Healing Medicare

Before enactment of Medicare in 1965, few elderly persons had reliable health insurance. When insurance was available, it was expensive and limited, and its renewal was uncertain. As a consequence, nearly 50 percent of the elderly had no health insurance at all, and faced bankruptcy from the costs of serious illness. Medicare provided all elderly Americans 65 or older with health coverage if they or their spouse had worked in a job subject to payroll taxation. As a reliable source of basic health insurance for the elderly, the Medicare program has been a tremendous success. Today, however, Medicare faces formidable challenges: an inadequate benefits package, an inefficient system of delivery, and a long-term budget gap.

In principle, getting rid of the inefficiencies would help expand coverage and reduce the budget shortfall. But how to achieve these goals is the subject of political and technical dispute.

Incomplete Coverage. Medicare was designed according to the medical and insurance practices of the mid-1960s. It excludes prescription drug coverage, "catastrophic" coverage (for conditions requiring long-term care), and coverage for many preventive services--all features of most private insurance plans. Medicare's deductibles and co-payments for inpatient hospital services are higher than most private plans. Compared to private health insurance in the United States, not to mention universal coverage abroad, Medicare's benefit package is parsimonious.

As a result of gaps in its benefits, Medicare covers only about 50 percent of total health care spending by the elderly. Less affluent households must devote more of their income to health spending than richer ones--or go without care. This regressivity is offset to some extent by Medicaid programs, which help low-income Americans cover their Medicare premiums, co-payments, and deductibles and buy prescription drugs. However, these programs are limited to the poorest Medicare beneficiaries, and coverage is incomplete.

Wasted Resources. Medicare's current structure, which relies on price controls to limit costs, fosters inefficiencies. Providers maximize income by encouraging utilization of services to compensate for their controlled prices; beneficiaries have weak financial incentives to use Medicare services in a cost-sensitive manner. Neither the Health Care Financing Administration (HCFA), the agency responsible for running Medicare, nor Congress can readily adjust the prices of tens of thousands of medical services to reflect rapidly changing market and technological conditions. So Medicare prices can diverge from market prices for long periods of time. Price controls also produce a strong inducement for fraud among Medicare suppliers. The free choice of provider, so popular with Medicare beneficiaries, is also an obstacle to the efficient management of care. Current law prevents HCFA from using such efficiency-enhancing measures as competitive bidding, selective contracting with preferred providers, and disease and case management techniques common to the private health care system. Legislative efforts to allow HCFA to use such techniques have been effectively blocked by supplier groups.

Medicare's inefficiencies are reflected in large and persistent regional differences in Medicare costs per beneficiary, differences that cannot be explained by living costs or demographics. At least one-quarter of the four-to-one regional differences in Medicare spending per capita is the result of differences in patterns of medical practice and utilization of medical services, with no observable effects on regional health outcomes. So reducing inefficiencies in the current Medicare system can slow spending growth without reducing the quality of care. The question is how--and how much?

A Budget Gap. Even with its meager benefits package and even after significant spending cuts mandated by the Balanced Budget Act of 1997, Medicare faces a long-term budget crunch. Medicare's financing difficulties reflect both the aging of the population and the growth of overall health care spending. By 2030, Medicare enrollment will double from 40 million to 80 million Americans. On average, medical outlays for persons aged 65 and older are nearly four times as large as outlays for persons aged 19 to 54. But changing demographics are not the main source of Medicare's projected growth. Rather, increases in per capita Medicare spending are driven by technological breakthroughs in medicine and their increasing utilization. Barring unforeseen--and highly unlikely--changes in medical technologies that significantly slow the growth of overall health care costs, the Medicare system will require additional revenues.

The Politics of Reform

The 1997 budget agreement mandated deep cuts in projected Medicare spending over 10 years mainly through the usual price-control mechanisms. The agreement also enhanced the ability of Medicare beneficiaries to choose among traditional fee-for-service Medicare, health maintenance organizations, and preferred-provider organizations, and broadened the number of experiments testing the applicability of such private-sector techniques as competitive bidding and case management to Medicare. In addition, the budget deal established a bipartisan commission to recommend financial and programmatic reforms to address Medicare's long-term challenges.

The commission split along partisan and philosophical lines. For a variety of reasons--most notably the absence of an immediate Medicare crisis--a supermajority of the commission's 17 members was unable to agree on recommendations, and the commission ended in the spring of 1999 without issuing a final report. Nonetheless, the commission's deliberations and the Clinton administration's July 1999 proposal in response are instructive for what they reveal about the likely direction of Medicare reform in the future. The major issues dividing the commission were drug benefits, the role of competition in a reformed Medicare program, and long-term strategies for reducing the budget gap.

Drug Benefits

Four out of five senior citizens are prescribed at least one daily drug treatment, and one in five takes five prescribed drugs per day. Per capita spending on drugs by the elderly is more than three times that of other adults and nearly 10 times that of children. On average, an elderly American spends more than $600 per year on drugs, with one in 10 spending more than $2,000 per year. Despite the obvious need for prescription drugs, about 35 percent of the elderly have no drug insurance. The remaining 65 percent obtain some drug coverage, much of it incomplete, through three channels: Medicaid (which affects only the 20 percent of the elderly living below the poverty line), supplemental insurance policies provided by former employers, and private insurance plans. The resulting patchwork system is riddled with inequities, inefficiencies, and excess administrative costs. Only about 30 percent of employers currently offer retiree health benefits, and that number is steadily declining. Annual premiums already exceed $1,000 for supplemental policies with limited nondrug benefits and an annual cap of $1,250 on drug spending. The majority of elderly Americans, whose annual incomes are less than $25,000, cannot afford such bare-bones Medigap policies.

According to a recent analysis by the National Academy of Social Insurance, a Medicare drug benefit will cost between $10 billion and $24 billion annually, depending on its terms. Adding prescription drug coverage to Medicare as a purely voluntary option won't work because of the "adverse selection" problem: The relatively healthy, whose premiums are needed to subsidize the sick, will tend not to buy drug coverage absent a generous government subsidy. Those most likely to choose the drug coverage option are those most likely to incur significant drug costs. But without an insurance pool that includes both the well and the sick, the sick would face astronomical and unaffordable premium costs.

According to the federal government actuary, a subsidy of about 50 percent of the premium price would be necessary to convince the majority of beneficiaries to enroll voluntarily in a Medicare drug coverage program, thereby solving the adverse selection problem. Without an adequate subsidy, this problem could be solved if all Medicare beneficiaries were required to participate in such a program. But mandatory participation is not viewed as a viable political option, in light of the resounding defeat of efforts to add a mandatory catastrophic-care option to Medicare in the late 1980s.

A related problem is that generously subsidizing Medicare drug coverage with federal dollars will drive out billions of private dollars already spent through employer-provided plans and private policies, increasing federal spending without a commensurate increase in prescription drug coverage. To many skeptics, this "substitution" of federal dollars for private dollars is simply wasteful. But there is no way around it if the adverse selection problem is to be resolved.

If Medicare did cover prescription drugs, how would they be priced? The federal government uses price controls for other services covered by Medicare. Would the price control mechanism be extended to drugs? If so, how would the government set the prices it would pay for drugs over time, and how would it deal with the rapid introduction of new drugs? Concerns over government price controls are the major reason the drug industry remains strenuously opposed to adding drug coverage to Medicare, and the industry stands ready to exercise its substantial political clout to stop it. Indeed, the industry even opposes a compromise solution, which would allow the same private beneficiary providers that act as drug purchasers for HMOs to act as drug purchasers for Medicare.

The chairmen of the National Bipartisan Commission on the Future of Medicare, Senator John Breaux and Congressman Bill Thomas, proposed a new subsidy to the poorest elderly (those with annual incomes of less than $11,000) for the purchase of limited drug coverage through private plans. The Democratic members of the commission rejected this recommendation on grounds of efficiency and equity. First, given the meager proposed subsidy, the unresolved adverse-selection problem would price drug coverage out of range for many low-income beneficiaries. And second, limiting a new Medicare drug subsidy to the poorest beneficiaries would violate Medicare's basic structure as a social insurance program available to all elderly Americans regardless of income.

Both criticisms helped shape the proposal for Medicare drug coverage offered by the Clinton administration in July 1999 and the proposals currently championed by Vice President Al Gore and Bill Bradley in their presidential campaigns. All three introduce a new voluntary Medicare prescription drug benefit, with a generous federal subsidy. The Clinton administration and Vice President Gore offer a voluntary program in which the government would pay half of the costs of prescription drugs for each beneficiary, up to a total of $5,000 with no deductible. Additional subsidies would be available for elderly with incomes below 150 percent of poverty, fully covering the premiums and cost sharing for those with incomes below 135 percent of poverty. To strengthen the incentives to participate, these plans would ensure that, even after the $5,000 cap is hit, beneficiaries would continue to receive price discounts on drug purchases. To counter concerns that Medicare drug coverage would drive out private coverage, the Clinton-Gore plans also provide financial incentives for employers to retain prescription drug benefits for retirees.

Bradley's plan does not cap overall drug benefits, but it does have a deductible and more stringent co-payment requirements. Although the latter features provide stronger incentives to discourage unnecessary drug expenditures than do the no-deductible Clinton-Gore plans, such features are unlikely to win favor with the Medicare beneficiaries who prefer first-dollar coverage. Even with these incentives, the absence of an overall cap means that Bradley's drug plan is about twice as expensive as Vice President Gore's.

Despite these and other technical differences between the Clinton-Gore and Bradley drug coverage proposals, all three of them share an important feature. They eschew price controls, relying instead on negotiations between private benefits managers and drug companies to determine the prices to be charged to Medicare beneficiaries. Medicare would operate much like HMOs that rely on such managers to negotiate drug prices for their covered patients. Because of its sheer size, however, Medicare would have significant clout in such negotiations--which is why the drug industry remains opposed.

Similar opposition to Medicare's bargaining strength has also blocked numerous proposals to allow Medicare to engage in competitive bidding to determine the prices of other covered services rather than to pay a set of controlled prices to "any willing provider" of such services. If Medicare is to move away from its antiquated and increasingly ineffective system of price controls, it will have to be granted the ability to pick and choose among providers and to exercise purchasing power against drug companies just as private plans do.

This conclusion motivated the Clinton administration to recommend that Medicare be allowed to adopt techniques already proven effective in the private sector at controlling costs without impairing the quality of care. These techniques include preferred-provider options in which beneficiaries are charged lower prices if they choose service from designated high-quality providers, competitive bidding and selective contracting in purchases of goods and services, and capitated disease- and case-management techniques in contracts for the treatment of high-cost chronic diseases like diabetes and coronary heart disease.

Competition and Market Reform

Most members of the National Bipartisan Commission on the Future of Medicare were convinced of the necessity of enhancing Medicare's ability to negotiate with private providers, except in the area of drugs. But a majority of commission members wanted to go even further by allowing private plans to compete directly with one another and with HCFA in the provision of all Medicare benefits. The purpose of such competition would be to encourage more efficient, cost-sensitive decision making by both producers and consumers in the Medicare program. Producers would be encouraged to become more efficient because they would be allowed to compete with one another for Medicare dollars. Consumers would be encouraged to become more cost conscious because they would be allowed to choose among competing plans for their Medicare services and to pay lower Medicare premiums if they chose lower-cost plans. Both economic logic and the experiences of the Federal Employees Health Benefits Program and other employersponsored programs allowing beneficiaries a choice among comparable insurance plans on the basis of price indicate that a similar system for Medicare could generate cost savings over time. There is, however, considerable uncertainty about the magnitude of this potential effect.

The competition approach espoused by the majority of commission members had two defining features. First, all plans granted permission by the government to compete for Medicare beneficiaries would be required to offer the same specified set of benefits as those in the standard fee-for-service Medicare plan. This condition was necessary to guarantee that competition among plans would be based on price and quality, not on coverage. Second, the government would make a contribution toward the payment of the premium of the plan chosen by each beneficiary, with the remainder paid by the beneficiary. The government's contribution would be set as a percentage of the weighted average of the costs of all plans, including the standard HCFA fee-for-service plan, competing to provide Medicare benefits. Setting and adjusting this percentage over time would be key political decisions, determining how Medicare costs were shared among beneficiaries and taxpayers. At least initially, commission members proposed setting this percentage at about 88 percent, the federal government's current share in Medicare spending per beneficiary.

The commission's competition approach preserved Medicare's entitlement to a specified set of benefits, allowed beneficiaries to pay lower premiums for plans covering such benefits at lower prices, and imposed a specified share of the risk of rising Medicare costs to its beneficiaries. Such an approach, which has come to be called a "premium support" approach in the health care field, should not be confused with a voucher approach. In a voucher system, such as the one proposed by the Republican Congress in 1995, the government would pay a fixed dollar amount to each beneficiary toward the purchase of a private plan without specifying the benefits to be covered. A voucher model would replace the Medicare entitlement to a package of specified health benefits with an entitlement to a government subsidy to purchase private health coverage. And if the value of this subsidy were set in nominal dollar terms, as the 1995 Republican plan proposed, all of the burden of rising health care costs would be borne by the beneficiaries.

Although the premium-support approach has several desirable features, it also has some serious downside risks. First, like any system based on choice in insurance markets, this approach would enhance the ability of providers to isolate the sickest population into small high-cost pools. This risk could be mitigated with a mandate that each participating health plan accept all Medicare beneficiaries who apply for coverage, and charge them the same premium regardless of difference in their age, sex, disability status, and other health risk indicators. The government's premium contribution to different plans would then be "risk adjusted" to vary depending on the risk characteristics of their enrollees, but the premium contributions of all enrollees choosing a particular plan would remain the same.

Second, given significant regional differences in health costs, there could also be significant regional differences in the premiums of private plans competing to provide such services in a premium-support system. This risk could be mitigated with an adjustment of the government's premium contributions so that Medicare beneficiaries would not be required to pay more for the same package of services because they lived in high-cost regions or because there were few competing plans in their regions.

Third, if the government's premium contribution were set as a percentage of the weighted average costs of participating plans, including traditional fee-for-service Medicare, and if its costs grew more rapidly than the costs of alternative plans, then those choosing the fee-for-service option could face significant increases in their premiums. As a result, although the Medicare entitlement to a specified set of services would be protected under the premium-support approach, the Medicare entitlement to receive these services with unlimited choice among providers in a fee-for-service setting would not be. Given the strong preference of the elderly, particularly those who are older and sicker, for choice among providers, and given mounting public skepticism about managed care, the risk of eroding Medicare's entitlement to an affordable fee-for-service option is the major reason for virulent opposition to the premiumsupport approach. Such an approach is based precisely on the idea that beneficiaries choosing more expensive plans would pay more while those choosing cheaper ones would pay less. This would mean that Medicare beneficiaries with limited incomes would be encouraged to choose low-cost plans and could be priced out of high-cost plans, including the fee-for-service plan to which they are currently entitled in the Medicare program.

The July 1999 Clinton plan proposed a compromise, allowing greater competition and choice in Medicare services while guaranteeing that beneficiaries selecting Medicare's traditional fee-for-service program continue to pay only the premium required by current law. The president's plan introduced a new "competitive defined benefit program," which, like the premium-support model, would allow private plans to compete with one another and with HCFA to offer a specified set of Medicare benefits. In contrast to the current Medicare+Choice system, in which the government pays managed-care providers a flat payment based on the costs of Medicare's fee-for-service plan, in the competitive defined-benefit model, the government would pay private plans on the basis of their actual price bids. Payment to competing plans on the basis of such bids has been used successfully by private employers and the Federal Employees Health Benefits Program to enhance efficiency and contain program costs. In the Clinton plan, beneficiaries could choose among plans offering Medicare benefits at prices lower than the cost of HCFA's traditional fee-for-service program, and they would be allowed to keep 75 percent of the resulting premium savings, with the remainder going to the Medicare program. Those who chose a more expensive plan than the traditional one would pay the full additional cost of their selection. And the government would continue to rely on price controls and whatever private-sector management techniques it was authorized to use over time to control the costs of the traditional fee-for-service program.

In contrast to the premium-support model, the president's competitive defined-benefits model would preserve Medicare's entitlement both to a specified set of benefits and to fee-for-service medicine with unlimited choice among providers. The preservation of the latter entitlement would come at the price of considerably smaller savings in projected Medicare expenditures. Indeed, the administration estimated that the savings resulting from its competitive defined-benefits plan would only amount to about $8 billion over its first 10 years. In contrast, the commission estimated that premium support would save $66 billion over its first decade. And over the long run, the commission optimistically assumed that the efficiencies fostered by premium support would slow Medicare spending about a percentage point per year. Even the most ardent proponents of such an approach among academic economists warn that its aggregate savings are likely to be far smaller and insufficient to resolve Medicare's long-run financing gap.

Medicare's Financial Health

As an additional measure to slow the long-run growth of Medicare spending, the chairmen of the commission recommended a gradual increase in the age of eligibility for Medicare from 65 to 67, to conform to Social Security's gradual increase in the age of eligibility to 67, by 2022. According to commission estimates, the proposed increase in the eligibility age for Medicare would eliminate about one-quarter of the projected gap between Medicare Part A expenditures and revenues by 2030. But this financial improvement would come at the cost of an increased number of uninsured Americans. According to recent estimates, between 10 percent and 20 percent of Americans aged 65-66 would lose health insurance if Medicare's age of eligibility were increased to 67, which would in turn aggravate one of the glaring failings of the American health care system: the large and growing number of the uninsured.

In contrast to the commission's proposal, the Clinton plan left Medicare's eligibility age at 65 and allowed Americans between ages 55 and 65 to purchase Medicare coverage at an actuarially fair market price. The new buy-in option was designed to pay for itself while addressing the problem of uninsured older Americans. In recent years, the number of uninsured has been growing most rapidly among Americans aged 55-65, who are twice as likely to have health problems as the 45-55 age cohort.

Neither the commission's recommendations nor the president's plan proposed introducing income testing into Medicare premium payments. Income testing might be justified on the grounds that Medicare has become a distinctively regressive system over time, with poorer elderly both paying a larger fraction of their disposable income for health and utilizing Medicare services less extensively than richer elderly. But in order for there to be significant Medicare savings from income testing, either the high-income threshold for paying a higher premium would have to be set quite low or the income-related premiums would have to be set quite high. The former approach would require politically sensitive increases in premium payments for a large fraction of the Medicare population. The latter would encourage the more affluent elderly to choose private coverage over Medicare, undermining its social insurance function and relegating it to a program primarily for the poorer, sicker population.

Medicare requires major reforms. Incremental measures that rely on tightening arbitrary price controls will only increase dissatisfaction with the Medicare program from providers, beneficiaries, and taxpayers. Given the dramatic changes that have occurred in the health care system since 1965 when Medicare was first introduced, at least four types of changes in Medicare are warranted.

First, a meaningful prescription drug benefit must be added to Medicare on a voluntary basis, with the government relying on private benefits managers as intermediaries to negotiate drug prices. Second, Medicare must be granted authority to use modern management techniques to pick and choose among providers and plans based on their cost and quality of service. Third, to encourage greater efficiency in the Medicare program, private plans must be allowed to compete with HCFA and with one another to provide a specified benefits package on the basis of price and quality, and beneficiaries must be allowed to share in the savings resulting from choosing low-cost plans. This condition implies that Medicare must be reinterpreted as an entitlement to a government subsidy and choice among competing health care plans for coverage of specified benefits, not as an entitlement to unlimited selection among providers in a fee-for-service setting. Finally, if Medicare's unintended regressivity is to be reduced, its premium should vary with income, and the higher premiums paid by high-income beneficiaries should be used to finance an expansion of benefits for low-income beneficiaries.

Even with structural reforms that encourage more competition in the Medicare program, its costs per beneficiary should be expected to continue to rise at about the same pace as per capita private health care spending. And without unforeseen changes in medical technology, both kinds of spending will continue to grow faster than the overall economy. Reform proposals that assume or require that per capita Medicare costs grow more slowly than private health care costs are neither credible nor justifiable. The reason for reforming Medicare is to reduce its inadequacies, inequities, inefficiencies, and projected financing gap, not, as some alarmists would have it, to save the American economy from toppling under the weight of Medicare spending.

A modified and phased-in premium-support approach such as the one proposed by President Clinton, along with the addition of a prescription drug benefit and a limited degree of income testing, is a sensible yet cautious way to begin modernizing Medicare to meet its real challenges. Even these structural reforms will not solve Medicare's long-term financing gap. But the economy is strong enough to provide decent health care for America's elderly, and we can find the additional revenues to do so. ยค