The Social Security Amendments of 1983 established the prospective payment system (PPS) under which Medicare pays hospitals for inpatient services provided to beneficiaries. Higher rates are paid to hospitals with teaching programs to cover their higher costs of caring for Medicare patients. Under the Balanced Budget Refinement Act of 1999, the additional percentage paid to teaching hospitals in 2000 will be approximately 6.5 percent for each 0.1 increase in a hospital's ratio of full-time interns and residents to its number of beds. Teaching hospitals will receive 6.25 percent more in 2001 and 5.5 percent more in 2002 and subsequent years for every 0.1 increase in the resident-to-bed-ratio. (Under the Balanced Budget Act of 1997, teaching hospitals would have received 6.0 percent more in 2000 and 5.5 percent more in 2001 and subsequent years for each 0.1 increase in the ratio.)

The Congress enacted the additional payments to teaching hospitals to compensate them for indirect teaching costs--such as the greater number of tests and procedures thought to be prescribed by interns and residents--and to cover higher costs from factors that are not otherwise accounted for in setting the PPS rates. Such factors might include more severely ill patients, a hospital's location in the inner city, and a more costly mix of staffing and facilities, all of which are associated with large teaching programs. (An alternative approach would base additional payments to teaching hospitals on the enhanced value of care in those settings. Such a proposal was recently made by the Medicare Payment Advisory Commission.)

The Prospective Payment Assessment Commission has estimated that a 4.1 percent adjustment to Medicare's payments would more closely match the increase in operating costs associated with teaching. If the teaching adjustment was lowered accordingly, outlays would fall by about $5.8 billion from current-law spending over the 2001-2005 period and by about $14.1 billion over the 2001-2010 period.

This option would better align payments with the actual costs incurred by teaching institutions. Furthermore, since the training that medical residents receive will result in a significant increase in their future income and since hospitals benefit from using residents' labor, it is reasonable for some or all of a hospital's indirect training costs to be borne by both residents and the hospital. Some of those costs are now passed on in the form of stipends that are lower than the value of the residents' services to the hospital. A lower teaching adjustment would probably lead to even lower stipends, however, as well as smaller residency programs. An additional consideration is that if the teaching hospitals now use some payments to fund such activities as charity care, people without health insurance could have less access to health services under this option.

Medicare's prospective payment system does not include payments to hospitals for the direct costs they incur in providing graduate medical education (GME)--namely, residents' salaries and fringe benefits, teaching costs, and institutional overhead. Instead, Medicare makes those payments separately on the basis of its share of a hospital's 1984 cost per resident indexed for increases in the level of consumer prices. Medicare's direct GME payments, which are received by about one-fifth of all U.S. hospitals, totaled about $2.2 billion for 1999. Under the proposed option, hospitals' direct GME payments would be based on the national average of salaries paid to residents in 1987, updated annually by the consumer price index for all urban consumers. Reimbursement would be based on 120 percent of the national average salary.

In effect, this option would reduce teaching and overhead payments for residents but continue to pay their salaries and fringe benefits. Unlike the current system, under which GME payments vary considerably from hospital to hospital, this option would pay every hospital the same amount for the same type of resident. (Although the Congress took action in 1999 to lessen some of the variation among hospitals in payments per resident, considerable differences remain under current law.) The option would also continue the current-law practice of reducing payments for residents who have gone beyond their initial residency period. The savings from current-law spending would total about $4.6 billion over the 2001-2005 period and about $10.3 billion over the 2001-2010 period.

The overall reduction in the level of subsidies might be warranted since market incentives appear to be sufficient to encourage a continuing flow of new physicians. Moreover, since hospitals use resident physicians to care for patients and since residency training helps young physicians earn higher incomes in the future, both hospitals and residents might reasonably contribute more to those training costs. Residents would contribute more to those costs if hospitals responded to the changes in reimbursements by cutting residents' salaries or fringe benefits.

If hospitals lowered residents' salaries or benefits, the costs of longer residencies--in terms of forgone practice income--could exert greater influence on the young physicians' decisions about pursuing a specialty. More residents might choose to begin primary care practice rather than specialize further. That outcome could be negative for the individual resident; by contrast, the Council on Graduate Medical Education and other groups believe that a relative increase in the number of primary care practitioners would be desirable. Finally, decreasing GME reimbursement could force some hospitals to reduce the resources they commit to training, possibly jeopardizing the quality of their medical education programs.

Under the prospective payment system for inpatient hospital services, Medicare pays hospitals an amount for each discharge that is intended to compensate the hospital for capital-related costs. Currently, teaching hospitals receive additional capital-related payments that are based on teaching intensity, measured as a hospital's ratio of residents to its average daily number of inpatients. Specifically, an increase of 0.1 in that ratio raises the hospital's capital-related payment by 2.8 percent.

Eliminating those extra payments would save the Medicare program about $0.3 billion in 2001. Five-year savings would equal about $1.6 billion, and savings over the 2001-2010 period would be $3.6 billion.

In contrast to higher operating costs, which analyses indicate are indeed associated with teaching intensity, a hospital's capital costs per case appear to be unrelated to intensity. Furthermore, paying teaching hospitals more than nonteaching hospitals for otherwise similar patients may discourage efficient decisionmaking by hospitals. In addition, Medicare's payment adjustments for teaching intensity may distort the market for residency training by artificially increasing the value (or decreasing the cost) of residents to hospitals. If residents' training raises the costs of patient care for a hospital, arguably the hospital should bear those costs in order to encourage an efficient amount of training. Hospitals are likely to shift such costs to residents in the form of lower stipends or greater workloads. Residents will engage in such training if they perceive that their future productivity, as reflected in their future incomes, will be great enough to outweigh those costs.

Eliminating the special capital-related payments would reduce revenues to teaching hospitals at a time when those hospitals already face pressures to reduce costs to remain competitive in the growing managed care environment. Teaching hospitals would probably have to reduce some services in response to the decline in their revenues. Those reductions in services could include less provision of public goods, such as research or providing medical care to the indigent.

Three types of Medicare graduate medical education (GME) payments are tied to the size or intensity of a teaching hospital's residency program: direct graduate medical education payments, the indirect medical education adjustment for inpatient operating costs, and the indirect medical education adjustment for inpatient capital-related costs. Under provisions in the Balanced Budget Act of 1997, teaching hospitals have begun to receive GME payments for participants in Medicare+Choice health plans in addition to the payments that they have traditionally received for fee-for-service Medicare patients. Several variables determine the total amount of GME payments that a hospital receives, including the number and diagnoses of Medicare discharges and numerical factors used for annually updating payments for inpatient operating costs and capital-related costs. Because of changes in those variables over time, the Congressional Budget Office expects GME payments under current law to grow at an average annual rate of 4.3 percent between 2001 and 2010.

This option would replace the current system with a consolidated block grant to fund the special activities of teaching hospitals. Under the current system, a hospital receives GME payments based on regulatory formulas, and total Medicare GME spending is the resulting sum of what Medicare owes each hospital. The option considered here assumes that a switch to the block-grant program would occur in 2001 and that the amount of the grant would be based on spending in 2000, increased for overall inflation. Compared with projected spending under current law, federal outlays would be reduced by $1.1 billion over the first five years and $6.0 billion over the 2001-2010 period.

Establishing a block grant for the three types of GME payments would allow the Congress to better monitor and adjust that funding. Another feature of the option is that Medicare would no longer pay different rates to hospitals for inpatient services merely because of differences in the size or presence of residency programs.

However, because this option would reduce total payments to teaching hospitals below the amounts expected under current law, such hospitals would, on average, receive less revenue than they would otherwise. In response, teaching hospitals might reduce the amount or quality of some of their services or their provision of some public goods, such as medical research or care for indigent people.

Under Medicare's prospective payment system (PPS) for inpatient hospital services, special rules apply to providers designated as sole community hospitals (SCHs). There are more than 700 SCHs, almost all of which are located in rural areas. Thus, about one-third of rural hospitals qualify for SCH status. Under the current rules, a hospital may be designated as an SCH if it meets specific criteria that define a sole provider of inpatient, acute care hospital services in a geographic area. In addition, some SCHs have been permitted to retain that status regardless of whether they meet the current sole-provider criteria.

Payments to SCHs are generally equal to the highest of three amounts: the regular federal PPS payment that would otherwise apply, an amount based on the hospital's costs in 1982 updated to the current year, or an amount based on the hospital's costs in 1987 updated to the current year. In addition, the Balanced Budget Refinement Act of 1999 allows certain SCHs to be paid according to their fiscal year 1996 costs. Hospitals that choose to receive the regular PPS payment--about half of all SCHs--are eligible to receive higher payment adjustments for disproportionate share status than are other rural hospitals. Hospitals that receive payments based on their updated costs are ineligible for those higher adjustments.

If all sole community hospitals received the regular PPS payment rather than their updated costs, total PPS payments would be about $100 million less in 2001 and $1.3 billion less for the 2001-2010 period. Those savings assume that SCHs would continue to be eligible for higher disproportionate share adjustments.

A primary objective of the SCH rules is to assist hospitals in locations where closings would threaten access to hospital care, but the federal support is not particularly well aimed at such essential providers. Moreover, whether an SCH actually receives higher payments under the special rules that permit payments to be based on a hospital-specific amount depends on whether its costs in any of the specified base years (1982, 1987, or 1996) were relatively high, not on its current financial condition.

If the special payment rules were eliminated, however, revenues of many sole community hospitals would be lower, which might cause financial distress for some of them. And because many SCHs are the sole providers of hospital services in their geographic areas, access to health care or the quality of care might be reduced in some rural locations.

Hospitals receive payments under Medicare's prospective payment system (PPS) for the operating costs of providing inpatient services to the program's beneficiaries. The payments are determined on a per-case basis; payment rates vary with the patient's diagnosis, which Medicare classifies within a system of diagnosis-related groups (DRGs), and the characteristics of the hospital. Those rates take into account reasonable variations in the treatment of patients with a given DRG and offer an incentive to the hospital to reduce the cost of treatment. PPS payments do not cover all services rendered to patients during the hospital stay. In particular, Medicare pays separately for physicians' services provided on an inpatient basis.

The Health Care Financing Administration (HCFA) has explored the feasibility of making a single global payment for high-cost, high-volume inpatient procedures. That payment would be lower than the separate payments that are now made for hospitals' operating costs and physicians' services. In a recent demonstration project involving heart bypass surgery, discounted payment rates were established through negotiations with participating hospitals in conjunction with teams of physicians. With a global payment, hospitals and physicians alike have an incentive to reduce operating costs while maintaining a satisfactory standard of care. The institutions hoped to offset the discounts in their Medicare payments by two means: improvements in efficiency (and their resultant cost savings) and increases (using new marketing efforts) in the volume of heart bypass patients. During the five-year project, Medicare outlays to the seven hospitals participating in the demonstration averaged about 10 percent less than would have been spent otherwise.

HCFA has also investigated ways to extend the global payment concept. One approach, similar to the heart bypass demonstration, identified other high-cost, high-volume inpatient procedures that might yield negotiated savings. (They included cataract surgery, coronary angioplasty, heart valve replacement, and joint replacement surgery.) That option might be attractive to hospitals, which could market themselves as "centers of excellence." However, such terminology would be controversial because it might be construed as suggesting that other hospitals did not offer high-quality care. Another disadvantage would be that only a modest number of institutions and high-cost procedures might become eligible for global payments. Expanding the use of global payments would yield savings of $100 million in 2001 and $1.2 billion for the 2001-2010 period.

Under Medicare's prospective payment system (PPS), payments for hospitals' operating costs for inpatient services provided to beneficiaries are determined on a per-case basis, according to preset rates that vary with the patient's diagnosis and the characteristics of the hospital. Payment rates are adjusted each year using an update factor that is determined, in part, by the projected increase in the hospital market-basket index (MBI), which reflects increases in hospital costs.

The Balanced Budget Act of 1997 reduced hospital update factors for 1998 through 2002. Specifically, the act froze the basic payment in 1998 and reduced the update by 1.9 percentage points in 1999, 1.8 percentage points in 2000, and 1.1 percentage points in 2001 and 2002. Without those reductions, the updates would have been 2.1 percent in 1998, 2.4 percent in 1999, 2.9 percent in 2000, and more than 3 percent each year in 2001 and 2002. (In several states, however, certain hospitals with negative PPS margins received a 0.5 percentage-point adjustment in 1998 and a 0.3 percentage-point adjustment in 1999.) After 2002, the update factor reverts to the full value of the MBI. If the factor was reduced to the MBI minus 1.8 percentage points in 2001 and stayed at that level throughout the 2001-2010 period, total savings during that time would be $75.8 billion.

In 1997, average profit margins for hospitals on Medicare inpatient services were about 17 percent. Moreover--although the data are not yet complete--MedPAC (the Medicare Payment Advisory Commission) estimates that despite the payment freeze in 1998 and the large reduction in the update factor for 1999, average Medicare inpatient profit margins exceeded 15 percent in both years. Thus, further reductions in update factors could be justified. The American Hospital Association, however, maintains that high inpatient margins reflect major efforts by hospitals to cut costs, which cannot continue indefinitely. Moreover, almost one-quarter of all hospitals have negative profit margins on Medicare inpatient services, so further reductions in payment update factors could cause considerable hardship for those facilities, especially as some hospitals are only now beginning to feel the effects of past payment reductions.

In 1992, Medicare revised its method of paying hospitals for their inpatient capital-related costs by replacing cost-based reimbursement with a prospective payment method. Under the prospective system, hospitals receive a predetermined amount for each Medicare patient to pay for capital-related costs, which include depreciation, interest, taxes, insurance, and similar expenses for buildings and fixed and movable equipment. The prospective system applies to about 5,000 hospitals paid under Medicare's prospective payment system for operating costs.

A fully prospective federal payment rate for capital costs is being phased in over 10 years. During the transition period, payments are determined by a complicated method based on a number of factors, including federal and hospital-specific payment rates. The federal and hospital-specific rates are increased annually. By 2001, all hospitals will receive the federal rate, adjusted for the hospital's mix of patients and certain other characteristics.

Analyses conducted by the Health Care Financing Administration (HCFA) suggest that the initial federal and hospital-specific rates were too high. The 1992 rates were based on actual 1989 and 1990 data (for the federal rate and hospital-specific rates, respectively) projected to 1992, but more recent data indicate that the rate of growth of capital costs between 1989 and 1992 was slower than expected. Moreover, the initial level of capital costs per case in 1989 was probably higher than would be optimal in an efficient market because of incentives provided by the Medicare payments. Factors such as changes in capital prices, the mix of patients treated by hospitals, and the "intensity" of hospital services contributed to the overestimate. On the basis of HCFA's analysis, the estimated 1992 capital costs would have been reduced by about 22 percent if those factors had been taken into account.

The federal rate was reduced by 7.4 percent in the Omnibus Budget Reconciliation Act of 1993 in a provision that expired in 1996. The Balanced Budget Act of 1997 reduced the federal rate by 17.8 percent for capital payments made to hospitals for patient discharges occurring in 1998 through 2002. A small part of that reduction, 2.1 percent, will be restored beginning in 2003. A further reduction of 5 percent (bringing the total reduction in capital payments to about the level estimated by HCFA) would yield savings of $300 million in 2001 and $4.6 billion for the 2001-2010 period.

Most hospitals would probably be able to adjust to the reductions by lowering their capital costs or partially covering them with other sources of revenue, because Medicare's payments for capital costs are a small share of hospitals' revenues--less than 5 percent of their total revenues from all sources. Hospitals that are in poor financial condition, however, might have difficulty absorbing the reductions. As a result, the quality of the care they offer might decline, and they might provide fewer services to people without insurance.

Medicare beneficiaries are responsible for certain deductible and coinsurance amounts when they receive hospital services. In calendar year 2000, the deductible amount is $766 per spell of illness, and beneficiaries must make coinsurance payments for inpatient care in excess of 60 days and for services furnished on an outpatient basis. Before enactment of the Balanced Budget Act of 1997 (BBA), if the hospital made a reasonable effort to collect the cost-sharing amounts from patients, Medicare would reimburse it for any remaining unpaid amounts. The BBA phased in a reduction in those bad-debt payments, cutting them to 55 percent of the amount that hospitals did not collect from beneficiaries. Eliminating all reimbursement for enrollees' bad debts would reduce Medicare's payments by $500 million in 2001 and $7.4 billion over the 2001-2010 period.

This option would give hospitals incentives to improve their collection efforts, but they would not be able to collect all of the money that their Medicare patients owed. In particular, low-income enrollees who were not covered by Medicaid might not be able to pay their hospital bills. As a result, this option would reduce revenues the most for those hospitals that were most likely to serve low-income Medicare patients. Moreover, a drop in their Medicare payments might lead some hospitals to cut back on the quality of their services or the amount of uncompensated care that they provide, or to raise the rates that they charge for the care of other patients.

The Medicare program pays a variety of providers, in addition to hospitals, for the bad debts of their Medicare patients. Providers incur such debts when Medicare patients do not pay the cost-sharing amounts that the program requires. Patients in skilled nursing facilities (SNFs), for example, must pay a coinsurance amount of $97 per day in calendar year 2000 for care received from the 21st through the 100th day of a benefit period (spell of illness). (A benefit period begins with the day the beneficiary is admitted to the hospital and ends after the patient has been out of the hospital or SNF for 60 straight days, or if the beneficiary remains in the SNF after 60 straight days without receiving skilled nursing care.) Providers that are eligible for bad-debt payments include SNFs, rural health clinics, and comprehensive outpatient rehabilitation facilities. The Balanced Budget Act of 1997 reduced Medicare's payments for hospitals' bad debts but did not affect bad-debt payments for nonhospital providers. This option would eliminate such payments for nonhospital providers, saving $600 million in 2001 and $10.3 billion over the 2001-2010 period.

As with hospitals, eliminating bad-debt payments would increase providers' incentives to improve their collection efforts, but some bad debts would remain. The policy could, therefore, cause financial problems for providers that serve a large number of low-income Medicare beneficiaries. Faced with unpaid bad debts, such providers might reduce the quality of their care or the amount of uncompensated care that they provide.

Medicare Supplementary Medical Insurance (Part B) paid providers about $4 billion in 1999 for certain outpatient drugs. Prescription drugs are covered under Part B when they must be administered under a physician's supervision, as is the case with many drugs requiring injection or infusion. Medicare also pays for drugs that must be delivered by durable medical equipment covered under the program. In addition, some oral chemotherapy and antinausea drugs for cancer patients as well as immunosuppressive drugs for organ transplant recipients are covered, as are certain vaccines.

Medicare payments for covered prescription drugs have varied over time and across settings of care. Since 1997, the amount Medicare has allowed as a reasonable charge for drugs delivered in physicians' offices and at home has been set at 95 percent of the average wholesale price, or AWP, which is a published list price established by the manufacturer. When several manufacturers make a product, the allowed charge is 95 percent of the median AWP among generic suppliers or the lowest brand-name AWP when that price is less than the median generic AWP. Medicare reimbursement for drugs delivered in hospital outpatient facilities is currently made on a cost basis, but under the Balanced Budget Act of 1997, Medicare is expected to implement a prospective payment system for hospital outpatient services beginning in July 2000. Most drugs will be exempted from the prospective payment system until at least 2002, however, and for them, the payment will be 95 percent of the AWP.

Because the AWP is a list price and not the actual price providers pay for drugs, pegging Medicare's payment to the AWP has meant that providers and suppliers could profit from administering or dispensing Medicare-covered drugs. The Inspector General of the Department of Health and Human Services reported that actual wholesale drug prices available to physicians were about 30 percent less than the AWP in 1997. This option would limit Medicare's reimbursements for prescription drugs by decreasing the allowed charge from 95 percent to 85 percent of the AWP and by limiting increases in the allowed charge for covered drugs to changes in the rate of inflation. (Changes in the allowed charge would track the consumer price index for all urban consumers, excluding food and energy.) As a result, Medicare Part B outlays would decrease by $7.2 billion between 2001 and 2010.

One disadvantage of the option is that it would encourage manufacturers to introduce new drugs at AWPs that were higher than they would otherwise be in order to restore the profit margins available to physicians and other suppliers. Physicians would prescribe newly introduced drugs more quickly as a result. Therefore, the option's effectiveness in limiting Part B spending growth would gradually erode as new drugs replaced older ones in the mix of covered drugs. Critics of the option also claim that the profit margins physicians now obtain when they administer drugs to Medicare patients subsidize the cost of drug administration, which is not now adequately reimbursed. Savings would be reduced and patient care might suffer if patients were diverted from physicians' offices to hospital outpatient settings, where Medicare reimbursements are higher. CBO's estimate includes a modest reduction in savings to account for that possibility.

Medicare offers insurance coverage for physicians' and hospital outpatient services through the Supplementary Medical Insurance (SMI) program. The program has a number of cost-sharing requirements. One way to achieve federal savings in SMI is to increase the deductible--that is, the amount that enrollees must pay for services each year before the government shares responsibility. The deductible is now $100 a year and has been increased only three times since Medicare began in 1966, when it was set at $50. In relation to average annual per capita charges under the SMI program, the deductible has fallen from 45 percent in 1967 to about 3 percent (projected) for 2000.

Increasing the SMI deductible for 2001 and later years according to the growth in SMI charges per enrollee would save $100 million in 2001, $2.6 billion over the five-year period, and $11.3 billion over the 10-year period. In 2001, the deductible would be $108.

An increase in the amount of the deductible would enhance the economic incentives for prudent consumption of medical care while spreading the impact of an increase in cost sharing among most enrollees. No enrollee's out-of-pocket costs would rise by more than $8 in 2001.

However, the additional out-of-pocket costs under this option might discourage some low-income enrollees who are not eligible for Medicaid from seeking needed care. In addition, costs to states would increase because their Medicaid programs pay the deductibles for Medicare enrollees who also receive benefits under Medicaid.

Medicare's cost-sharing requirements in its fee-for-service sector are varied and difficult for beneficiaries to understand. Moreover, in contrast to most private insurance plans, Medicare places no limit on the cost-sharing expenses for which enrollees may be liable. As a result, most fee-for-service enrollees seek supplementary coverage (either through their employers or by purchasing individual medigap plans) to protect them from the potentially catastrophic expenses they might be left with under Medicare. Those enrollees with the nearly first-dollar coverage that medigap plans provide no longer have financial incentives to use medical services prudently. Consequently, Medicare's costs are higher than they would be if there were no medigap supplements.

Medicare could simplify and limit cost-sharing requirements in the fee-for-service sector while also reducing federal costs. For example, the current complicated mix of cost-sharing requirements could be replaced with a single deductible, a uniform coinsurance rate of 20 percent for amounts above the deductible, and a cap on each beneficiary's total cost-sharing expenses-- whether they arose from Part A or Part B of the Medicare program. If those provisions were in place beginning in January 2001 with a deductible of $800 and a cap on total cost sharing of $2,000, federal savings would be $0.3 billion for 2001, $3.9 billion over five years, and $14.7 billion over 10 years. Those estimates assume that both the deductible and the cap would be indexed to growth in per capita benefits paid by Medicare.

For three reasons, such changes in Medicare's cost-sharing requirements would increase the incentives for enrollees to use medical services prudently. First, because about 40 percent of the medigap plans purchased do not now cover the deductible, more of the services used by those policyholders would be exempt from medigap coverage under Medicare's higher deductible. Second, over time, fewer enrollees would purchase medigap plans because their cost-sharing expenses would be capped under Medicare. Third, the uniform coinsurance rate on all services would encourage enrollees without supplementary coverage to consider relative costs appropriately when choosing among alternative treatments.

Although this option would generally reduce out-of-pocket costs for enrollees who had serious illnesses or were hospitalized during the year, it would increase out-of-pocket costs for most enrollees. On average, enrollees' cost-sharing expenses under Medicare would increase by about $45 a year in 2001. Expenses would fall for about 10 percent of enrollees, rise for about 70 percent, and be unchanged for all others. The option would also introduce cost-sharing requirements for services--such as home health care--that are not now subject to them, increasing administrative costs for the affected providers.

Savings from option 570-13-A could be substantially increased by restricting or prohibiting medigap coverage in addition to changing Medicare's cost-sharing provisions. Alternatively, some or all of the additional savings from restricting medigap coverage could be used to improve Medicare's coverage by reducing the deductible or cap.

If, for example, medigap plans were prohibited from covering any part of Medicare's new deductible (as discussed in option 570-13-A), savings would be $23.1 billion over five years and $63.5 billion over 10 years. By raising Medicare's deductible and prohibiting medigap plans from covering it, the incentives for more prudent use of health care services would be appreciably strengthened for enrollees who now have medigap plans. Those incentives would be still greater if medigap coverage was prohibited altogether. However, despite Medicare's new copayment cap, which would protect enrollees against very large cost-sharing expenses, some enrollees would object to any policy that denied them access to first-dollar coverage.

Medicare currently pays 100 percent of the approved fee for clinical laboratory services provided to enrollees. Medicare's payment is set by a fee schedule, and providers must accept that fee as full payment for the service. For most other services provided under Medicare's Supplementary Medical Insurance (SMI) program, beneficiaries are subject to both a deductible and a coinsurance rate of 20 percent.

Imposing the SMI program's usual deductible and coinsurance requirements on laboratory services would yield appreciable savings. If this policy was in place beginning on January 1, 2001, federal savings would be $490 million in 2001, $5.1 billion over five years, and $13.9 billion over 10 years.

In addition to reducing Medicare's costs, this option would make cost-sharing requirements under the SMI program more uniform and therefore easier to understand. Moreover, enrollees might be somewhat less likely to undergo laboratory tests with little expected benefit if they paid part of those costs.

However, enrollees' use of laboratory services would probably not be substantially affected because decisions about what tests are appropriate are generally left to physicians, whose judgments do not appear to depend on enrollees' cost-sharing liabilities. Hence, the Congressional Budget Office assumes that a small part of the expected savings under this option would stem from more prudent use of laboratory services, but the greater part would reflect the transfer to enrollees of costs now borne by Medicare. Billing costs for some providers, such as independent laboratories, would be higher under the option because they would have to bill both Medicare and enrollees to collect their full fees. (Currently, they have no need to bill enrollees directly for clinical laboratory services.) In addition, states' Medicaid costs would increase for enrollees who also received Medicaid benefits.

The use of home health services and the resulting costs are growing rapidly under Medicare. One reason for the unrestrained growth of such costs is that the services are free to enrollees--enrollees are not currently required to pay any portion of the cost of home health services under Medicare.

If a copayment of $5 was required for each home health visit covered by Medicare beginning in January 2001, net federal savings would be $0.6 billion in 2001, $6.0 billion over five years, and $17.2 billion over 10 years. If the copayment was $10, five-year savings would be $10.9 billion and 10-year savings would be $31.1 billion. Those estimates assume that the copayment would be indexed to the consumer price index after 2001.

This option would reduce Medicare's costs for home health care not only by shifting a small part of the cost per visit to users but also by reducing enrollees' use of the service--at least among the 15 percent of fee-for-service enrollees with no supplementary coverage for their cost-sharing expenses. However, little or no drop in use would be expected among the 85 percent of enrollees who have either Medicaid, medigap, or employment-sponsored supplementary coverage. Further, the option would increase private insurance premiums for the 35 percent of enrollees with medigap supplements, and it would increase Medicaid program costs on behalf of the 15 percent of enrollees who also receive Medicaid benefits. Moreover, it would increase the risk of very large out-of-pocket costs for those with no supplementary coverage.

About 35 percent of Medicare's fee-for-service enrollees purchase individual supplementary private insurance (medigap coverage) that covers all or most of the cost sharing that the Medicare program requires. On average, medigap policyholders use at least 25 percent more services than they would if they did not have first-dollar coverage. However, the federal government through Medicare and not medigap insurers pays most of the costs of those additional services.

Federal costs for Medicare could be reduced if medigap plans were prohibited from offering first-dollar coverage for Medicare's cost-sharing requirements. If, for example, medigap plans were barred from paying any portion of the first $1,500 of an enrollee's cost-sharing liabilities for calendar year 2001, use of medical services by medigap policyholders would fall and federal savings in 2001 would total $3.8 billion. Assuming that the medigap limit was linked to growth in the average value of Medicare's costs for later years, savings over the 2001-2005 period would total $37.8 billion. Over 10 years, savings would total $98.0 billion.

Only enrollees who have medigap policies would be directly affected by this option, and most of them would be financially better off under it. Because their medigap premiums would decrease more than their out-of-pocket liabilities would increase, most medigap enrollees would have lower yearly expenses under this approach. Indirectly, all enrollees might be better off because Medicare's premiums would be lower than under current law.

Medigap policyholders, however, would have to assume a higher level of financial risk for Medicare-covered services than they do now. Because they might feel more uncertain about their expenses, some policyholders might object to eliminating their option to purchase first-dollar coverage, even if in most years they would be financially better off. Moreover, in any given year, about a quarter of people with medigap policies would actually incur higher expenses under this option, and those with expensive chronic conditions might be worse off year after year. Finally, the decrease in use of services by medigap policyholders that would generate federal savings under this option might not be limited to unnecessary care, so the health of some policyholders might be adversely affected.

Benefits under Medicare's Supplementary Medical Insurance (SMI) program are partially funded by monthly premiums paid by enrollees, with the remainder funded by general revenues. Although the SMI premium was initially intended to cover 50 percent of the cost of benefits, premium receipts between 1975 and 1983 covered a declining share of SMI costs--falling from 50 percent to less than 25 percent. That drop occurred because premium increases were limited by the cost-of-living adjustment (COLA) for Social Security benefits (which is based on the consumer price index) but the per capita cost of the SMI program rose faster. Since 1984, premiums have been set to cover about 25 percent of average benefits for an aged enrollee, a provision that was made permanent in the Balanced Budget Act of 1997.

If the SMI premium was set to cover 30 percent of costs for 2001 and all years thereafter, outlay savings would be $2.9 billion in 2001, $23.3 billion over five years, and $60.9 billion over 10 years. The premium for 2001 would be $59.10 a month instead of $49.30. Those estimates assume a continuation of the current hold-harmless provision, which ensures that no enrollee's monthly Social Security benefit will fall as a result of the Social Security COLA (which is based on the whole benefit) being smaller than the SMI premium increase.

Most SMI enrollees would pay a little more under this option, in contrast to proposals--such as increasing cost-sharing requirements--that could substantially raise the out-of-pocket costs of those who become seriously ill. This option need not affect enrollees with income below 120 percent of the federal poverty threshold because all of them are eligible to have Medicaid pay their Medicare premiums. (Some people who are eligible for Medicaid do not apply for benefits, however.)

Low-income enrollees who are not eligible for Medicaid could find the increased premium burdensome. A few might drop SMI coverage and either do without care or turn to sources of free or reduced-cost care, which could increase demands on local governments. In addition, states' expenditures would rise because states would pay part of the higher premium costs for those Medicare enrollees who also receive Medicaid benefits.

Instead of increasing the basic premium to 30 percent of costs for all enrollees in the Supplementary Medical Insurance (SMI) program (see option 570-17), this option would collect relatively more from higher-income people. For example, people with modified adjusted gross income of less than $50,000 and couples with income lower than $75,000 would pay only the basic premium, set at 25 percent of SMI costs per aged enrollee. Premiums would rise progressively for higher-income enrollees, however. The maximum total premium would be set to cover 50 percent of costs for people with income exceeding $100,000 and for couples with income exceeding $150,000. The income- related premiums would have to be collected through the income tax system so that rates could be aligned with income. Current premiums are deducted automatically from Social Security checks for most enrollees.

If this option was in place in calendar year 2001, savings would total $480 million in fiscal year 2001, $8.8 billion over five years, and $28.2 billion over 10 years. Those estimates assume that the current hold-harmless provisions would continue only for people subject to the basic 25 percent premium. (The hold-harmless provisions ensure that no enrollee's Social Security check will decrease because an increase in the SMI premium exceeds the cost-of-living adjustment.)

Most SMI enrollees would be unaffected by tying a portion of the program's premium to income. Roughly 86 percent of enrollees would face the basic 25 percent premium, about 3 percent would pay the maximum premium, and 11 percent would pay a premium somewhere in between.

Enrollees subject to the income-related premium would pay substantially more, however. The maximum monthly premium for 2001 would be $98.60 instead of the $49.30 premium projected under current law. That increase might lead some enrollees to drop out, although it is estimated that fewer than 0.5 percent would do so. Those enrollees with retirement health plans that do not require Medicare enrollment (mainly, retired government employees) would be most likely to drop out. Some healthy enrollees who have no other source of health insurance might do so as well, if they were not averse to the risk that they might incur large health care costs.

Under current law, the normal retirement age (NRA) for Social Security will gradually increase from 65 to 67 in the first quarter of the next century. However, eligibility for Medicare based on age will remain at 65. Because the two programs affect the same population and because eligibility is based on the same work history, some people have argued that the age requirements should be the same.

If the age at which a person became eligible for Medicare was raised in step with increases in the NRA for Social Security, the first cohort to be affected would be people who turned 65 in 2003--for that group, eligibility for Medicare would be delayed by two months. The age of eligibility would be increased by an additional two months each year through 2008 and then remain at 66 for 12 years. Beginning in 2020, the age of eligibility would again increase by two months a year until it reached 67 in 2025. Under that option, federal budget savings would total $390 million in 2003, $3.2 billion through 2005, and $29.5 billion through 2010. Reduced spending for Medicare would be partially offset by increased spending under Medicaid, the Federal Employees Health Benefits program, and the Civilian Health and Medical Program of the Uniformed Services (reflected in the savings estimates). In addition, off-budget outlays for Social Security would fall by $8.5 billion over the 10-year period because some people who were affected would delay retirement. (That drop in costs is not reflected in the estimates.)

The same reasons that have been used to justify increasing the NRA for Social Security apply to this option as well. Life expectancy has increased substantially since Social Security and Medicare began, and a majority of workers now live well beyond the age of eligibility. When Social Security was established in 1935, average life expectancy at birth was less than 65 years; now average life expectancy is greater than 75 years. Unless changes are made in those programs, longer expected lifetimes, together with the population bulge of the baby-boom generation, will increase costs enormously under Social Security and Medicare after 2010. Only three general options for change are available: reduce the number of people eligible for benefits, reduce benefits per eligible person, or increase taxes. As a practical matter, it is likely that all three options will be called into play.

However, about 70 percent of Social Security beneficiaries retire before the normal retirement age--generally at Social Security's early retirement age of 62, which entitles them to benefits at a reduced level. Increasing Medicare's age of eligibility would also raise the number of years during which early re-tirees would be at risk of having no health insurance--just when their need for health care would be expected to increase significantly and their access to private individual insurance would be limited.

One way to alleviate the problem that early retirees may have in continuing health insurance coverage until they are eligible for Medicare would be to introduce an early age of eligibility (62) for nondisabled retirees. (Disabled people are already eligible for Medicare regardless of their age.) That change would make the conditions for age-based eligibility under Medicare wholly consistent with those for Social Security.

Allowing people to "buy in" to Medicare at age 62 beginning in January 2001, together with the gradual move to a later normal age of eligibility (67) described in option 570-19-A, would reduce federal costs by $1.3 billion over the 2001-2005 period and by $23.8 billion through 2010. Social Security costs would increase in the early years when only the buy-in was in place, but (off-budget) savings would occur after 2004 as delays in retirement due to the increase in the eligibility age for Medicare more than offset earlier retirement among those taking advantage of the buy-in option. Those estimates assume that people who used the early buy-in option would pay an actuarially fair premium for their age group during the buy-in years. The estimates also assume that once buy-in participants reached the normal age of eligibility, they would pay a premium surcharge to compensate for any excess costs incurred during their buy-in years. (Buy-in participants are likely to be more costly to Medicare than the average person in their age group.)