Skip Navigation

Madam Chairman and Members of the Committee, I am pleased to be here today to discuss some of the major issues affecting the design of an outpatient prescription drug benefit for Medicare beneficiaries. Those design issues present some difficult choices among desirable, but potentially conflicting, objectives and need to be considered in the context of the growing financial pressures facing the Medicare program.
 

FINANCIAL PRESSURES FACING THE MEDICARE PROGRAM

The growth of Medicare spending has been much slower in the past few years than it has been historically. In fiscal years 1998 through 2001, the Congressional Budget Office (CBO) estimates that benefit payments will grow at an average annual rate of 3.1 percent, compared with 10.0 percent per year over the previous decade. That growth estimate reflects CBO's May 2001 projections of baseline spending.

CBO further estimates that Medicare will spend $237 billion on benefits for 40 million elderly and disabled people in fiscal year 2001. Despite the recent slowdown in spending growth, that amount is almost 25 percent more than Medicare spent five years ago. The program now accounts for about 13 percent of estimated total federal spending, or 2.3 percent of gross domestic product (GDP).

Moreover, CBO is projecting faster Medicare growth over the next decade. We estimate that Medicare spending will more than double--reaching $499 billion--by fiscal year 2011, reflecting an average increase of 7.9 percent per year (see Figure 1). At that rate, Medicare spending in 2011 will constitute 19 percent of the federal budget, assuming that no change occurs in current tax and spending policies. In fact, the program will account for 36 percent of the projected increase in federal spending by the end of the decade.

The latest report from the Board of Trustees for Medicare projects that total Medicare spending will increase substantially in the long run, rising from 2.3 percent of GDP in 2001 to 8.5 percent in 2075. In addition, the difference between projected total Medicare spending and total federal revenues specifically dedicated to the program is expected to grow substantially. Sources of those dedicated revenues include the Medicare payroll tax, the portion of the income taxes on Social Security benefits that is paid to the Hospital Insurance (HI) trust fund (Part A of Medicare), and premiums paid by enrollees for Supplementary Medical Insurance (SMI, or Part B of Medicare). According to the Medicare trustees, the discrepancy between total Medicare expenditures and dedicated revenues will be $64.0 billion in 2001, or 0.6 percent of GDP (see Figure 2). By 2075, that gap is projected to grow to 6.0 percent of GDP. The growing difference between spending and dedicated revenues indicates the Medicare program's increasing dependence on general revenues to pay its bills.

These financial pressures have focused policymakers' attention on restructuring the Medicare program. There are two potentially conflicting considerations:

• First, Medicare spending is expected to grow at a rapid rate, making the program increasingly dependent on general revenues and, ultimately, unsustainable in its present form.



• Second, Medicare does not provide the protection offered by most private insurance, since it lacks a stop-loss provision and coverage for prescription drugs.

PROVIDING MEDICARE BENEFICIARIES WITH COVERAGE FOR PRESCRIPTION DRUGS

Modernizing Medicare's benefit package by adding a prescription drug benefit could close a significant gap in program coverage but only at a sizable cost to the federal government or to enrollees.

Beneficiaries' Current Spending on Prescription Drugs

In recent years, growth in prescription drug spending has far outpaced growth in spending for other types of health care. Those rising expenditures have had a significant impact not only on Medicare beneficiaries but on employers who offer retiree health coverage and on state governments as well.

Between 1990 and 2000, annual spending on prescription drugs in the United States grew at nearly twice the rate as that for total national health expenditures, and it has maintained a double-digit pace since the mid-1990s. For the U.S. population as a whole, three factors explain most of that growth: the introduction of new and costlier drug treatments, broader use of prescription drugs by a larger number of people, and lower cost-sharing requirements by private health plans. Within some therapeutic classes, new brand-name drugs tend to be much costlier than older drug therapies, which has also contributed to growth in spending. Use of prescription drugs has broadened as well, because many new drugs provide better treatment or have fewer side effects than older alternatives and more people are aware of new drug therapies through the "direct to consumer" advertising campaigns of pharmaceutical manufacturers.

Even without a Medicare drug benefit, CBO expects prescription drug costs for Medicare enrollees to grow at a rapid pace over the next decade (see Table 1). At an average annual rate of 10.3 percent per beneficiary, drug costs are expected to rise at nearly twice the pace of combined costs for Medicare's HI and SMI programs, and much faster than growth in the nation's economy. (CBO's estimates of rising drug spending are based on the latest projections for prescription drug costs within the national health accounts.)

CBO's baseline estimate of prescription drug costs for Medicare enrollees is up significantly over last year because of higher projections of the rate of growth in per capita drug costs. Last year's analysis indicated that spending by Medicare enrollees on outpatient drugs not covered by Medicare would total $1.1 trillion over the period 2001 through 2010 (see Table 2). This year, our projection for the same period is $1.3 trillion, or about 18 percent higher.

Our estimate for 2002 through 2011, the current 10-year projection period, is roughly $1.5 trillion--which is about 33 percent higher than last year's projection for 2001 through 2010. The jump results from assuming a higher growth rate and replacing an early low-cost year (2001) with a late high-cost year (2011).

Those changes to CBO's baseline estimate--higher per capita drug spending and the inclusion of a new high-cost year in the projection window--imply that proposals for a prescription drug benefit will have a higher price tag than they did last year. But for any given proposal, the exact magnitude of the difference between CBO's estimate for last year and its estimate for this year will also depend on the bill's specific features.

Existing Coverage

While third-party coverage for prescription drugs has become more generous over time for the population as a whole, that trend is less clear for Medicare beneficiaries. In 1997, nearly one-third of the Medicare population had no prescription drug coverage. On average, Medicare beneficiaries paid about 45 percent of their drug expenditures out of pocket (see Figure 3). By comparison, all people in the United States paid an average of 39 percent of the cost of their prescriptions. Because Medicare beneficiaries are elderly or disabled, they are more likely to have chronic health conditions and use more prescription drugs: nearly 89 percent filled at least one pre-scription in 1997. Medicare beneficiaries made up 14 percent of the population that year, yet they accounted for about 40 percent of the $75 billion spent on prescription drugs in the United States.

Those factors suggest that growth in drug spending has a larger financial impact on the Medicare population than on other population groups. However, aggregate statistics mask a wide variety of personal circumstances. Nearly 70 percent of beneficiaries obtain drug coverage as part of a plan that supplements Medicare's benefits, but those supplemental plans vary significantly in their generosity.

Traditionally, retiree health plans have provided prescription drug coverage to more seniors than any other source, and their benefits have been relatively generous. In 1997, about one-third of Medicare beneficiaries had supplemental coverage through a current or former employer, and most of those plans provided drug coverage (see Table 3). Although specific benefits vary, it is common to find relatively low deductibles and copayments in employer-sponsored drug plans.

However, because prescription drug spending by elderly retirees has become a significant cost to employers, many have begun to restructure their benefits. For example, a 1997 Hewitt Associates' study for the Kaiser Family Foundation found that among large employers, drug spending for people age 65 or older made up 40 percent to 60 percent of the total cost of their retiree health plans. Average utilization of prescription drugs among elderly retirees was more than double that for active workers. Although relatively few employers in the Hewitt survey have dropped retiree coverage altogether, most have taken steps to control costs, such as tightening eligibility standards, requiring retirees to contribute more toward premiums, placing caps on the amount of benefits that plans will cover, and encouraging elderly beneficiaries to enroll in managed care plans.

Medicare+Choice (M+C) plans are another means by which the elderly and disabled have obtained prescription drug coverage. In 2000, for example, 64 percent of Medicare beneficiaries had access to M+C plans that offered some drug coverage, although a significantly smaller fraction of elderly people signed up for those plans. Many M+C plans have scaled back their drug benefits in response to rising costs and slower growth in Medicare's payment rates. Nearly all such plans have annual caps on drug benefits for enrollees--many at a level of only $500 per year--and a growing share of plans charge a premium for supplemental benefits.

While 26 percent of the Medicare population relied on individually purchased (often medigap) plans as their sole form of supplemental coverage in 1997, less than half of that group had policies that covered prescription drugs. Medigap plans with drug coverage tend to be much less generous than retiree health plans; medigap plans have a deductible of $250, 50 percent coinsurance, and annual benefit limits of either $1,250 or $3,000. Premiums for plans that include drug coverage also tend to be much higher than premiums for other medigap plans, due in part to their tendency to attract enrollees who have higher-than-average health expenses.

Certain low-income Medicare beneficiaries also may be eligible for Medicaid coverage, which generally includes a prescription drug benefit. All state Medicaid programs offer prescription drug coverage (usually involving little or no cost sharing) to people whose income and assets fall below certain thresholds. In addition, as of January 2001, 26 states had authorized (but had not necessarily yet implemented) some type of pharmaceutical assistance program, most of which would provide direct aid for purchases to low-income seniors who did not meet the Medicaid requirements. About 64 percent of the Medicare population lives in those states.

Thus, middle- and higher-income seniors can usually obtain coverage through retiree or M+C plans, while seniors with the lowest income generally have access to state-based drug benefit programs. However, beneficiaries with income between one and two times the poverty level are more likely to be caught in the middle, with income or asset levels that are too high to qualify for state programs and less access than higher-income enrollees to drug coverage through former employers.

Design Choices for a Medicare Drug Benefit

A Medicare drug benefit might address a number of objectives. The most fundamental would be to ensure that all beneficiaries had access to reasonable coverage for outpatient prescription drug costs--but this fundamental notion allows for considerable debate about what that would mean. The various objectives that might be thought desirable in the abstract are often mutually incompatible; as a result, difficult choices must be made. For example, it is not possible to provide a generous drug benefit to all Medicare beneficiaries at low cost--either enrollees' premiums or the government's subsidy costs would be high. If most of the costs were paid by enrollees' premiums to keep federal costs low, some Medicare beneficiaries would be unwilling or unable to participate in the program. If costs were limited by covering only catastrophic expenses, few enrollees would benefit in any given year, possibly reducing support for the program. If, instead, costs were limited by capping the annual benefits paid to each enrollee, the program would fail to protect participants from the impact of catastrophic expenses.

In designing a drug benefit, policymakers must make four fundamental decisions:

• Who may participate?



• How will program costs be financed?



• How comprehensive will coverage be?



• Who will administer the benefit and under what conditions?

Participation. Although most Medicare enrollees use some prescription drugs, the bulk of such spending is concentrated among a much smaller group. In 1997, about 13 percent of enrollees had expenditures of $2,000 or more, accounting for 45 percent of total drug spending by the Medicare population. Forty-six percent had expenditures of $500 or less, making up about 8 percent of total spending. Most spending is associated with treatment of chronic conditions--such as hypertension, cardiovascular disease, and diabetes. The skewed distribution of spending and the need for people with chronic conditions to stay on drug therapies over the long term makes stand-alone drug coverage particularly susceptible to adverse selection, where enrollment is concentrated among those who expect to receive more in benefits than they would pay in premiums.

Because of the likelihood of adverse selection, a premium-financed drug benefit offered as a voluntary option for Medicare enrollees must restrict participation in some way. If Medicare beneficiaries were free to enroll in or leave the program at will, only those who expected to gain from the benefit would participate each year. That would drive premiums up, which would further reduce enrollment as enrollees with below-average drug costs dropped out.

Most of the drug benefit proposals developed in 2000 would have provided a voluntary drug option, but they attempted to mitigate the potential for adverse selection by one of two approaches: either they gave enrollees only one opportunity to choose the drug benefit at the time enrollees first became eligible, or they imposed an actuarially fair surcharge on premiums for those who delayed enrollment. Another approach to avoiding the problem of adverse selection would be to couple the drug benefit with Part B of Medicare, so that enrollees could choose either Part B plus a drug benefit or no Part B and no drug benefit. In that case, even if the drug portion of the benefit was not heavily subsidized, the current 75 percent subsidy of Part B benefits would ensure nearly universal participation in the coupled benefit.

Financing. Program costs could be entirely financed by enrollees' premiums, or some or all of the costs could be paid by the federal government. Given a one-time-only enrollment option, participation rates would be reasonably high, even if the program was largely financed by enrollees' premiums. If enrollees lived long enough, virtually all of them would benefit from drug coverage, and the erosion now occurring in the comprehensive coverage provided by private plans would also spur participation. Further, employer-sponsored health plans would probably require that retirees eligible for a new Medicare benefit participate in it, just as they now effectively require that retirees participate in Part B. And state Medicaid agencies, even if not mandated to do so, would choose to enroll dual eligibles (people eligible for both Medicare and Medicaid) in a new Medicare drug benefit if their costs under the new program were less than the cost of the drug benefits now provided under Medicaid. However, if a generous drug benefit was fully financed by enrollees, premiums would be high, making the benefit difficult to afford for lower-income beneficiaries ineligible for Medicaid. The drug proposals developed last year all provided full subsidies to low-income people for both cost-sharing and premium expenses, in addition to partially subsidizing premium costs for all other enrollees.

Coverage. A Medicare drug benefit could be designed to look like the benefit typically provided by employer-sponsored plans. If so, it would be integrated with the rest of the Medicare benefit. Further, it would have low cost-sharing requirements (ranging from 20 percent to 25 percent coinsurance or a copayment per prescription of $10 to $25) and stop-loss protection--a dollar limit above which no cost sharing would be required. Such comprehensive coverage would provide good protection for enrollees, but it would be very costly. Not only would it transfer most of the costs of drugs currently used by enrollees to the Medicare program, but it would also increase utilization among those who now have less generous coverage.

One way to constrain costs and utilization is by limiting coverage--covering only catastrophic costs, for example, or imposing a cap on benefits paid per enrollee each year. If Medicare provided coverage only for catastrophic costs, most enrollees would receive no benefit payments in any given year. Nevertheless, it would be inaccurate to say that those enrollees would receive no benefit, since they would be protected against the possibility of catastrophic expenses--the main function of insurance. Public support for a drug benefit might be stronger, though, if most enrollees could reasonably expect to receive some benefit payments each year.

Alternatively, policymakers could take the other approach to limiting costs: covering a portion of all drug costs but only up to a benefit cap. However, because that approach would not protect those enrollees who were most in need, most of last year's proposals included stop-loss protection. The end result was a benefit unlike anything available in the private sector--a hybrid that had a capped benefit, then a "hole" with no drug coverage, and finally a stop-loss provision, beyond which the program would pay all drug costs (see Figure 4). The larger the range of spending encompassed by the hole, the less costly the program would be--but also the less coverage the benefit would provide.

An approach to limiting costs within the context of a more traditional benefit would be to have a higher initial deductible amount, relatively high cost-sharing requirements, and a high stop-loss threshold. Or the program could provide a more generous benefit similar to those provided by employer-sponsored plans, with federal costs limited by financing most of the program's costs through enrollees' premiums.

Administration. The way in which a drug benefit is administered can also have a significant effect on how costly it is. All recent proposals have envisioned adopting the now common private-sector approach of using pharmacy benefit managers (PBMs) in each region. Proposals have differed, however, in whether only one or several PBMs would serve a region, in whether the responsible entities would assume any insurance risk, and in the kind of restrictions that would be placed on them.

Private health plans use PBMs to process claims and negotiate price discounts with drug manufacturers and dispensing pharmacies. PBMs also try to steer beneficiaries toward lower-cost drugs, such as generic, preferred formulary, or mail-order drugs. In addition, because of their centralized records for each enrollee's prescriptions, they can help prevent adverse drug interactions. The likelihood that PBMs could effectively constrain costs depends on their having both the authority and the incentive to aggressively use the various cost-control mechanisms at their disposal. In the private sector, PBMs often have considerable leeway in the tools they can use, but they do not assume any insurance risk for the drug benefit. At most, they may be subject to a bonus or a penalty added to their administrative fee, based on how well they meet prespecified goals for their performance.

Some of the proposals developed last year (such as the one developed by the Clinton Administration) adopted the typical private-sector model, with a single PBM selected periodically to serve each region and with all insurance risk borne by Medicare, not the PBM. There are two main concerns about that model: it might prove politically difficult to allow the designated PBMs to use cost-control tools aggressively if enrollees have no choice of provider in each region, and non-risk-bearing PBMs might have too little incentive to use strong tools, even if they were permitted.

Other proposals (such as the Breaux-Frist bills and the House-passed drug bill) adopted a different model, more akin to the risk-based competitive model characteristic of Medicare+Choice plans. Those proposals envisioned multiple risk-bearing entities (such as PBM/insurer partners) that would compete to serve enrollees in each region. Enrollees would have some choice among providers, so that beneficiaries who were willing to accept more-restrictive rules (such as a closed formulary) in return for lower premium costs could do so, while others could select a more expensive provider with fewer restrictions. If the entities bore all of the insurance risk for the drug benefit, they would have strong incentives to use whatever cost-control tools were permitted. However, they would also have strong incentives to try to achieve favorable selection by avoiding enrollees most in need of coverage.

One of the concerns raised about this model was that no entities might be willing to participate if they had to assume the full insurance risk for a stand-alone drug benefit. To mitigate that concern, the proposals included federally provided reinsurance for high-cost enrollees. (Reinsurance means that the federal government shares part or all of the costs of high-cost enrollees.) However, reinsurance would tend to weaken the plans' incentives to control costs. Another concern was that differences among plans in benefit structures or strategies for cost control could result in some plans attracting low-cost enrollees and others attracting more costly enrollees. The risk of that kind of selection would lead plans to raise the cost of the benefit. Moreover, to avoid such risks, plans would, over time, come to offer very similar plan designs.

The Cost of Covering Prescription Drugs for Medicare Enrollees

There are numerous design parameters that must be specified in developing a Medicare prescription drug benefit, and decisions concerning those parameters can greatly affect the benefit's cost to the taxpayer and to the beneficiary. This testimony provides some examples of how costs would be affected by varying certain aspects of the benefit's design.

The estimates that follow are approximate and subject to change; the cost of a detailed proposal would vary depending on its precise specifications. The estimates are for 2004 only.

Base Case. For purposes of this testimony, the base case is a benefit that provides coverage for all of the outpatient drug costs of Medicare enrollees (see Table 4). The enrollee would be responsible for coinsurance equal to 50 percent of the cost of prescription drugs up to $8,000 of spending. The new benefit would cover the entire cost of drugs above that amount. Thus, the enrollee would be liable for up to $4,000 in out-of-pocket spending before reaching the stop-loss amount.

To pay for this program, enrollees would be charged a monthly premium designed to cover 50 percent of the cost of the benefit. The federal government would pay for the other 50 percent. We assume that a subsidy of that size would be sufficient to ensure that all enrollees in Medicare Part B would participate in the prescription drug program.

Low-income enrollees would receive a subsidy to enable them to participate in the Medicare drug program. Enrollees with income up to 135 percent of the federal poverty level would receive a full subsidy of premiums and cost-sharing amounts. Those with income between 135 percent and 150 percent of the poverty level would receive a premium subsidy (on a sliding scale that declined with income) but would be responsible for any cost sharing. States and the federal government would share in those subsidy costs for enrollees with income of less than 100 percent of the poverty level and for those who were dually eligible for Medicare and Medicaid.

The base case also assumes that a single PBM would administer the program in each region, with all insurance risk borne by Medicare. The cases presented in this testimony do not consider the other major alternative for delivering a Medicare drug benefit: instead of a single PBM, the program could be operated through multiple risk-bearing entities who would compete for enrollees. Competing PBM/insurer partners who bore insurance risk would have a strong incentive to use such tools as restrictive formularies and three-tier copayment structures to aggressively manage costs. However, they would also incur certain "load" costs--such as marketing expenses to attract enrollees and a premium for accepting insurance risk--that a single PBM would not. The net impact on program costs would depend on the specific details of the proposal.

The benefit design assumed for the base case would cost the federal government about $30.9 billion in 2004. The Medicare benefit portion of that total is $26.5 billion, and the low-income subsidy (and interactions with the Medicaid program) account for the remaining $4.3 billion (see Table 5). As we will see in comparisons with other cases, a less generous drug benefit would decrease Medicare costs but increase the cost of the low-income subsidy.

In the aggregate, enrollees would pay a total of $26.5 billion in premiums, reflecting a $56.80 monthly premium that they would pay under the base case plan. That total includes premiums that are paid by Medicaid on behalf of low-income enrollees. In addition, enrollees would face about $44.5 billion in cost sharing for the prescription drugs that they used. Again, that amount includes some cost sharing that would be picked up by supplemental payers, including employer-sponsored insurance and medigap plans. As we will demonstrate below, a less generous benefit would lower premiums but raise the amount of cost sharing paid by enrollees.

Federal costs could be reduced by imposing more cost sharing on enrollees or by varying other aspects of the design. The following discussion of alternative cases examines how the costs imposed on taxpayers and beneficiaries would change if one or more features of the program are varied.

Change Beneficiaries' Cost Sharing. The overall federal cost of a prescription drug proposal would fall if beneficiaries were responsible for a greater share of program costs. Higher cost sharing would, of course, increase the cost of the low-income subsidy.

Case 1-A is identical to the base case except for a $250 annual deductible. Nearly 89 percent of enrollees have some prescription drug spending during the year and would thus be liable for at least part of the deductible. Including a deductible would lower Medicare costs but raise low-income costs compared with the base case. On balance, the federal cost of the program would fall to $28.7 billion in 2004, and monthly premiums would decline to $52.10. Beneficiaries who had more than $250 in drug spending that year would face higher costs under this option because the added cost of the deductible would be only partly offset by the reduced premium.

An even higher deductible would further reduce program costs. Case 1-B imposes a $500 deductible on the base case, and the federal cost drops to $26.9 billion in 2004. Doubling the deductible amount from Case 1-A does not double savings from the base case, however, because some enrollees who would pay the full $250 deductible would spend less than $500 on drugs in a year and thus would not pay the full amount of the higher deductible.

Lowering the coinsurance rate could alter program costs dramatically. The base case assumes a 50 percent coinsurance rate, while Case 1-C lowers that rate to 25 percent. That adjustment increases the program's net federal cost by nearly 40 percent, to $42.6 billion in 2004. Medicare's cost would increase to $38.4 billion, while the low-income subsidy would fall to $4.1 billion.

The lower coinsurance would drive premiums upward as program costs rose. Beneficiary premiums would increase by nearly half, to $82.30 monthly. In the aggregate, beneficiaries would pay about $38.4 billion in premiums. However, aggregate cost sharing would decline precipitously as well, to nearly $25 billion. While all enrollees would face the higher premiums, the lower coinsurance rate would primarily benefit enrollees with significant drug costs.

Raise the Stop-Loss Amount. The net federal program cost also could be reduced by raising the stop-loss amount, although the additional financial exposure would increase the cost of the low-income subsidy. Under the base case, the stop-loss amount is set at $4,000 paid out of pocket: a beneficiary who had used $8,000 in covered prescription drugs and paid 50 percent coinsurance would not be liable for any additional costs incurred during the year. (Enrollees who spend more than $8,000 account for about 23 percent of total baseline spending in 2004.)

Case 2-A raises the stop-loss amount to $6,000 in out-of-pocket spending. That higher level is equivalent to total spending by an enrollee of $12,000, which will account for less than 10 percent of total baseline spending in 2004. Under this option, the federal cost of the program would fall to $29.8 billion, a reduction of 3 percent from the base case. The low-income subsidy rises to $4.4 billion compared with the base case. Total premiums fall to about $25 billion, and aggregate cost sharing increases to nearly $47 billion.

Raising the stop-loss amount by an additional $2,000--to $8,000--lowers program costs by less than the previous incremental difference found in Case 2-A. The federal cost for Case 2-B is estimated to be $29.5 billion, or about 5 percent lower than the base case.

Cap Benefits. A third approach would place a limit on drug costs covered under the Medicare benefit. Case 3 would impose such a limit when the enrollee reached $2,500 in total drug spending. That is, the enrollee would receive up to $1,250 in reimbursement for drug expenses before reaching the benefit cap. Such a cap could be absolute, with no additional reimbursement for spending at any level above the cap. However, Case 3 keeps the same stop-loss provision as in the base case, so that the beneficiary faces no cost sharing beyond $5,250 in total charges. That structure leaves a "hole" in covered spending--a range of prescription drug spending for which most enrollees must pay all of their costs. (Individuals with income below 135 percent of the poverty level, whose cost sharing is fully subsidized, would be unaffected by this provision.)

Relative to the base case, the limit on coverage in Case 3 would lower Medicare costs but increase the low-income subsidy. The net federal cost would total approximately $27.0 billion in 2004. The option's benefit cap would lower premiums to about $22.5 billion but raise aggregate cost sharing to nearly $52 billion. Lower premiums under Case 3, compared with the base case, reflect a less-generous benefit.

Combine Features. The above options were designed to show how varying one parameter of a prescription drug benefit would affect program costs. This section looks at alternatives that combine several changes at the same time.

Case 4-A combines the base case with many of the features described above: a $250 deductible, benefits capped at $1,125 (after the enrollee reaches $2,500 in total drug spending), and stop-loss protection after the beneficiary spends $6,000 out of pocket. The costs of enrollees with income below 135 percent of the poverty level would be fully subsidized inside the benefit "hole."

Such a benefit would be significantly less generous than the base case, but the costs of financing it would be significantly lower as well. In 2004, federal costs would be approximately $21.5 billion, or about one-third less than the base case. Likewise, monthly premiums would fall from $56.80 under the base case to $35.90 under Case 4-A. That causes total premiums to drop to $16.8 billion, with a corresponding increase in aggregate cost sharing to $61.8 billion.

Case 4-B is identical to Case 4-A, except that low-income individuals would not be subsidized inside the benefit "hole." CBO estimates that in 2004, federal costs would total $20.4 billion. Nearly all of that savings comes from reductions in the cost of the low-income subsidy. Premiums would drop negligibly compared with Case 4-A.

Case 4-C extends the low-income subsidy to individuals with higher income than in previous cases. Specifically, it includes all of the features of Case 4-A but provides a full subsidy for premiums and cost sharing to enrollees who have income at or below 150 percent of the federal poverty level. Enrollees with income between 150 percent and 175 percent of the poverty level would receive a premium subsidy on a sliding scale. Medicare costs would remain roughly unchanged compared with Case 4-A, but the low-income subsidy would increase to $5.7 billion in 2004.

Increasing the federal subsidy for beneficiary premiums would substantially raise program costs. Case 4-D is identical to Case 4-A except that the federal subsidy is raised to 75 percent of premiums. That change increases Medicare costs by 50 percent compared with Case 4-A but lowers the cost of the low-income subsidy somewhat. The net federal cost would rise to over $28 billion in 2004. The sharp increase in Medicare costs is mirrored by the sharp drop in beneficiary premiums, which fall from about $17 billion in Case 4-A to about $8 billion in Case 4-D.

Because we have assumed throughout this discussion that the federal subsidy would be at least 50 percent, the increase in Case 4-D does not yield an increase in participation by Medicare enrollees. However, if the federal subsidy declined below 50 percent, CBO assumes that enrollment would decline somewhat.
 

CONCLUSIONS

While policymakers are well aware of Medicare's long-run financial problems, they also know that its benefit package has deficiencies relative to the benefits typically provided by private-sector insurance plans. One such deficiency is that the program provides only very limited coverage for outpatient prescription drugs--an increasingly important component of modern medical care. But adding a drug benefit would significantly increase Medicare's costs, and unless it was fully financed by enrollees' premiums, it would exacerbate the imbalance between the program's projected spending and its dedicated revenues.

We are extremely unlikely to see a new drug benefit that has no adverse impact on Medicare's long-term financial status. But, as I have discussed today, there are important design features that could be built in to such a benefit to limit federal costs while providing important insurance protection for enrollees. In developing a realistic policy proposal, hard decisions must be made to establish the proper balance among competing objectives.
 

FIGURE 1. ANNUAL AVERAGE MEDICARE SPENDING GROWTH FOR VARIOUS PERIODS

SOURCE: Historical data from the Health Care Financing Administration and projections by the Congressional Budget Office.

FIGURE 2. PROJECTED MEDICARE OUTLAYS AND DEDICATED REVENUES AS A PERCENTAGE OF GDP, CALENDAR YEARS 2000-2075

SOURCE: Board of Trustees, Federal Hospital Insurance Trust Fund (2001).

FIGURE 3. DISTRIBUTION OF DRUG SPENDING FOR MEDICARE ENROLLEES, BY PAYER, 1997

SOURCE: Congressional Budget Office tabulations from the 1997 Medicare Current Beneficiary Survey.

NOTE: Drugs currently covered by Medicare are not included here.

FIGURE 4. POSSIBLE FEATURES OF A PRESCRIPTION DRUG INSURANCE BENEFIT

SOURCE: Congressional Budget Office.