantage of the efficiencies, competition, and productivity of the private sector. This study will show how that can be done and how it is possible to close the long-term financial gaps of the system.

The Medicare System

Medicare consists of two components--Part A, or Hospital Insurance (HI), and Part B, or Supplemental Medical Insurance (SMI). HI pays for hospital care for up to 90 days for each spell of illness, plus an additional 60 lifetime reserve days of hospitalization that each beneficiary can use once. Medicare recipients must pay a deductible of $760, a $190 coinsurance fee after the first 60 days of care, and a $380 coinsurance fee for each of the lifetime reserve days. The deductible and coinsurance fees are indexed to increase with health costs each year. HI also pays for up to 100 days of care in a skilled nursing facility after hospital discharge and hospice care.

HI benefits are financed through part of the payroll tax. The HI tax rate is 2.9 percent, which is assessed against all wages without limit. That tax is assessed half against the employer and half against the employee. But the employer's share effectively comes out of wages that would otherwise go to the worker, meaning that workers bear the entire tax burden.

Medicare Part B, or SMI, pays for doctors' services, diagnostic tests, certain medical supplies and equipment, and some forms of home health care. Those benefits are subject to a fixed deductible of $100 per year not indexed to health costs and to a coinsurance fee of 20 percent of costs after the deductible. Health care providers can also charge patients up to 15 percent more than the maximum Medicare-approved fees.

SMI is financed in part by a monthly premium, currently $43.80 per month, paid by each retiree. That amounts to $525.60 per year for each retiree, or $1,051.20 annually for a retired couple. Those premiums currently cover only about 25 percent of program costs. General tax revenues finance the remaining expenditures.

Medicare does not pay for long-term nursing home care, custodial home health care, outpatient prescription drugs, dental care, eyeglasses, hearing aids, or similar items.

About three-fourths of the elderly have private medical coverage that supplements Medicare.(1) At a minimum such insurance covers the Medicare coinsurance fees, hospital costs after Medicare coverage runs out, and, in some cases, prescription drugs and other services not covered by Medicare. Private insurance is also available for nursing home care, but most of the elderly have not purchased it.

Medicare expenditures are projected to total $218.8 billion in fiscal year 1998, accounting for about 13 percent of the entire federal budget.(2)

The 1997 Reforms

The 1997 balanced budget agreement included a number of Medicare reforms, some good, some ill-conceived.

Private Options

The most important reform adopted in the 1997 budget act has been mostly overlooked. Retirees are now allowed to use their Medicare funds to buy private health insurance coverage from a full range of private alternatives to Medicare. Private alternatives include health maintenance organizations (HMOs), medical savings accounts (MSAs), preferred provider organizations, provider-sponsored organizations, and traditional fee-for-service insurance plans. Those private plans generally would enable retirees to get better services and treatment at a better price than Medicare provides.

Cost Controls

The most significant cost reductions in the 1997 reforms result from simply reducing prospective payments to doctors and hospitals for providing services under Medicare. But, as discussed below, Medicare payments were already well below market rates for most services, and inadequate reimbursements threaten the quality and availability of health care for the elderly.

Private Contracting

A provision in the 1997 act effectively prohibits retirees from using their own money to buy the health care they want. The provision states that doctors and hospitals who contract privately with Medicare beneficiaries for Medicare-covered services will be banned from providing Medicare services for two years. As a result, retirees would have great difficulty finding doctors who would be willing to provide such services.

That provision greatly restricts the freedom of the elderly to control their own health care. Suppose a retiree wants a service or treatment that Medicare won't pay for, or suppose the doctor or hospital is unwilling to provide a particular service or treatment because of Medicare's low reimbursement rates. The new provision effectively prevents a retiree from using his own funds to pay for care that may be critical to restoring his health.

That policy makes no sense, and it is also financially detrimental to Medicare. When retirees choose to forgo Medicare payments altogether and pay for services or treatment out of their own pockets, they save Medicare the expense of their care. By preventing retirees from making that choice, the new provision increases Medicare costs, in addition to reducing health care freedom, control, and quality for the elderly.

Major Benefit Changes

The 1997 budget act shifted home health benefits that had been provided by Medicare Part A to Part B. By 2004, that change will shift about $24 billion per year in spending from Part A to Part B. That is a sham that simply increases the tax burden on general taxpayers.

As discussed above, Part A is financed by the payroll tax. If payroll tax funds are insufficient to pay promised benefits, then the program is unable to pay those benefits. But 75 percent of Part B is financed by general revenues. If more money is needed to finance Part B benefits, then the federal government simply devotes more general revenue funds to the program. Therefore, only Part A is subject to the threat of bankruptcy in the sense of being unable to pay promised benefits.

By shifting Part A benefits to Part B, Congress was able to delay the date on which Part A would be unable to pay benefits. Delaying the date was, indeed, the most important consideration in making the switch. But that move does nothing to reduce the overall financial burden of Medicare. It just increases the drain on general taxpayers due to Part B. As will be shown below, without reform the 75 percent general revenue contribution to Part B will grow in 12 years to an overwhelming burden that will massively increase the deficit.

The bill also expanded Medicare coverage for several preventive services, such as mammograms, pap smears, and prostate and colorectal cancer screening tests.

Medicare Premiums

The 1997 act also requires the Part B premiums paid by the elderly to continue to finance 25 percent of Part B expenses each year. The premiums, consequently, will automatically increase each year to the level sufficient to cover that portion of the program's expenses. Part B premiums had varied in recent years and covered as much as 33 percent of Part B expenses, but the proportion was slated to decline below 25 percent and continue to fall in future years.

The Problems of Medicare

Despite the euphoria of last year's bipartisan budget agreement, Medicare still faces financial disaster. Medicare Part A, which pays hospital bills, will be running a deficit again in 2004. The Medicare trust fund, which is more an internal set of IOUs than a real asset, will be exhausted by 2010. At the same time, Medicare Part B, which pays for physician care, is spending far more than it takes in. In fact, by 2010 the combined deficit for Medicare Parts A and B will be more than $150 billion! And that doesn't even begin to account for the baby-boom retirements to come.

The bigger problem, however, is that over the long run the financial gap in Medicare becomes overwhelming. Under the intermediate projections in the Medicare trustees' reports, paying all promised Part A benefits to today's young workers when they retire would require increasing today's Medicare payroll tax rate about 2.5 times, from 2.9 percent to well over 7 percent (see Table 1).

Even with that tax increase, the general-revenue-financed deficit in Medicare Part B would continue to soar. With premiums paid by the elderly covering 25 percent of Part B expenditures, by 2015 the remaining Part B deficit alone would climb to about $150 billion in 1997 dollars. By 2040 that deficit would reach more than $340 billion (see Table 2). Without tax increase to cover Part A, the total Medicare deficit would grow to over $200 billion in 2015, over $350 billion in 2025, and almost $600 billion in 2040, all in 1997 dollars. The total federal deficit would be increased each year by the amounts shown in Table 2.

In addition, the premiums that cover a portion of Part B expenses will become an even larger burden on the elderly.

Table 1
Payroll Tax Rates Needed to Finance Promised Medicare Part A Benefits (intermediate assumptions)

Year Percentage Year Percentage
2010 3.66 2040 7.10
2015 4.19 2045 7.32
2020 4.85 2050 7.46
2025 5.54 2055 7.59
2030 6.21 2060 7.78
2035 6.75

Source: Calculated from data provided by the Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Hospital Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); and 1997 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington: Government Printing Office, April 24, 1997), Table 1.

Table 2
Projected Medicare Deficits (intermediate assumptions) in Billions of Constant 1997 Dollars

Year Medicare Part A Deficitsa Medicare Part B Deficitsb Increase in Total Federal Deficit due to Medicare
2005 3 97 100
2010 25 126 151
2015 62 154 216
2020 119 189 308
2025 132 226 358
2030 177 267 444
2035 219 307 526
2040 254 341 595
2045 283 371 654
2050 307 398 705
2055 330 425 755
2060 361 457 818

Source: Calculated from data provided by the Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Hospital Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Supplementary Medical Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); and 1997 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington: Government Printing Office, April 24, 1997).
aProjected deficits in Part A assuming the payroll tax continues at 2.9 percent.
bAnnual general revenue contribution to Medicare Part B each year, assuming that premiums paid by retirees continue to finance no more than 25 percent of annual Part B expenditures.

Under intermediate assumptions, the Health Care Financing Administration projects that in 2007 premiums per retiree will soar to $104 per month compared to $43.80 today, an increase of almost 250 percent. That amounts to almost $1,250 per year for each retiree or about $2,500 for each retired couple. That burden will continue to increase along with the expenses of Medicare in future years.

Furthermore, the intermediate assumptions in the trustees' reports may be overoptimistic. As former Social Security chief actuary A. Haeworth Robertson has noted, in the past actual experience has more often been closer to the trustees' so-called pessimistic projections.

Analysis of the key assumptions underlying the projections in the trustees' reports also indicates that future experience may be closer to the pessimistic assumptions. For example, a key assumption for the future financial prospects of Medicare is the rate of growth of real wages. With Part A financed by a payroll tax, the growth of wages over time will determine how much Part A revenues will grow. The intermediate projections in the latest trustees' reports assume that real wages will grow at about 0.9 percent per year over the next 75 years. But real wages have in fact grown at less than half that rate for the past 25 years, according to the Social Security Administration's own data.(3) The so-called pessimistic projections, however, assume that real wages will continue growing at about the same rate as they have since 1970.

Another critical assumption is the rate of increase in life expectancy at retirement. The longer the elderly live in retirement, the higher Medicare expenditures will be. The so-called intermediate projections assume that the rate of increase in life expectancy will slow over the next 75 years.(4) The pessimistic projections assume that life expectancy over the next 75 years will continue to grow at about the same rate as it has for the past 50 years (though it assumes that the increase for men and women will be much closer than in the past).(5) Indeed, the pessimistic projections may even understate the problem. Given high-tech advances already developing in genetics, biotechnology, and other medical fields, the increase in life expectancy may well accelerate over the next 75 years rather than decline.

Finally, the intermediate projections assume a substantial reduction over the next 75 years in the rate of increase in health costs. Even the pessimistic projections assume some reduction in the growth of health costs despite the fact that health costs have been steadily increasing.

Under the pessimistic projections, Medicare's future financing problems are twice as bad. Paying all promised Part A benefits to today's young workers when they retire would require increasing today's Medicare payroll tax about fivefold, to well over 14 percent (see Table 3).

Moreover, even with that tax increase, the Medicare Part B deficit would soar to about $100 billion by 2005 and to more than $200 billion by 2020. By 2040 the Part B deficit alone would be well over $400 billion (see Table 4). Without the tax increase for Part A, the total Medicare deficit would reach almost $200 billion by 2010, $400 billion in 2020, and over $900 billion in 2040.

Table 3
Payroll Tax Rates Needed to Finance Promised Medicare Part A Benefits (pessimistic assumptions)

Year Rate (percentage) Year Rate (percentage)
2010 4.61 2040 13.52
2015 5.72 2045 14.13
2020 7.19 2050 14.38
2025 8.99 2055 14.60
2030 10.86 2060 14.95

Source: Calculated from data provided by the Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Hospital Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); and 1997 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington: Government Printing Office, April 24, 1997).

That assumes, of course, that premiums paid by the elderly continue to cover 25 percent of Part B expenditures. If that is so, premiums in retirement for those entering the workforce today will be growing about three times as fast as their retirement incomes.

Clearly, the financial problems of Medicare cannot be solved with current revenue sources. Attempting to do so would bankrupt workers, the federal government, and the elderly. Moreover, with current revenue sources financing only about one-half of projected expenditures under intermediate assumptions, or about one-third or less under pessimistic assumptions, cutting benefits to match revenues would effectively end the program.

The financial problems of Medicare are already hurting the elderly. To rein in the program's runaway costs, the government has increasingly restricted what it will pay doctors and hospitals for services under Medicare. According to the Medicare Payment Advisory Commission, the government's Medicare fees cover only about 70 percent of the cost of providing care and services to the elderly.(6) As a result, many doctors and hospitals are refusing to treat retirees under Medicare or are providing lower cost, lower

quality services to compensate for the government's low payments. Consequently, the elderly suffer lower quality care. That problem will only get worse under the 1997 reforms, which further reduce Medicare reimbursement payments.

Table 4
Projected Medicare Deficits (pessimistic assumptions) in Billions of Constant 1997 Dollars

Year Medicare Part A Deficita Medicare Part B Deficitb Increase in Total Federal Deficit due to Medicare
2005 32 101 133
2010 66 128 194
2015 116 165 281
2020 183 213 396
2025 267 270 537
2030 356 331 687
2035 436 385 821
2040 492 423 915
2045 523 445 968
2050 534 454 988
2055 543 461 1004
2060 556 470 1026

Source: Calculated from data provided by the Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Hospital Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); Health Care Financing Administration, 1997 Annual Report of the Board of Trustees of the Federal Supplementary Medical Insurance Trust Fund (Washington: Government Printing Office, April 24, 1997); and 1997 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington: Government Printing Office, April 24, 1997).
aThese are the projected deficits in Part A assuming the payroll tax continues at 2.9 percent.
bThis amount is the annual general revenue contribution to Medicare Part B each year, assuming that premiums paid by retirees continue to finance no more than 25 percent of the annual Part B expenditures.

In addition, under the Medicare payment system, doctors and hospitals receive a standard fee for their services regardless of the quality of care. Since the lowest quality providers receive the same payments as the best providers, there is no economic incentive to provide service above the lowest quality care acceptable. Moreover, the Medicare payment system pays hospitals the same standard fee for each illness regardless of costs. That was intended to give hospitals incentives to control costs; however, it creates incentives for hospitals to maximize their net income by discharging patients earlier, regardless of their health conditions, and otherwise to cut corners whenever they can.

The government is also slow to approve new medical technologies and treatments for coverage under Medicare. While the delay reduces Medicare expenditures in the short term, retirees lose access to the latest, best treatments and care.

Finally, if retirees are unhappy with the quality of care they get under Medicare, they are now effectively prohibited from going outside the system with their own money to buy the services and care they want, because of the provision on private contracting.

Clearly, fundamental reform of Medicare, as well as other entitlement programs, is inevitable.

Reforming Medicare

Effective Medicare reform would include two components. One component would allow retirees to use their Medicare funds to pay for private-sector coverage. The second would allow workers to save and invest their tax payments in private-sector accounts and use the funds accumulated by retirement to finance their health care. Those components could be adopted separately, together, and in either order.

The Option for the Elderly

The 1997 reforms already began to adopt an option for the elderly to choose private coverage, as discussed above. But now that option must be restructured to fit within the revenue available to the system. That can be done as follows.

Available Resources. First, the revenues that can reasonably be made available for Medicare should be determined. The Medicare payroll tax rate would remain 2.9 percent of wages. At the same time, Medicare premiums should be indexed to grow at the rate of growth in incomes for the elderly--no faster. That would protect retirees from an increased financial burden. Finally, the general revenue contribution to Medicare Part B should be fixed to grow at no more than the rate of general economic growth.

Freedom to Choose Private Alternatives. Medicare then must be redesigned to fit within the available resources. The first step is to allow each retiree to choose to have health coverage provided through private alternatives to Medicare, as under the 1997 reforms. Each retiree should be free to choose from the full range of such private alternatives. Those include MSAs, HMOs, preferred-provider organizations, traditional fee-for-service insurers, provider service networks, and plans offered by associations such as the American Association of Retired Persons, unions, or past or present employers. Any private alternatives left out by the 1997 reforms should be added. MSAs in particular should be available without limits.

As under the 1997 reforms, each retiree should be free to use his share of Medicare funds to purchase a private alternative of his choosing. But the key is to add the restriction that the total amount of Medicare funds that would be available to all retirees each year to purchase those private alternatives would be limited to the amount of available revenues as described above. Each retiree would then be able to use his share of the available funds for the private alternatives. Then, to the extent retirees chose private alternatives, total Medicare expenditures would be limited to the amount of available resources. Thus there would be no financial crisis in Medicare.

The amount of Medicare funds that would be available to each retiree for private alternatives would be risk adjusted. That means each retiree's share would vary depending on age, health status, and geographical location to reflect the expected variation in health costs among retirees and what insurers would be expected to charge to cover them. Retirees who were older and sicker would receive more from Medicare to buy their private coverage. Retirees who were younger and healthier would receive less from Medicare as their lower expected health costs would result in lower charges for their private coverage.

No system can predict exactly what the health costs of each retiree will be, but that is not necessary. All that is necessary is that the risk adjustments be close enough that the private market can function, and that adverse selection against Medicare (with the higher cost retirees staying in Medicare and the lower cost retirees leaving) can be avoided. The long experience with cost variations under Medicare, and the additional experience that would be gained with the private options each year, would make such risk adjustments quite feasible. The 1997 reforms in fact provide for the Health Care Financing Administration to adopt such risk adjustment.

Each private plan offered as an alternative to Medicare would have to accept all Medicare beneficiaries who chose that plan, as a condition of participation in the new system. The private plans consequently could not sign up only the healthiest, lowest cost retirees and leave the sicker, higher cost retirees in Medicare. Insurers that did not want to submit to this condition would not have to, or be allowed to, offer their plans as private alternatives to Medicare.

The private plans would have to offer at least the same core benefits as Medicare. They could charge additional premiums beyond the amounts that each retiree would receive from Medicare to buy the private options, as long as those additional premiums reflected expected costs under standard insurance pricing principles. However, those additional premium charges would have to be standard for all retirees under Medicare and could not vary on the basis of the health or other characteristics of each retiree, except perhaps for geographic location. The private plans would receive varied payments to compensate for varied health risks through the risk-adjusted payments from Medicare as described above. That means that all retirees would face the same supplemental premium charges for a particular private health plan, regardless of health condition, again except perhaps for geographic variations. At the same time, the private plan would not be limited to receiving the same compensation for each retiree no matter how costly or sick because the risk-adjusted payments from Medicare would be varied to compensate for such differing costs.

The private plans could offer additional benefits not offered by Medicare for additional premium charges. Those additional charges would not have to be the same for all retirees but could vary on the basis of expected health costs. That is justified because the plans would not be receiving risk-adjusted payments from Medicare reflecting varied costs for these benefits.

The new system would include an open season each year during which retirees could switch from one private plan to another if they desired. If a private plan cost less than the amounts to be paid from Medicare for the retiree, the retiree would keep the difference as a cash rebate. That would provide a strong incentive for retirees to choose the lowest cost plan and enable retirees to share the benefits of controlled costs.

A Continuing Medicare Option. A key question in this new system is what happens to retirees who do not choose the private alternatives. Retirees should be assured that they can stay in Medicare if they desire. But they cannot have the option of staying in the current Medicare system without change because that system is hopelessly bankrupt. Moreover, if retirees can choose a Medicare option that spends far more per beneficiary than do the private alternatives, then over time retirees will be forced to choose to stay in the public system with unlimited spending, and Medicare's financial crisis will continue unresolved.

This problem can be solved by providing that no more will be spent on average for those beneficiaries who choose to stay in Medicare than Medicare will pay on average for those who choose the private alternatives. That will ensure that total Medicare spending will stay within the available resources described above. Thus, the Medicare financing crisis will be completely solved, regardless of whether retirees choose Medicare or the private alternatives. Moreover, these reforms would put the choice between Medicare and the private alternatives on a level playing field, enabling the system to benefit from the advantages of competition and private alternatives.

Keeping Medicare spending per beneficiary within the above limits will require some restructuring of Medicare benefits. It is difficult to determine how that can be done in a way that will be acceptable to future retirees, and the answer is likely to vary among retirees. Consequently, the best approach is probably to offer retirees a range of options within the Medicare program. Retirees could be offered the following Medicare options:

None of those continuing Medicare options would be particularly appealing, as they all involve placing a substantial burden on retirees. But this is the best that can be done within the bankrupt Medicare framework if the program is to become solvent. The key to the reform is that these continuing Medicare options are just a back-up to assure people that they will have a chance to continue to rely on the public system if they desire. But retirees would be likely to choose the private alternatives. Many of those private alternatives have powerful cost control mechanisms so that desired benefits can be provided for much less than Medicare. As a result, through private alternatives, retirees can continue to receive the benefits now provided by Medicare for many years with little if any additional burden. Indeed, many private alternatives will offer better benefits than Medicare.

Medical Savings Accounts. The most appealing private option is probably MSAs. With an MSA, instead of all health funds going to an insurance company, only a modest portion is paid to an insurer for catastrophic insurance, which typically covers all bills over a high deductible like $3,000 per year. The rest of the funds would be paid into an individual account for each retiree or worker. The covered patient could then use the funds in the account to pay medical bills below the deductible amount. Moreover, a patient could use funds in his MSA for any medical services or treatments he needed. Whatever account funds the patient did not spend on health care could be withdrawn at the end of the year and used for any purpose, or saved for future use.

With MSAs, therefore, patients would effectively be spending for noncatastrophic health care funds that they could otherwise keep for any use. As a result, they would have full market incentives to control costs for such care. They would consequently seek to avoid unnecessary care or tests, look for doctors and hospitals that would provide good quality care at the best prices, and consider whether the health care or service was worth the cost. That in turn would stimulate true cost competition among doctors and hospitals. Since consumers would be choosing among them on the basis of cost as well as quality, they would compete to minimize costs as well as to maximize quality.

That is quite different from what happens with traditional insurance. Generally, when an insurer pays the bills, neither the patient nor the doctor is concerned about costs. The patient consequently does not seek or request lower medical costs. And, when the patient is unconcerned about costs, doctors and hospitals have no reason to keep costs low. The result is rapidly escalating health costs.

In 1995 the National Center for Policy Analysis conducted a study of how MSAs would work under Medicare, with the assistance of a major actuarial firm.(7) The study showed that with the funds Medicare would pay, retirees could choose an MSA that works as follows:

Those figures are for insurance under which there is no restriction on the doctors or hospitals the patient can choose, or the needed services or treatments he may receive. Even less expensive would be a managed care alternative under which the retiree can choose from an established network of doctors and hospitals organized to reduce costs. This network would provide the care when costs exceeded $3,000 per year, while the retiree would still choose any doctor or hospital and needed service or treatment for costs below $3,000. With a cost-reducing network, the MSA would work as follows:

These MSA plans would actually provide better benefits than Medicare in several ways:

Medicare, by contrast, has no cap on out-of-pocket expenses. A retiree under the program can be liable for tens of thousands of dollars in expenses each year, even for services that are covered by Medicare. That is why about 70 percent of the elderly buy Medigap insurance that will cost them an average of $1,200 in premiums this year.

Indeed, with the MSA, the retiree would not need the Medigap insurance and could contribute the $1,200 instead to the MSA. That would virtually eliminate any other out-of-pocket expenses for services covered by Medicare. The MSA would have close to or even more than $3,000 put in it each year to pay for expenses below $3,000, and the insurance would pay for all expenses over $3,000.

MSAs would be better than Medicare for sicker retirees as well as healthier ones. Sicker retirees would prefer the catastrophic coverage and cap on out-of-pocket expenses provided by MSAs. They would value the ability to use the MSA funds for any health service or treatment they chose, including services not covered by Medicare. They would also value the freedom from Medicare's restrictive payment system, which would result in greater freedom of choice and better quality care.

The same is true for lower income retirees. Lower income retirees would greatly prefer the catastrophic coverage and cap on out-of-pocket expenses provided by MSAs. They would also greatly prefer the availability of MSA funds for services not covered by Medicare and the cash benefits provided by MSAs.

While providing such superior, highly attractive benefits, MSAs can also be expected to reduce costs by about 30 percent, based on experience, data, and analysis. With such reduced costs, the funds Medicare would pay for retirees to be covered by MSAs would be sufficient for several years, perhaps 10 to 15, to pay for the MSA coverage with little or no additional cost to retirees, especially considering the financial benefits of MSAs discussed above. Avoiding additional burdens on retirees after that time would require the second reform component, the option for workers described below.

Other types of private health plans can also reduce costs and improve benefits compared to Medicare. HMO cost controls can perhaps reduce costs as much as MSAs, although HMO bureaucrats, rather than doctors and patients, effectively decide what care should be given. HMOs have been offered previously as an option for a limited number of Medicare recipients and have generally provided benefits that traditional Medicare does not, such as prescription drugs and eyeglasses. Networked plans, such as preferred-provider organizations, are also able to offer cost controls by negotiating discounts with providers in exchange for a regular volume of covered patients.

Figure 1 shows that, regardless of the reason, private health insurance plans have been more effective than Medicare in controlling costs throughout the 1990s. Despite the fact that Medicare traditionally reimburses providers at a lower rate than private insurance, the average per patient cost of private insurance is now lower than that of Medicare.(8)

Protecting the Poor. For retirees in or near poverty, the government could offer additional funds for the purchase of private alternatives. Those additional funds would be sufficient to ensure that low-income retirees could purchase the most cost-effective private plans with no significant out-of-pocket costs. The supplemental funds could also be available to those who chose to stay in Medicare. That would ensure that low-income retirees would not suffer any burden under the reform.

If desired, supplemental benefits could be financed by means testing the highest income retirees. Wealthier retirees would then receive reduced amounts from Medicare for the private alternatives and would supplement those Medicare benefits with their own funds. Such a means test would have to apply in an equivalent way to the continuing Medicare options described above.

Little or no such supplements would be necessary for the first 10 to 15 years after reform, as Medicare funds would probably be sufficient to pay for private alternatives during that time. If the second component of the reform was also adopted, the advantages of that component would probably obviate the need for such supplements over the long term. Still, such a provision creates a reassuring safety net.

Given that the alternatives to this reform plan would be massive tax increases or massive benefit cuts, or both, this plan is highly attractive for retirees, workers, and the nation's economy.

Figure 1
Medicare vs. Private-Sector Costs (average per patient cost)

img.gif (4342 bytes)







Source: Council for Affordable Health Insurance.

The Option for Workers

Medicare's long-term financial shortfall is so huge that even the cost reductions envisioned as a result of the reforms discussed above would not be enough to eliminate the projected future deficits. As a result, the amount Medicare could pay for the private alternatives would eventually not be enough to cover the costs of even the most efficient alternatives such as MSAs. Retirees would then have to add more and more of their own funds over time to pay for even the least costly alternatives.

That is why the second component of the reform plan is necessary. Under this component, workers would be allowed to choose to save and invest the Medicare taxes assessed on them and their employers in personal investment accounts. Those accounts would, by retirement, provide the funds to finance the best private alternatives with no additional burden on retirees.

Medicare currently operates on a pay-as-you-go basis, which means that the funds paid into the program are not saved and invested for future benefits. Rather, those funds are immediately paid out to finance benefits for today's retirees. The future benefits of today's workers when they retire are supposed to be paid with taxes assessed on those working at that time.

Since little investment is made through this pay-as-you-go system, essentially no investment returns are earned to help finance future benefits.(9) If we shifted to a fully funded private system instead, workers would be saving and investing their current tax payments in the private sector, and, by retirement, the investment returns would accumulate to huge amounts that could be used to finance health coverage.

Under this second reform component, then, workers could choose to direct the full 2.9 percent of their Medicare tax, including both the employee and employer shares, into a personal investment account called a Health Bank IRA. The same rules, regulations, and restrictions would apply to the private accounts as apply to IRAs today, except no withdrawals would be allowed before retirement.

Investment of the funds in each worker's Health Bank IRA would be conducted through a system of private professional investment management firms. Firms would apply to the government to be recognized as designated investment management firms eligible to handle investment of Health Bank funds. Workers would then choose among the designated firms to handle investment of their accounts. The firms would choose the investments for the workers' accounts. Workers could switch from one designated firm to another on short notice.

This system would make investing Health Bank funds easy for inexperienced investors. They would need only to choose an investment management firm. The firm would then choose the investments. This system would also help protect workers from fraudulent operators, as only established firms that received government designation could invest Health Bank funds.

At retirement, the Health Bank funds would finance an annuity that would pay a specified amount each year for the rest of the worker's life. Those amounts could be adjusted so that they would increase with inflation over time. The annual annuity payments would then be used to pay for the private health coverage alternatives created by the first component of the reform. The retiree could choose whatever private alternative he preferred, including MSAs.

Workers already in the workforce at the time of the reform will have already paid substantial taxes into the current Medicare system for all their working years. If they chose the private Health Bank option, the government would pay into their accounts recognition bonds compensating them for past taxes. The amount of the bonds would be set so that with interest they would pay a proportion of future Medicare benefits equal to the proportion of lifetime Medicare taxes the worker and employers had paid. At retirement, Medicare would pay each year that proportion of what it would have otherwise paid for the private alternatives.

Studies calculating the benefits that could be paid by a similar private replacement for Social Security indicate how much benefit can be derived from such Health Banks.(10) Assume that over a worker's career, Health Bank investments earn a real rate of return of 4 percent, which is just half the average return earned in the stock market over the last 70 years. A conservative reading of the studies suggests that at such a return, a private investment account like a Health Bank IRA would pay three times the benefits of a pay-as-you-go system like Medicare. Indeed, most of the recent studies suggest the private investment system can pay even more.(11)

Since the payroll tax funding Medicare Part A would be used for the Health Bank, a 4 percent real return on the funds accumulated in the Health Bank would pay about three times the benefits payable by Medicare Part A. In other words, the annual annuity amount payable through the Health Bank would be about three times what Medicare Part A could spend per beneficiary under the current payroll tax. If in retirement the worker chose the most efficient private alternative such as MSAs, then the costs for the benefits paid by both Part A and Part B would be reduced 30 percent. The combination of the effects of both components of the reform would be roughly sufficient to eliminate the Medicare financing gap under the intermediate projections. In other words, with both components of the reform adopted, under the intermediate projections for Medicare, the funds payable through the Health Banks should be roughly sufficient to pay the costs of private MSAs providing retirement health benefits in place of Medicare, with little or no additional cost burden to the retiree.

The reform would be sufficient to close most of the long-term financing gap under the pessimistic projections for Medicare as well. If retirees are going to be living substantially longer, as assumed under those projections, then it would also be reasonable to delay retirement for a few years. The Health Bank funds would also generate substantial tax revenues in addition to the returns earned directly by the investor. That revenue would come from taxes assessed at the business level on the full before-tax return to capital, before the returns to the investor were paid. As discussed further below, those additional revenues would offset the transition costs of the Health Bank proposal in less than 25 years and start to generate substantial surpluses, which would grow to huge amounts by the time today's young workers retire. These eventual surpluses could be paid to retirees for private health coverage to close any remaining gap under the pessimistic assumptions, perhaps on a means-tested basis focusing the funds on low- and moderate-income retirees.

Those two further elements, delaying the eligible recipient age and additional revenues from the private investments, would likely be sufficient to eliminate the remaining financing gap under the pessimistic assumptions. But there is an additional possible source of new funds that would more than close the gap. If the entire Social Security system were privatized, the additional benefits paid from the private invested system would be more than enough to cover the remaining Medicare financial gap under the pessimistic assumptions.

Work by Professor Thomas B. Saving of Texas A&M University, who has recently endorsed similar reforms, also shows how the Medicare crisis could be avoided as a result.(12) Saving calculates the results under the new reform for a worker entering the workforce at age 22. Under that system, this worker in retirement would be covered by a private health insurance policy, with an annual deductible of $2,500, that would cover all health expenses exceeding $2,500 each year. The worker would also pay no premiums in retirement for this policy, unlike the current Medicare system under which premiums are already $1,000 per year for a retired couple and will grow over time to much more.

Saving calculates under the equivalent of intermediate assumptions, and with only a 3 percent real return on investment, that the private investment accounts would generate enough funds to finance the above-described health insurance in retirement, with only 1.15 percent of annual earnings invested in the account. Even under pessimistic assumptions, the account would generate enough funds to finance the insurance policy described above with no more than the 2.9 percent of earnings taken by the current Medicare tax.

The government can assure all workers with Health Banks a guaranteed minimum benefit, supplementing what the Health Banks would pay with additional payments each year to the extent necessary to achieve the minimum. The minimum benefit would ensure a basic level of support for retirement health expenses in any event. Yet, with the likely performance of the private investment system providing much higher benefits than Medicare, government expenditures for the minimum benefit are likely to be minimal. The public can consequently be assured that no one will be left without essential assistance for retirement health spending, while workers and the nation will still be able to enjoy the freedom and better benefits offered by the private system.

During the transition to the new system, the government would lose the payroll tax revenues it currently uses to finance Medicare Part A benefits, to the extent workers chose the private Health Bank option. If about half of all workers exercised the option in the first year, this revenue loss would total about $50 billion. To pay benefits to current retirees, the government would have to reduce other expenditures by an equivalent amount or find other sources of financing. Sen. Phil Gramm (R-Tex.) has suggested using revenues from the forthcoming tobacco settlement.

However, over time the government would receive an important new source of revenue to offset the loss of payroll taxes. The new savings and investment in the Health Banks would generate the full before-tax return to capital. Substantial taxes on that return would be paid by the business investing the funds before the interest, dividends, or other returns were paid back to the original Health Bank investors. As the Health Bank funds grow, this new revenue could be used to help pay the cost of transitioning to the new system.(13)

Whatever the financing mechanism, the transition to a new Medicare system will be far less costly than allowing the current system to continue.


The Medicare system cannot be saved by raising taxes and cutting benefits. That would impose too much harm on both workers and retirees. If allowed to continue on its present course, Medicare will ultimately bankrupt workers, the federal government, and the elderly.

The solution lies in reforming the system to take advantage of the efficiencies, incentives, competition, and productivity of the private sector. This study showed how that could be achieved through reform involving two basic components. First, retirees would each be allowed to use their share of Medicare funds to purchase more cost-effective private coverage that in many instances would provide better benefits than Medicare, as was begun in the 1997 reforms. Second, workers would be allowed to save and invest their Medicare funds through the private sector instead.

The cost-reducing effects of the private alternatives under the first component along with the increased resources resulting from the second component would be sufficient to eliminate the enormous long-term financial deficits of the current Medicare system, with little or no cost burden to retirees. Indeed, through the reform retirees would be guaranteed that Medicare premiums would grow no faster than their incomes, unlike the current system under which such premiums would grow enormously over time. Retirees would also be able to get better benefits through the private alternatives than are offered by Medicare. Retirees would enjoy as well greater freedom of choice, power, and control over their health and its care.


1. See Peter J. Ferrara, "Medicare and the Private Sector," Yale Law and Policy Review 6, no. 1 (1988): 65.

2. Budget of the United States Government, Fiscal Year 1999 (Washington: Government Printing Office, 1998), p. 2 and appendix, pp. 416-19.

3. 1997 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington: Government Printing Office, 1997), Table II.D.1.

4. Ibid., Table II.D.2.

5. Ibid.

6. Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (Washington: Government Printing Office, 1998), p. 2.

7. Peter J. Ferrara, John Goodman, and Merrill Matthews, "Saving the Medicare System with Medical Savings Accounts," National Center for Policy Analysis Policy Report no. 199, Dallas, September 1995.

8. Mark Litow et al., "What's Wrong with Medicare?" Council for Affordable Health Insurance Issue Brief, Arlington, Virginia, March 1997, pp. 16-17.

9. A pay-as-you-go system would be able to pay a real return equal to the rate at which real tax revenues grow over time. This return would be equal to the rate of growth of the workforce plus the rate of growth of real wages. Even this return would involve only a redistribution from others rather than net new income earned as in an invested, fully funded system. In any event, the real redistribution return from a pay-as-you-go system can be expected to be about 1 percent, at best. The before-tax real rate of return to capital, which would be generated by private investments in a fully funded system, is over 9 percent. See Martin Feldstein, "The Missing Piece in Policy Analysis: Social Security Reform," American Economic Review 86 (May 1996): 12.

10. Peter J. Ferrara, Social Security Rates of Return for Today's Young Workers (Washington: National Chamber Foundation, 1986); William G. Shipman, "Retiring with Dignity: Social Security v. Private Markets," Cato Institute Social Security Paper no. 2, August 14, 1995; and Marshall N. Carter and William G. Shipman, Promises to Keep: Saving Social Security's Dream (Washington: Regnery, 1996); and Robert Genetski, A Nation of Millionaires (Chicago: Heartland Institute, 1996).

11. Ibid.

12. Thomas B. Saving and Andrew J. Rettenmaier, "Privatizing Medicare: The Permanent Solution to the Crisis," Perspectives on Policy, Private Enterprise Research Center, Texas A&M University, Fall 1996.

13. This "revenue feedback" has been explored in depth in discussions of Social Security privatization. See Peter J. Ferrara, "A Plan for Privatizing Social Security," Cato Institute Social Security Paper no. 8, April 30, 1997. The effect would work in the same way for Health Bank accounts.

Published by the Cato Institute, Policy Analysis is a regular series evaluating government policies and offering proposals for reform. Nothing in Policy Analysis should be construed as necessarily reflecting the views of the Cato Institute or as an attempt to aid or hinder the passage of any bill before Congress.

Contact the Cato Institute for reprint permission.Printed copies of Policy Analysis are $6.00 each ($3.00 each for five or more). To order, or for a complete listing of available studies, write to: Cato Institute, 1000 Massachusetts Avenue NW, Washington, DC, 20001. (202)842-0200 FAX (202)842-3490 E-mail

| Hot Topics | Policy Analysis | Cato Home Page |