Medicare, the federal health insurance program for the elderly and some people with disabilities, is divided into three basic programs. Part A, Hospital Insurance, pays for inpatient care in hospitals and skilled nursing facilities. It also pays for some home health care and hospice services. Part B, Supplementary Medical Insurance, pays for physicians' services, hospital out-patient services, some home health care, and other services. Part D, the prescription drug benefit added in 2006, is used to reduce what participants pay for medicine. (Medicare's Part C specifies the rules under which private health care plans can assume responsibility and be paid for providing benefits covered under Parts A, B, and D.)

Total Medicare spending has grown at an average annual rate of about 9 percent in recent years. The Congressional Budget Office estimates that net outlays will total $370 billion in 2007, including discretionary outlays of about $5 billion for Medicare's administrative costs. Roughly $60 billion will be offset by premium payments (mostly from participants in Parts B and D), by payments from states, and by recovery of improper payments to providers. In the next few years, Medicare enrollment—and its cost—will expand substantially as the first members of the baby-boom generation become eligible because of age or disability.

Under current law, the age at which workers become eligible for full Social Security retirement benefits—the normal retirement age (NRA)—is gradually increasing until it reaches 67 for people who were born in 1960 or later. (Workers can receive a reduced retirement benefit as early as age 62, however.) The eligibility age for Medicare will remain at 65, although people can qualify for coverage earlier if they are disabled or have end-stage renal disease. Because the two programs affect the same population, some people have argued that the eligibility age for Medicare should be identical to Social Security's NRA.

This option comprises two alternatives for raising the eligibility age for Medicare. Each alternative assumes that the eligibility age would not be increased until 2017, so people who are currently nearing retirement would not be affected. The first alternative would increase the eligibility age by two months every year beginning in 2017 until it reached 67 in 2028, where it would stay indefinitely. Although the increases under that alternative are consistent with increases currently scheduled for Social Security's NRA, the Medicare eligibility age would remain below Social Security's NRA until 2028 (because the NRA increases under Social Security started sooner). The second alternative would increase the eligibility age by two months every year beginning in 2017 until it reached 70 in 2046, at which point it would stabilize. That alternative is analogous to the option for raising Social Security's NRA (see Option 650-5), but it would be phased in more slowly and would not raise the eligibility age above 70.

In 2050, Medicare spending would fall by about 3 percent under the first alternative and by about 10 percent under the second. Because those estimates are not within the Congressional Budget Office's 10-year budget window, no year-by-year table is shown. Spending would fall by less than enrollment because younger beneficiaries are healthier and less costly than average.

The reduced spending for Medicare would be partially offset by higher spending under Medicaid and the Federal Employees Health Benefits program—both of which would pick up part of the health care costs of those beneficiaries whose eligibility for Medicare had been delayed. Spending under the military's Tricare For Life program would decline, however, because eligibility for that program is limited to people who are enrolled in Medicare.

The primary rationale for this option is that it would restrain the growth of Medicare spending, which would ease long-term budgetary pressures. Life expectancy has risen since the Medicare program began in 1965, and the life expectancy of 65-year-olds is expected to continue increasing. Therefore, on average, people will spend a longer time covered by Medicare, which will boost the program's costs. In addition, raising the eligibility age would reinforce incentives created by increases in Social Security's NRA for people to delay retirement. Disability among the elderly has declined over time, and jobs are generally less physically demanding, suggesting that a larger fraction of the population may be capable of working past age 65. Many who do so could have access to employment-based insurance.

An argument against this option is that many workers retire before age 65. For those early retirees, raising the eligibility age for Medicare would lengthen the time they might be at risk of having no health insurance. Furthermore, raising the eligibility age for Medicare would shift costs that are now paid by that program to individuals and to employers that offer health insurance to their retirees. Those higher costs might lead more employers to reduce or eliminate health coverage for their retirees. Also, raising the eligibility age for Medicare would strengthen the incentive for people to apply for Social Security disability benefits, reducing the net savings to the federal government.

The Medicare Advantage program is the vehicle through which private health plans can participate in Medicare. Plans that want to participate in the program submit bids reflecting the per capita payment for which they are willing to provide Medicare's covered benefits. The government compares those bids with benchmarks that are determined in advance through statutory rules. Plans are paid their bids (up to the benchmark) plus 75 percent of the amount by which the benchmark exceeds their bid. Plans must return that 75 percent to beneficiaries as additional benefits or rebates on their Medicare premium. Plans whose bids are above the benchmark are required to charge enrollees the full difference between the bid and benchmark as an add-on to their regular Medicare premium. (Private plans submit separate bids to provide Medicare's prescription drug benefit; this option pertains to their bids for all other Medicare benefits.)

Benchmarks are established for each county and are required to be at least as high as local per capita spending in Medicare's fee-for-service (FFS) program. (The county-level benchmarks are also used to establish benchmarks for the program for regional preferred provider organizations; see Option 570-4.) In many counties, the benchmark is higher than per capita FFS spending, in some cases substantially. Benchmarks were derived from a payment mechanism for private plans that was established in the Balanced Budget Act of 1997 and modified through subsequent legislation. Those rules resulted in rates in many counties that are higher than local per capita spending in the FFS program.

This option would set the benchmark in each county equal to local per capita Medicare fee-for-service spending. That change would reduce Medicare spending by about $8.1 billion in 2008 and $64.8 billion over five years.

An argument in favor of this option is that the Medicare program should be neutral as to whether beneficiaries decide to enroll in private plans or remain in the fee-for-service sector. (Most beneficiaries—about 82 percent—are enrolled in the FFS program.) The current payment system gives an advantage to private plans because they can operate in areas where their bids exceed FFS spending levels and, if their bids are less than the benchmark, provide additional benefits to attract enrollees. Under that system, Medicare pays more for enrollees in some private plans than it would have paid if they had remained in the FFS sector. Setting the benchmark equal to per capita FFS spending in each county would encourage private plans to operate only in areas where they could provide Medicare services at a lower cost than the FFS sector, without encouraging them to operate in areas where they could not.

An argument against this option is that, in many geographic areas, it would reduce the revenue that private plans receive from Medicare, which could lead many plans to limit the benefits they offer, raise their premiums, or withdraw from the program. Another argument is that private plans should not be expected to provide Medicare services in all markets at a cost that is less than per capita FFS spending because Medicare may be able to use its market power to set FFS payment rates at levels below those that are determined through private-market forces. Below-market payments to health care providers may result in a less-efficient allocation of resources than would be achieved if more beneficiaries were enrolled in private plans that paid providers at rates determined in the market.

Hospitals with teaching programs receive additional payments from Medicare for costs associated with graduate medical education (GME). One component of those additional payments covers the direct costs of GME (such as residents' compensation). Two other components are indirect medical education (IME) adjustments to Medicare's payments for hospitals' operating costs and for their capital-related costs; they are designed to account for the fact that teaching hospitals tend to have greater expenses than other hospitals for various reasons. (For instance, teaching hospitals typically offer more technically sophisticated services and treat patients with more complex conditions than other hospitals do.) Medicare makes those three types of medical education payments to hospitals for the inpatient stays of all Medicare beneficiaries, including those who are enrolled in private health plans that participate in the Medicare Advantage program.

About 18 percent of Medicare beneficiaries are enrolled in Medicare Advantage plans, which assume responsibility and financial risk for providing Medicare benefits. The government's maximum payment, or benchmark, for an enrollee in such a plan is set for each county and updated annually. Benchmarks were derived from a setof payment rates for private plans that were in effect in 2004; under that system, the rate for each county was the greatest of four amounts: a minimum or "floor" rate; a blend of a local rate and the national average rate; a minimum increase from the previous year's rate; and local per capita spending in Medicare's fee-for-service (FFS) program. Beginning in 2005, the benchmark (or rate) in each county is equal to the previous year's benchmark (or rate) updated by the national growth in per capita Medicare spending or by 2 percent, whichever is greater. The government is required to reestimate local per capita FFS spending at least once every three years, and when it does so, the benchmark in each county is the greater of that new estimate of local FFS spending or the previous year's benchmark updated in the usual manner.

The estimates of local per capita FFS spending in 2004 and the revised estimates generated in later years include payments for IME even though the Medicare program makes IME payments directly to teaching hospitals for the inpatient stays of Medicare Advantage enrollees. As a result, the Medicare program is paying twice for IME for those enrollees—first, as an allowance for IME payments in the benchmark and, second, as a payment to teaching hospitals.

This option would remove payments for IME from the benchmarks for private plans, leaving the payment to teaching hospitals as the only compensation for IME. Making that change would reduce Medicare outlays by $700 million in 2008 and by $5.2 billion through 2012.

A rationale for this option is that it would reduce Medicare expenditures. According to proponents, there is no basis for making double payments for IME for Medicare Advantage enrollees.

A potential drawback of this option is that eliminating the double payment for IME would reduce the revenue that private health plans earn from Medicare, which could lead some plans to limit the benefits they offer, raise their premiums, or withdraw from the program. Plan withdrawals could reduce the number of Medicare beneficiaries with access to private health plans and the additional benefits they provide.

The Medicare Modernization Act of 2003 (MMA) established incentives for preferred provider organizations (PPOs) to participate in Medicare and serve broad regions of the country. One of those incentives is a stabilization fund that the Medicare program may use to increase the maximum payment amounts (or benchmarks) for regional PPOs to encourage them to enter into and remain in the Medicare Advantage program. Regional PPOs, like other types of Medicare Advantage plans, submit bids reflecting the per capita payment for which they are willing to provide Medicare's covered benefits. The plans are paid their bids (up to the benchmark) plus 75 percent of the amount by which the benchmark exceeds their bid. Plans must return that 75 percent to beneficiaries as additional benefits or as rebates on their Part B or Part D premiums. Plans whose bids are above the benchmark are required to charge enrollees the full difference between the two amounts as an additional premium for the Medicare benefit package.

The benchmarks for regional PPOs are a weighted average of two components: a statutory component and a bid component. The statutory component is the weighted average of the county-level benchmarks in the region, with each county weighted by the number of Medicare beneficiaries who live there. (The county-level benchmarks are determined each year by statutory rules. They are required by law to be at least as high as per capita fee-for-service spending in the county; in many counties, they are higher than such spending.) The bid component is a weighted average of the bids of the PPOs in the region, with each PPO weighted by its enrollment. To determine the regional benchmark, the statutory component is weighted by the number of Medicare beneficiaries nationally who are enrolled in the fee-for-service program, and the bid component is weighted by the number who are enrolled in private plans. (In contrast, for all other types of Medicare Advantage plans, the benchmark is the weighted average of the county-level benchmarks in the plan's service area, with each county weighted by the number of the plan's enrollees who live there. The plans' bids do not affect their benchmarks.)

The MMA established the stabilization fund for the regional PPO program and mandated that $10 billion be available to the fund from 2007 through 2013. The Tax Relief and Health Care Act of 2006 reduced the amount of money that will be available to the fund to $3.5 billion and limited the period during which it can be used to 2012 and 2013. The fund will also receive a portion (12.5 percent) of the difference between the bids of regional PPOs and their benchmarks when those bids are below the benchmarks. The Medicare program can use the stabilization fund to increase the benchmarks in regions that were not served by regional PPOs in the previous year and, under certain conditions, to increase the benchmarks in regions in which PPOs inform the government that they intend to leave the program. The stabilization fund can also be used to increase the benchmarks for organizations that participate in every region.

This option would eliminate the stabilization fund for the regional PPO program. Because no money will be available to the fund until 2012, this option would have no effect on Medicare spending until that year. It would reduce Medicare spending by $1.6 billion in 2012 and by $3.5 billion over 10 years.

An argument in favor of this option is that the stabilization fund could give regional PPOs an advantage relative to other plans that participate in Medicare Advantage. In addition, the evidence suggests that increased benchmarks may not be necessary to encourage regional PPOs to participate in Medicare—such plans participated in 21 of the 26 regions in 2006, with no payment from the fund.

An argument against this option is that it could decrease the number of regional PPOs that participate in Medicare and reduce the number of regions that are served by such plans. Without the increased benchmarks, PPOs may not be willing to participate in some regions, particularly those that are largely rural. (Health plans typically find it more costly to develop provider networks in rural areas because of the lack of competition among providers there.) Consequently, elimination of the stabilization fund could reduce the number and types of private plans that are available to Medicare beneficiaries; as a result, the additional benefits and premium rebates that many plans offer might be reduced or eliminated for beneficiaries in some areas.

Medicare pays hospitals for the inpatient stays of its beneficiaries through a prospective payment system. Under that system, hospitals with teaching programs receive additional amounts for costs associated with graduate medical education (GME). One component of the education-related payment is called direct GME, which covers a portion of a hospital's costs for residents' compensation and institutional overhead. Payments are made on the basis of a hospital's 1984 cost per resident (indexed for changes in consumer prices) and Medicare's share of inpatient days. Direct GME payments for physician residents, received by about one-fifth of U.S. hospitals, totaled $2.3 billion in 2006. (Option 570-6 covers Medicare's indirect payments for medical education.)

Under this option, hospitals' direct GME payments would be set at 120 percent of the national average salary paid to residents in 1987 and updated annually for changes in consumer prices since 1987. In effect, this option would reduce payments for teaching and overhead while continuing payments for residents' compensation. It would also maintain the current practice of reducing payments for residents who have exceeded their initial period of residency. (Such a resident is treated as one-half of a full-time-equivalent resident.) The savings from this option would total $900 million in 2008 and $5.9 billion over five years.

An argument in favor of this option is that market incentives appear sufficient to entice people to enter medicine, so a reduction in the federal subsidy for medical education may be warranted. In addition, because hospitals benefit from the services that residents provide, they should shoulder more of the costs of residents' training. Although residents would bear more of the cost of their education if hospitals responded by cutting residents' salaries or benefits, the training ultimately enables them to earn higher future incomes.

An argument against this option is that reducing the federal subsidy for graduate medical education could lead some hospitals to cut the resources they devote to medical training, possibly compromising the quality of their education programs.

Under Medicare's prospective payment system for in-patient medical services, hospitals with teaching programs receive additional funds for costs related to graduate medical education (GME). One part of the additional payment to teaching hospitals covers the costs of indirect medical education (IME), or those costs that are not attributable either to residents' compensation or to other direct costs of running a teaching program. Examples of IME expenses are the added demands placed on staff as a result of teaching activities and the greater number of tests and procedures ordered by residents. IME payments also compensate for the higher proportion of severely ill patients treated at teaching hospitals. (Option 570-5 discusses direct GME payments.)

The IME adjustment provides teaching hospitals with about 5.5 percent more in payments for inpatient services for every increase of 0.1 in the ratio of full-time residents to the number of beds. (The adjustment for 2007 is 5.35 percent.) This option would lower the IME adjustment to 2.2 percent—an amount that the Medicare Payment Advisory Commission has estimated would more accurately reflect indirect costs—saving $3.9 billion in 2008 and $21.6 billion through 2012.

An argument in favor of this option is that it would bring payments into line with actual teaching costs, thus reducing the federal subsidy without unduly affecting teaching activity. It also would remove an incentive for hospitals to have a higher number of residents than is necessary.

Possible drawbacks of this option are that a lower teaching adjustment could prompt teaching programs to train fewer residents or devote less time and resources to beneficial educational activities. Also, because some teaching hospitals use a portion of the additional payments they receive to fund charitable care, reducing those payments could limit the number of low-income patients they were able to serve or decrease the quality of care they were able to provide.

Under the prospective payment system for inpatient hospital services, Medicare pays hospitals an amount for each discharged patient that is intended to compensate hospitals for capital-related costs such as depreciation, interest, rent, and other expenses related to property. Hospitals with teaching programs receive additional capital-related payments that are made on the basis of "teaching intensity," which is measured as the ratio of residents to the average daily number of hospitalized patients. An increase of 0.1 in that ratio raises a hospital's capital-related payment by 2.8 percent.

This option would eliminate those extra payments to teaching hospitals, saving the Medicare program $400 million in 2008 and $2.3 billion over five years.

One argument in favor of this option is that paying teaching hospitals more than nonteaching hospitals for treating otherwise similar patients may promote inefficient practices at teaching 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. According to that argument, if residents' training raised the costs of patient care for a hospital, the hospital should bear those costs in order to encourage an efficient amount of training. Finally, although residents would bear more of the cost of their education if hospitals responded by cutting their salaries or benefits, their training would still enable them to eventually earn a high income.

A possible drawback of this option is that it could prompt teaching programs to train fewer residents or to devote less time and resources to beneficial educational activities. Also, because some teaching hospitals use a portion of their additional payments to fund charity care, reducing those payments could limit the number of low-income patients they were able to serve or decrease the quality of care they were able to provide.

Hospitals that serve a disproportionately large number of low-income patients can receive higher payment rates under Medicare than other hospitals do. The Medicare disproportionate share hospital (DSH) adjustment was introduced in 1986 to account for what were assumed to be the higher costs of treating Medicare patients in such hospitals. The DSH adjustment has also come to be seen as a way to protect low-income patients' access to care by providing financial support to hospitals that serve a large share of people from low-income populations.

Between 1992 and 1997, annual outlays for Medicare DSH payments rose from $2.2 billion to $4.5 billion. Restrictions established by the Balanced Budget Act of 1997 caused those outlays to decline for a few years, but they resumed growing in 2000. In 2003, the Medicare Modernization Act further boosted DSH payments to rural and small urban hospitals by adjusting the payment formulas. As a result, Medicare DSH payments totaled $9.5 billion in 2006.

This option would convert DSH payments into a block grant to the states. In 2008, each state's grant would be 10 percent less than the estimated sum of Medicare DSH payments made to hospitals in that state in 2006. In subsequent years, the block grant would be indexed to the change in the consumer price index for all urban consumers minus 1 percentage point. In return for the lower Medicare DSH payments, states would be granted increased flexibility in how they used their DSH funds. Those changes would decrease Medicare outlays by $1.2 billion in 2008 and by $11.2 billion over five years. (The estimated savings include the lower payment updates that plans participating in the Medicare Advantage program would receive.)

An argument in favor of this option is that the added flexibility provided to states under this option could result in DSH funds being targeted more appropriately and equitably to facilities and providers that serve low-income populations. For example, rather than going solely to hospitals, such funds might also be used to support outpatient clinics that treat low-income patients.

An argument against this option is that the net reduction in federal payments to hospitals, unless made up for by states with their own funds, would result in some hospitals' receiving less public funding than they do now. That drop in funding could limit the number of low-income patients they were able to serve or decrease the quality of care they were able to provide.

Medicare compensates hospitals for their operating costs tied to providing inpatient services to Medicare beneficiaries under a prospective payment system (PPS). Payments are determined on a per-case basis, according to preset rates that vary with a patient's diagnosis and the characteristics of the hospital. Medicare adjusts those payment rates each year using an update factor that is determined in part by the projected rise in the hospital market-basket index (MBI), which reflects increases in hospitals' costs per case or their unit costs.

Under current law, hospitals that submit quality performance data each year to the Department of Health and Human Services will receive the full MBI update for that year. Hospitals must report on a set of measures approved by the Hospital Quality Alliance (HQA). The current set of measures—which is continuously being expanded by the HQA—reflects recommended treatments for three serious medical conditions (heart attack, heart failure, and pneumonia) and guidelines for patient safety (such as the prevention of surgical infection). Hospitals that do not submit the required information will receive the MBI update factor minus 2 percentage points. That reduction will apply only for the year in which the hospital does not submit the required information and will not be taken into account in subsequent years. (The Congressional Budget Office expects that nearly all hospitals will submit the required data and receive the full update.)

This option would reduce the Medicare PPS update factor set under current law by 1 percentage point. That lower rate would take effect in 2008 and continue through at least 2017. Savings from this option would total $1 billion in 2008 and $17.8 billion over five years.

Supporters of this option reason that granting the full MBI update factor will overcompensate hospitals for their average growth in operating costs. To the extent that the MBI is intended to approximate how much providers' costs would rise if the quantity, quality, and mix of inputs they use to provide care remained constant, the MBI would generally overstate cost inflation because of productivity improvements (such as the tendency of providers to adopt cost-saving technological advances in response to the fixed payments established under the PPS).

Critics of this option contend that Medicare's payments for inpatient services should not be reduced without carefully evaluating the adequacy of payments for other hospital services (such as outpatient care). The overall Medicare margin (which includes both inpatient and outpatient care) has decreased continuously since 2000 (falling to -3 percent in 2004), and further reductions in the update factor could cause considerable hardship for hospitals.

In 1992, Medicare changed its method of paying hospitals for capital expenses associated with providing in-patient services; specifically, it switched from a cost-based reimbursement system to a prospective payment system. Under the revised system, hospitals receive a predetermined amount to cover capital-related costs for every Medicare patient treated at their facility. (Those costs include depreciation, insurance, interest, taxes, and similar expenses for the maintenance of buildings and the purchase and upkeep of equipment.) The prospective payment system for capital-related costs applies to hospitals that are also reimbursed by Medicare for inpatient operating costs under that system. A hospital's payment rate is adjusted to reflect its case mix of patients and other characteristics, such as whether the hospital is new and where it is located.

Analyses by the Centers for Medicare and Medicaid Services (CMS), which administers the Medicare program, suggest that the rates for capital payments set in 1992 were too high. Those rates were based on 1989 data projected to 1992; but in actuality, capital costs grew more slowly than expected during those years. Moreover, the level of capital costs per case that was used to set rates in 1989 was probably higher than would be optimal because of incentives created under cost-based reimbursement before 1992. Factors such as changes in capital prices, the mix of patients treated at a given hospital, and the "intensity" (technological complexity) of hospital services contributed to the inflated estimates for the initial capital rates set in 1992, which the Medicare Payment Advisory Commission and CMS calculated were between 15 percent and 28 percent, with an average of about 22 percent. In response, as part of the Balanced Budget Act of 1997, lawmakers reduced by 17.8 percent the federal rate for capital payments made to hospitals for patient discharges occurring between 1998 and 2002. (A small part of that reduction, 2.1 percentage points, was restored effective October 1, 2002.)

This option would further reduce the prospective payment rate for hospitals' capital-related costs by 5 percentage points. That change would lower Medicare outlays by $400 million in 2008 and $2.4 billion through 2012.

A rationale for this option is that it would reduce the overestimate that might remain in Medicare's capital payment rates. Moreover, since Medicare's payments for capital-related costs represent a small share—about 5 percent—of hospitals' total revenues, most hospitals would probably be able to adjust to the reductions by lowering their capital costs or by partially covering those expenses through other sources of revenue.

An argument against this option is that hospitals in poor financial condition could have difficulty absorbing the reductions. As a result, the quality of the care that they offered could decline, and they might provide fewer services to people without health insurance.

In 2006, Medicare paid about $13 billion for home health care services (which include skilled nursing care, physical and speech therapy, and home health aide services for beneficiaries deemed to be homebound). Medicare spending on home health services grew rapidly in the late 1980s and early 1990s, when home health agencies were reimbursed separately for each home health visit, but it fell sharply after a new payment system was implemented under the Balanced Budget Act of 1997. Since 2000, however, Medicare spending on home health care has again been increasing rapidly.

Home health agencies currently receive a single payment from Medicare for providing all covered services to an individual beneficiary for a 60-day period (known as a home health episode). The Centers for Medicare and Medicaid Services sets the payment rates for different types of episodes prospectively, meaning that payment rates are set in advance to reflect the expected costs of each episode and are not determined by the costs that home health agencies actually incur. In calendar year 2007, the base payment rate per home health episode is $2,339. Under current law, that rate is updated from year to year, partly on the basis of annual changes in the prices of inputs (such as wages for home health aides).

The Medicare Payment Advisory Commission, or MedPAC, has calculated that among freestanding home health agencies, the aggregate Medicare margin—the excess of Medicare payments over providers' costs expressed as a percentage of payments—was high in 2004, at about 16 percent. (MedPAC did not report the aggregate Medicare margin for hospital-based agencies in 2004.) MedPAC projects that the aggregate Medicare margin will remain at a high level (15 percent) in 2006, even though the Deficit Reduction Act of 2005 eliminated the 2006 update for home health agencies. The continuing high margins appear to be the result of reductions in home health agencies' costs in response to the incentives created by the new prospective payment system.

This option would freeze the base payment for each home health episode under Medicare at its calendar year 2007 level ($2,339) through 2012, with the goal of gradually narrowing the gap between payments and costs. That change would reduce federal outlays by $300 million in 2008 and by $8.5 billion over five years.

A rationale for this option is that margins for home health care are likely to remain high under current law. MedPAC estimates that home health agencies' costs per episode have grown by less than 1 percent a year in recent years, and if that trend continues, home health agencies would still receive more than adequate margins under this option.

A drawback of this option is that it could reduce access to home health services for Medicare beneficiaries. Home health agencies that had substantially higher costs than average and that were not able to reduce their operating expenses sufficiently would cease participating in the program. As a result, some beneficiaries might have difficulty obtaining home health services. Also, although MedPAC has not thus far identified problems with the quality of care provided under the new payment system, lower payment rates could lead some home health agencies to reduce the level or quality of the services they provide.

Medicare's coverage of post-acute care is generally limited to patients who require skilled nursing care or rehabilitation. Post-acute care is offered by four types of providers: skilled nursing facilities, home health agencies, long-term care hospitals, and inpatient rehabilitation facilities. In 2004, Medicare outlays for post-acute care accounted for over 12 percent of total Medicare spending. In each of the four post-acute care settings, providers are paid by Medicare under prospective payment systems in which payment rates reflect "base" payment rates. The payment for a specific case equals the base payment rate adjusted to reflect local practice costs, the clinical characteristics of the patient, and other factors.

Annual increases in Medicare's base payment rates are referred to as "update factors." Under current law, update factors generally are determined by increases in the prices of various "inputs," such as labor and equipment, that medical providers use to produce medical services. Those increases in input prices are measured by market-basket indexes, which combine various price increases into a single number for each type of provider.

This option would change the update factors for each type of post-acute care provider to equal the market basket index minus 1 percentage point for each year beginning in 2008. This option would reduce Medicare outlays by $350 million in 2008 and by $8.1 billion over five years.

An argument in favor of this option is that Medicare's payment rates for post-acute care have been found, in general, to be more than adequate relative to providers' costs. The Medicare Payment Advisory Commission (MedPAC) came to that conclusion in its March 2006 report to the Congress. MedPAC recommended that the Congress eliminate the update to payment rates for all types of post-acute care providers for 2007 and stated that doing so would be unlikely to harm beneficiaries' access to post-acute care. A second argument for this option is that it could provide a stronger incentive for post-acute care providers to increase their efficiency and reduce their operating costs.

An argument against this option is that the reduced federal payments that would result might increase the incentive of post-acute care providers to avoid admitting to their facilities patients with complex conditions who require costly care. Reducing update factors, therefore, might lead to certain patients having difficulty obtaining post-acute care. To the extent that patients faced limited access to post-acute care, they might either remain longer in a short-stay hospital, return home without receiving post-acute care, or be discharged to receive long-term care not covered by Medicare. By reducing the revenue of providers, this option might also limit their ability to provide high-quality care.

Each year, according to federal law, Medicare sets fees for physicians' services using the "sustainable growth rate" (SGR) mechanism. That mechanism establishes both yearly and cumulative targets for Medicare's combined spending for physicians' services and those services furnished "incident to" (in connection with) a visit to a physician (for instance, diagnostic laboratory services or physician-administered drugs). Those targets are updated annually to reflect inflation, overall economic growth, the increase in the number of Medicare enrollees in the fee-for-service sector, and any changes in Medicare outlays that stem from new laws or regulations. If spending exceeds the target (measured on both an annual and a cumulative basis), as it currently does, the SGR mechanism is designed to reduce payment rates to physicians each year so that cumulative spending and the cumulative target eventually converge. (The reverse happens when spending is below the target.)

Since 2002, Medicare spending for physicians' services has consistently been above the targets established by the formula and, consequently, the SGR mechanism has called for reductions in physician payment rates. In 2003, physicians were scheduled to receive a negative 4.4 percent update, after having seen a drop in payment rates of 4.8 percent in 2002. Lawmakers responded to that imminent reduction by boosting the cumulative target, thereby producing a 1.6 percent increase in payment rates for physicians= services for 2003. Since 2003, legislation has overridden scheduled payment reductions each year, further widening the gap between the target for cumulative spending and actual cumulative spending. As a consequence, unless overridden by legislation again, payment rates under the SGR mechanism will be reduced by about 10 percent in 2008 and around 5 percent annually for at least several years thereafter.

Given that the SGR formula's results have been over-ridden in each of the past five years, there is interest among policymakers in considering other payment mechanisms. The option considered here presents four alternatives, three of which would adjust the SGR mechanism to provide temporary relief from projected payment cuts and one that would repeal the SGR mechanism and instead increase payment rates each year by the Medicare economic index, or MEI.

• Alternative 1 would increase physician payment rates by 1 percent in 2008 but would not treat the update as a change in law or regulation. As a result, the SGR expenditure targets would remain the same, and the difference between cumulative actual spending and the cumulative expenditure targets would be larger than under current law. The increase in spending attributable to the higher payment rate under this alternative would eventually be recouped by the SGR, causing payment rates to be lower in the future than they would otherwise have been. This alternative would increase net federal outlays by $2.8 billion in 2008 and by $24.0 billion over the 2008-2012 period.
  
  
• Alternative 2 would increase payment rates by 1 percent in 2008 and would treat the update as a change in law or regulation. Thus, the SGR would be adjusted to account for the increased payment rate, and the difference between cumulative actual spending and the cumulative target would be largely unchanged from what it was under current law. This alternative would increase net federal outlays by $2.8 billion in 2008 and by $24.0 billion through 2012. (Because the Congressional Budget Office's baseline assumes that the maximum cuts are effective for at least the next several years, Alternatives 1 and 2 change outlays equally over the 2008-2012 period. However, Alternative 2 would cost more than Alternative 1 over the full 10-year period.)
  
  
• Alternative 3 is largely the same as Alternative 1, but it would include a provision specifying that Medicare's Part B premium would not be adjusted to reflect changes in spending resulting from changes in payment rates for physicians' services. Because that hold-harmless provision would uncouple premiums from program costs, this alternative would increase federal costs relative to their projected level under Alternative 1. In particular, this alternative would boost net federal outlays by $3.6 billion next year and by $30.7 billion over five years.
  
  
• Alternative 4 would fully replace the SGR targets with annual updates based on changes in the prices of inputs that are used to provide physicians' services, minus a productivity adjustment (as measured by the Medicare economic index). Instead of declining by about 10 percent in 2008 and around 5 percentannually for at least several years thereafter, payment rates would climb by about 2 percent annually. Those updates would not be subject to further adjustments, and excess spending, as defined by the SGR, would not be recouped. This alternative would increase net federal outlays by $3.2 billion in 2008 and by $65.0 billion from 2008 to 2012.

An advantage of all the alternatives is that replacing the 2008 projected negative update with a positive update (either a 1 percent update in Alternatives 1, 2, and 3 or an MEI-based update in Alternative 4) could forestall the possibility that beneficiaries might find it harder to locate a physician who accepts Medicare patients. (Several organizations have examined that issue, including the Government Accountability Office, the Medicare Payment Advisory Commission, and the Center for Studying Health System Change. Although they have not identified problems with Medicare beneficiaries' current access to care, it is uncertain at what point changes in physicians' participation because of lower fees would have a significant negative effect on Medicare patients' access to physicians' services.)

The first three alternatives, which would preserve the current SGR mechanism for updating payments for physicians' services, would temporarily lift the fee reductions scheduled under the SGR mechanism with the expectation that future payment rates would be reduced to recoup spending already incurred that exceeded the SGR targets. A rationale for preserving the SGR mechanism (as in Alternatives 1, 2, and 3) is that failing to recoup past excess spending under the SGR mechanism and increasing fees paid to physicians would add to the already substantial long-term costs of the Medicare program and to the broader budgetary pressures posed by the aging of the baby-boom generation.

Proponents of replacing the SGR mechanism (as in Alternative 4) argue that the current system is flawed because as a national target, it does not provide incentives for individual physicians to control the volume of services they provide. In addition, the SGR mechanism cannot differentiate between increases in the volume of physicians' services that are desirable (for example, for preventive care) and increases that are not.

An argument against modifying the SGR mechanism according to all of the alternatives considered here is that over the long term, higher spending by Medicare for physicians' services would boost federal spending, requiring cuts elsewhere in the budget, higher taxes, or larger deficits. In addition, because all of the alternatives to the current SGR formula result in higher Medicare spending, beneficiaries would face higher cost-sharing obligations and (except under Alternative 3) higher Medicare Part B premiums, which are set at 25 percent of the program's average costs.

The Medicare Modernization Act of 2003 (MMA) included a provision designed to limit Medicare's total outlays relative to its "dedicated financing sources," which include the Hospital Insurance payroll tax revenue and premiums paid by beneficiaries. That provision directs the Medicare trustees to calculate and report their projection of the general-revenue Medicare funding (GRMF) percentage, which equals total Medicare outlays minus dedicated Medicare financing as a percentage of total Medicare outlays. If the Medicare trustees report that the GRMF share will exceed 45 percent in one or more of the next seven years, then they are required to make a determination of "excess general revenue Medicare funding." If the trustees make such a determination in two consecutive years, the MMA requires the President to submit legislation designed to eliminate the excess general-revenue funding, and it establishes expedited procedures for considering legislation with that objective.

In their 2006 report, the Medicare trustees projected that the GRMF percentage would exceed 45 percent in 2012 and in all years thereafter. If the trustees make a second such determination in their 2007 report, the President would be required to propose legislation to address the excess. The GRMF percentage could be reduced in three ways: by reducing Medicare outlays, increasing Medicare's payroll tax revenues, or raising the premiums paid by beneficiaries for Parts B and D of Medicare.

This option would institute an automatic mechanism to reduce Medicare payments to fee-for-service providers across the board so that the GRMF percentage would not exceed 45 percent in any year from 2008 through 2017. It would apply to all payments that are made on the basis of a fee schedule, which includes payments to hospitals, physicians, and providers of post-acute care. The reductions in fee-for-service payment rates also would indirectly reduce Medicare's payments to private Medicare Advantage plans by reducing the so-called benchmark payment rates.

Because the Congressional Budget Office projects that the GRMF percentage will not exceed 45 percent until 2014, this option would have no effect on Medicare spending until that year. In 2014, Medicare spending would decline by $5.1 billion. Over 10 years, Medicare spending would fall by $91.0 billion.

An argument in favor of this option is that it would reduce the federal government's Medicare outlays compared with their level under current law. Reductions in Medicare's Part B expenditures would also reduce beneficiaries' premiums and out-of-pocket payments.

An argument against this option is that, once the payment reductions take effect, Medicare beneficiaries might face difficulties obtaining access to medical services if providers became less willing to serve them. Because providers would receive less revenue from Medicare, they would either have to reduce their operating costs or earn lower margins on Medicare patients. By reducing the revenue of providers, this option might also limit their ability to provide high-quality care. In addition, reductions in Medicare's payments to Medicare Advantage plans would result in beneficiaries' paying higher premiums for those plans or receiving a narrower benefit package, and some plans might withdraw from the Medicare program.

Medicare's Supplementary Medical Insurance (SMI) program, or Part B, allows beneficiaries to obtain coverage for physicians' and other outpatient services by paying a monthly premium ($93.50 in 2007). The premium was originally intended to finance 50 percent of the SMI program's costs, with the remainder funded from general revenues. Legislation enacted in 1972, however, limited growth in the SMI premium to the Social Security cost-of-living adjustment (COLA), and the premium's share subsequently fell below 25 percent of the program's costs because medical spending grew more quickly than inflation. After setting the premium's share at 25 percent during much of the 1980s, the Congress passed the Omnibus Budget Reconciliation Act of 1990, which specified dollar amounts for 1991 through 1995; but when per capita health care spending grew more slowly than anticipated, the premium's share of the program's costs rose to more than 31 percent. The Balanced Budget Act of 1997 permanently set the Part B premium at 25 percent of SMI spending. General revenues still fund the remainder. (Since January 2007, some higher-income enrollees have faced greater premiums for Part B, but the basic premium of 25 percent still applies to about 96 percent of enrollees; the "income-related" SMI premium is described in Option 570-16.)

This option would raise the basic SMI premium to 30 percent of the program's costs starting in 2008 while preserving the income-related-premium shares specified in current law, saving $6.8 billion in 2008 and $42.2 billion over five years. The estimate assumes a continuation of the hold-harmless provision, which protects SMI enrollees (other than those paying the income-related premium) from a drop in their monthly net Social Security benefits when the premium increase exceeds Social Security's COLA. The hold-harmless provision would apply to more enrollees in 2008 because of the initial increase in premiums from 25 percent to 30 percent under this option.

The main rationale for this option is that it would ease the budgetary pressures posed by rising SMI costs, which are expected to accelerate as the baby-boom generation ages. Even under this option, the public subsidy for most SMI enrollees would remain quite high, at 70 percent—a subsidy far greater than what was intended at the program's outset. Moreover, because Medicaid pays the SMI premiums for certain low-income Part B enrollees with limited assets, those people would be unaffected.

An argument against this option is that it would reduce disposable income for many SMI enrollees below what it would be otherwise. In addition, expenditures for states would rise if they paid higher premiums for people eligible for coverage through both Medicare and Medicaid.

Medicare's Supplementary Medical Insurance (SMI) program, or Part B, offers subsidized coverage for physicians' and other outpatient services. Before 2007, one SMI premium applied universally to all beneficiaries. Beginning in January 2007, however, the SMI premium is tied to enrollees' modified adjusted gross income (AGI). Although the premium for most beneficiaries will remain at 25 percent of SMI costs per aged enrollee, those in higher income categories will face progressively greater shares of 35 percent, 50 percent, 65 percent, and 80 percent, to be phased in over three years. For 2007, the income categories are defined using the following thresholds: $80,000, $100,000, $150,000, and $200,000. (For married couples, the corresponding income thresholds are twice the values shown.) Those thresholds are set to rise annually with changes in the consumer price index for urban consumers.

This option would apply the higher SMI premiums to more beneficiaries by one of two methods. Alternative 1 would eliminate the scheduled inflation adjustments to the income thresholds after 2008. Alternative 2 would reduce each 2008 income threshold by 20 percent while retaining the scheduled adjustments for inflation. As under current law, only those enrollees who pay the basic 25 percent premium would be covered by the hold-harmless provision, which ensures that the amount of an enrollee's Social Security check does not decline if the cost-of-living adjustment is insufficient to cover an increase in the SMI premium resulting from the annual update. Alternative 1 would not reduce outlays in 2008 but would decrease outlays by $3.3 billion from 2009 to 2012. Alternative 2 would reduce outlays by $600 million in 2008 and $8.2 billion over five years.

A rationale for this option is that it would provide savings amid the growing budgetary pressures posed by mandatory spending while leaving most SMI enrollees unaffected—particularly enrollees with lower income. Under Alternative 1, a large majority of beneficiaries, about 92 percent, would still pay no more than the basic SMI premium in 2012, while less than 1 percent would face the highest premium. More enrollees would be affected under Alternative 2, but about 90 percent would still see no increase in their premium in 2012. Under either alternative, the added cost of the higher premiums would be small compared with enrollees' income. SMI enrollees, including those paying an income-related premium, would still receive a substantial subsidy from taxpayers, including workers with modest earnings.

An argument against this option is that enrollees who faced the higher premiums would effectively see a reduction in their disposable income. Some of those enrollees might drop out of the SMI program as a result. (Few might be expected to do so, however, because enrollment rates for SMI have been historically quite high, even though enrollment for that program is voluntary.)

Currently, most enrollees in Part D of Medicare pay premiums that, on average, are intended to cover about 25 percent of the costs of providing that program's standard drug benefit. Enrollees with low income and few assets can have most or all of their premiums paid by Medicare, but middle-income and higher-income seniors pay the same drug premiums. Middle-income enrollees in Part B of Medicare also pay a premium that covers about 25 percent of that program's average costs, but some higher-income enrollees must pay a larger share starting in 2007. That increase applies to single individuals with an annual income of more than $80,000 and to married couples with a combined income of more than $160,000, and the increase rises in steps until it reaches a maximum amount for those with income above $200,000 (for singles) or $400,000 (for married couples). As modified by the Deficit Reduction Act of 2005, those changes will be phased in over three years and fully implemented in 2009—by which time the premiums paid by those enrollees will cover between 35 percent and 80 percent of average Part B costs.

Under this option, the higher-income enrollees who paya greater premium for Part B—that is, receive a reduced subsidy from Medicare—would see their subsidy decline to the same degree for Part D. If that option was implemented in 2008 and phased in over three years, federal savings would be $150 million in the first year and $2.9 billion over the 2008-2012 period. That estimate assumes that Medicare's subsidy payments to employers that provide qualified drug coverage would also be reduced for those higher-income retirees and that (as under Part B) the income thresholds would be indexed to general inflation. When fully implemented in 2010, this option would require that affected Part D enrollees pay an increment ranging from about $15 to about $80 per month on top of the average drug premium of about $35 per month for that year.

An argument for this option is that the additional payment represents only a small share of income for the enrollees who would be affected by it. The Congressional Budget Office estimates that less than 5 percent of enrollees would be subject to a higher Part D premium in any given year.

An argument against this option is that some higher-income enrollees would avoid the greater premium by opting out of the program—particularly those who have relatively low drug costs. (CBO assumed that about 1 percent of projected Part D enrollees would ultimately decline to enroll or delay their enrollment as a result of the higher premiums.) Critics maintain that such an outcome could eventually raise premiums for the remaining enrollees, who would have higher average costs. Concerns have also been raised about the administrative costs and burdens of tying enrollees' Medicare premiums to their income. But the incremental costs would be limited under this option because the steps needed to implement it—for example, income verification—are already being taken to carry out Part B provisions for income-related premiums.

Many people enrolled in Medicare Part B (Supplementary Medical Insurance, or SMI) have their premium payments automatically deducted from their Social Security benefit checks. The Medicare Part B premium is set to cover 25 percent of the program's costs. Under current law, the dollar amount of any increase in the Part B premium is limited to the dollar amount of the annual cost-of-living adjustment (COLA) for Social Security benefits. Under that hold-harmless provision, if the calculated increase in the SMI premium is greater than the dollar increase in the Social Security benefit, the premium is reduced by the amount needed to ensure that there is no reduction in the dollar amount of the net Social Security benefit.

This option would apply a similar hold-harmless provision to the increase in premiums for Medicare Part D (the new prescription drug benefit) beginning in 2008. (The option would not affect the initial reduction in the net Social Security benefit that occurs when enrollees first sign up for Part D.) Because Part D premiums will vary among beneficiaries (depending on the particular drug plan they choose), the hold-harmless calculations described here are based on the average premium for Part D plans. In other words, the net Social Security benefit of a beneficiary in an average plan could not fall from year to year. If beneficiaries enrolled in a plan whose increases in premiums were significantly higher than that of the average Part D plan, however, they could see reductions in their net Social Security benefit.

Expanding the current hold-harmless provision to include the Part D premium would increase Medicare spending by $250 million in 2008 and by $800 million over five years. The number of Medicare beneficiaries subject to both the current and proposed hold-harmless provisions would vary considerably over time, primarily because of significant year-to-year fluctuations in the rates of increase in the Part B and Part D premiums.

A rationale for this option is that it would limit the extent to which the rising cost of prescription drugs reduced the amount of income available to the elderly for spending on other goods and services. It would especially protect the net Social Security benefit of beneficiaries with relatively low lifetime wages (and thus low Social Security benefits) because the dollar amount of their COLAs would be relatively small.

An argument against this option is that, by insulating beneficiaries from the full impact of the cost of higher premiums for the drug benefit, the policy might reduce pressures to curb growth in Medicare's drug spending.

In Medicare's fee-for-service program—consisting of Part A (Hospital Insurance) and Part B (Supplementary Medical Insurance)—enrollees' cost sharing varies significantly depending on the type of service provided. For example, enrollees who are hospitalized in 2007 must pay a Part A deductible of $992 for each "spell" of illness they incur and are subject to daily copayments for extended hospital stays or skilled nursing care. Meanwhile, the deductible for outpatient services covered under Medicare Part B is $131. Beyond that deductible, enrollees generally pay 20 percent of allowable costs for most Part B services, but cost sharing can be significantly higher for outpatient hospital care. At the same time, certain Medicare services, such as home health visits and laboratory tests, require no cost sharing. As a result of those variations, enrollees are not given consistent incentives to weigh relative costs when choosing among treatment options. Moreover, if Medicare patients incur extremely high medical costs, they can face significant cost sharing, because the program does not cap those expenses.

This option would replace the current complicated mix of cost-sharing provisions with a single combined deductible covering all services in Parts A and B of Medicare, a uniform coinsurance rate of 20 percent for amounts above that deductible (including inpatient expenses), and an annual cap on each enrollee's total cost-sharing liabilities. Specifically, the combined deductible would be $500 in 2008, and the cap on total cost sharing would be $5,000; in later years, those amounts would grow at the same rate as per capita Medicare costs. If this option took effect on January 1, 2008, federal outlays would be reduced by $1.6 billion in that year and by $11.6 billion over five years.

One argument in favor of this option is that it would provide greater protection against catastrophic costs while reducing Medicare's coverage of more-predictable expenses. Capping enrollees' out-of-pocket expenses would especially help people who develop serious illnesses, require extended care, or undergo repeated hospitalizations but lack supplemental (medigap) coverage for their cost sharing. This option would also increase incentives for enrollees to use medical services prudently. Deductibles and coinsurance rates expose enrollees to some of the financial consequences of their health care treatments and are aimed at ensuring that services are used only when their benefits exceed those costs. Although this option's combined deductible would be lower than the Part A deductible, the vast majority of Medicare enrollees are not hospitalized in a given year; thus, most people without supplemental coverage would face the full cost for a larger share of the Part B services that they used. The uniform coinsurance rate across services would also encourage enrollees to compare the costs of different treatment options in a more consistent way. In addition, the resulting reductions in costs for Medicare's Part B program would translate into lower premiums for all enrollees.

An argument against this option is that it would increase cost-sharing liabilities for most Medicare enrollees. Specifically, those liabilities would rise modestly in 2008 for about three-fourths of enrollees (by about $500, on average) and would stay the same for another 15 percent. (For the remaining 10 percent of enrollees, cost-sharing liabilities would fall by an average of about $4,000.) Enrollees who are hospitalized only once in a year would generally face higher costs because of the coinsurance that would apply to that care; however, most Medicare enrollees would be insulated from those direct effects because they have supplemental coverage. (Some enrollees might see the effects in the form of higher premiums for their supplemental policies.) In addition, the option would make enrollees responsible for paying coinsurance for certain services—such as home health care—that are not currently subject to cost sharing. That requirement would increase administrative costs for some types of health care providers and could discourage enrollees from seeking treatment in some cases.

Cost-sharing requirements in Medicare's fee-for-service sector can be substantial, so most enrollees obtain supplemental coverage from some source (including the Medicaid program or their former employer). About 25 percent of fee-for-service enrollees buy individual insurance—or medigap—policies that are designed to cover most or all of the cost sharing that Medicare requires. Some studies have found that medigap policyholders use at least 25 percent more services than Medicare enrollees who have no supplemental coverage and at least 10 percent more services than enrollees who have supplemental coverage from a former employer (which tends to reduce, but not eliminate, their cost-sharing liabilities). Because enrollees are liable for only a portion of the costs of those additional services, it is taxpayers (through Medicare) and not medigap insurers or the policyholders themselves who bear most of the resulting costs. Federal costs for Medicare could be reduced if medigap plans were restructured so that policyholders faced some cost sharing for Medicare services but still had a limit on their out-of-pocket costs.

This option would bar medigap policies from paying any of the first $500 of an enrollee's cost-sharing liabilities for calendar year 2008 and would limit coverage to 50 percent of the next $4,500 in Medicare cost sharing. (All further cost sharing would be covered by the medigap policy, so enrollees could not pay more than $2,750 in cost sharing in that year.) If those dollar limits were indexed to growth in average Medicare costs for later years, savings would total $1.9 billion in 2008 and $14.4 billion over five years. Those estimates—which assume that all current and future medigap policies will be required to meet the new standards—reflect a reduction in Medicare costs of about 5 percent for the population of medigap policyholders that would be affected. (Two similar designs for medigap policies were authorized by the Medicare Modernization Act of 2003, but enrollment in them is optional.)

An argument in favor of this option is that most Medicare enrollees who have medigap policies would be better off financially as a result. Because insurers that offer medigap plans must compete for business, they would most likely reduce premiums to reflect the lower costs of providing the new policies. Indeed, most medigap policyholders would have smaller annual expenses under this option because their medigap premiums would decline to a greater extent than their cost-sharing liabilities would increase. (Part of the reason is that premiums for medigap policies are generally somewhat higher than the average cost-sharing liabilities that the policies cover, because of the administrative and other costs that medigap insurers incur. But the primary reason is that most of those liabilities are generated by a minority of policyholders.) Greater exposure to Medicare's cost sharing would also lead some medigap policyholders to forgo treatments that would yield them few or no net health benefits. Indirectly, the decline in Medicare's costs would also cause that program's monthly premiums (which cover 25 percent of costs for Medicare Part B) to fall, so other Medicare enrollees would also benefit.

An argument against this option is that medigap policyholders would face more uncertainty about their out-of-pocket costs. For that reason, some policyholders might object to being barred from purchasing coverage for all of their cost sharing, even if they would be better off financially in most years under this option. (Most medigap policyholders buy optional coverage for the Part B deductible; high-deductible medigap policies have attracted only limited enrollment despite their lower premiums.) Moreover, in any given year, about one-quarter of medigap policyholders would incur higher total costs under this option than they would under the current system, and those with costly chronic conditions might be worse off year after year. Finally, the decline in use of services by medigap policyholders (which would generate the federal savings under this option) might adversely affect their health in some cases.

The savings from modifying Medicare's cost-sharing requirements (see Option 570-19) could be increased by limiting medigap coverage at the same time (see Option 570-20). That is, the savings that would result from instituting both changes simultaneously would exceed the sum of the savings derived from implementing each option in isolation. That synergy arises because medigap policyholders would not be insulated from the changes in Medicare's cost-sharing requirements if their medigap plans were also restructured.

Under this option, medigap plans would be prohibited from covering any of the new $500 combined deductible that would be required by Medicare in 2008 (described in Option 570-19) and could cover only 50 percent of the program's remaining cost-sharing requirements. Such a medigap policy would correspond to the one described in Option 570-20, with coverage limited to 50 percent of the next $4,500 in Medicare cost sharing (thus capping enrollees' out-of-pocket expenses at $2,750 in 2008). Under this combined option, the point at which the medigap policy's cap on out-of-pocket costs was reached would also be the point at which the Medicare program's new cap was reached. Between the deductible and the catastrophic cap, policyholders would face a uniform coinsurance rate of 10 percent for all services. If those various dollar limits were indexed to growth in per capita costs for the Medicare program, this option would save $3.8 billion in 2008 and $27.9 billion over five years. Those estimates assume that participation in Medicare's new cost-sharing requirements will be mandatory and that all medigap policies will be required to follow the new standards.

An argument in favor of this option is that it would appreciably strengthen incentives for more prudent use of medical services—both by raising the initial threshold of health care costs that most Medicare beneficiaries faced and by ensuring that enrollees generally paid at least a portion of all subsequent costs (up to the out-of-pocket limit). As a result, the five-year savings from this option would be $1.9 billion more than the sum of savings achieved from Options 570-19 and 570-20.

An argument against this option is that even with the new catastrophic cap—which would protect Medicare enrollees against substantial out-of-pocket expenses—some enrollees would object to any policy that denied them access to full supplemental coverage for their cost sharing. Furthermore, in any given year, a significant number of enrollees would see their combined payments for premiums and cost sharing rise as Medicare's average subsidies were reduced and medigap plans were restructured.

Medicare's spending for home health care dropped during the late 1990s following enactment of the Balanced Budget Act of 1997, which introduced a prospective payment system for home health services. Since 2000, spending for home health care has been rising, however. And the use of home health services, and the resulting costs to the Medicare program, will grow rapidly over the next 10 years, the Congressional Budget Office projects. One reason for the projected rapid growth is that Medicare beneficiaries are not currently required to pay any of the costs of home health services covered by the program.

This option would charge beneficiaries a copayment amounting to 10 percent of the total cost of each home health episode—a 60-day period of services—covered by Medicare, starting on January 1, 2008. That change would yield net federal savings of $1.6 billion in 2008 and $12.9 billion over five years.

An argument in favor of this option is that it would directly offset a portion of Medicare's home health outlays and encourage beneficiaries to be cost-conscious in their use of those services. The use of services would also decrease, most likely among the approximately 10 percent of beneficiaries with fee-for-service Medicare only (in other words, beneficiaries who are not enrolled in Medicaid or a health maintenance organization and who do not have supplemental insurance, such as medigap or "wraparound" retiree coverage).

An argument against this option is that it would increase the risk of significant out-of-pocket costs for the 10 percent of Medicare enrollees with only fee-for-service coverage and thus could reduce the use of services among that population. Those enrollees tend to have lower income than do beneficiaries with private supplemental insurance. (Among the majority of enrollees who have supplemental insurance, little or no drop in use would be expected, because their supplemental policies would presumably be expanded to cover the home health copayment proposed in this option.) Also, the 25 percent of enrollees with individually purchased medigap policies would probably face higher premiums, and the costs of employer-sponsored medigap policies could also rise (again under the assumption that supplemental policies would cover the proposed home health copayment). Finally, this option would result in increased Medicaid outlays for home health care. (The federal share of higher Medicaid outlays is included in the estimated change in outlays.)

For enrollees who have been hospitalized and need continuing skilled nursing care or rehabilitative services on a daily basis, Medicare currently covers up to 100 days of care in a skilled nursing facility (SNF). The average SNF stay covered by Medicare lasts about 20 days, and a substantial share of Medicare's SNF payments are for the first 20 days of such a stay. The first 20 days of SNF care are free to the beneficiary, but the next 80 days require a copayment that is projected to be $124 per day in 2008. That copayment is set at one-eighth of Medicare's deductible for each hospital inpatient "spell," and thus the copayment grows over time along with increases in average daily hospital costs. Total payments to SNFs under Part A of Medicare are projected to average about $396 per day in 2008, so the $124 copayment corresponds to an average coinsurance rate of more than 30 percent. The Congressional Budget Office projects that total Medicare spending for SNF services provided under Part A will rise from $20.4 billion in 2008 to $34.5 billion in 2017.

This option would impose a copayment for each of the first 20 days of care in a skilled nursing facility equal to 5 percent of the inpatient deductible, which would be $51.60 per day in 2008. The maximum additional liability for a beneficiary would thus equal the inpatient deductible (projected by CBO to be $1,032 in 2008) and would rise at the same rate over time. Imposing that copayment would reduce federal outlays by $1.3 billion in 2008 and by $9.6 billion over five years.

The effect of this option on the use of SNF services and beneficiaries' out-of-pocket payments would depend on whether participants had supplemental coverage for their Medicare cost sharing. Most individual medigap policies include full coverage of current SNF copayments, so beneficiaries with such policies would be insulated from the direct impact of the higher copayments but could expect to see the additional costs reflected in their medigap premiums. This option would not affect Medicare beneficiaries who received full Medicaid benefits or those considered qualified Medicare beneficiaries, because their Medicare cost sharing would be paid by Medicaid. The savings shown in this option reflect the additional federal Medicaid spending that would occur as a result. (State Medicaid programs would also pay correspondingly more.)

Overall, 2 percent to 3 percent of all Medicare beneficiaries would incur higher out-of-pocket costs under this option in any given year, CBO estimates. For those beneficiaries, the absence of cost sharing for the first 20 days of SNF care under current law probably encourages additional use of those services. An advantage of imposing a copayment, therefore, would be that those beneficiaries would have to balance the costs and benefits of receiving care in a skilled nursing facility.

One argument against this option is that enrollees who use SNF care would already have been liable for the in-patient deductible as a result of their initial hospital admission. The added copayment could lead some beneficiaries to forgo services that would help avoid further complications from surgery or improve their health in other ways. Some beneficiaries might choose instead to receive similar services as a home health care benefit, which currently has no cost sharing. (The resulting added payments for home health services are reflected in the estimate of net program savings for this option.)

Medicare currently pays 100 percent of approved fees for 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 ($131 in 2007 and updated annually by the increase in the Part B premium) and a coinsurance rate of 20 percent.

This option would impose the SMI program's usual deductible and coinsurance requirements on laboratory services beginning January 1, 2008. The change would yield federal savings of $1.1 billion in 2008 and $8.3 billion over five years.

A rationale for this option is that, besides reducing costs to Medicare, such a change would make cost-sharing requirements under the SMI program more uniform and therefore easier for enrollees to understand. Moreover, although decisions about the appropriateness of tests are generally left to physicians, some enrollees might be less likely to request or undergo laboratory tests of little expected benefit if they had to pay part of the costs themselves.

An argument against this option is that only a small portion of the expected savings would stem from more prudent use of laboratory services; the rest would reflect the transfer to enrollees of costs now borne by Medicare. Moreover, the billing costs of some providers, such as independent laboratories, would be higher under this option because those providers 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.)

More than a million Medicare beneficiaries suffer from respiratory illnesses, such as chronic obstructive pulmonary disease, that require them to use supplemental oxygen. Until recently, Medicare made monthly payments to private suppliers for the rental of oxygen equipment indefinitely if it was deemed medically necessary. Beneficiaries were responsible for 20 percent of the costs of such rentals (including fees to rent the equipment and costs to service it).

The Deficit Reduction Act of 2005 replaced that system with a "rent-to-own" system, which became effective January 1, 2006. Under the new system, Medicare will pay for up to 36 months of continuous rental of oxygen equipment, after which the supplier must transfer ownership of the equipment to the beneficiary. Once the equipment has been transferred, the beneficiary no longer has to pay rental fees. After the beneficiary assumes ownership, Medicare will pay to maintain, service, and refill the equipment. As before, beneficiaries continue to be responsible for 20 percent of the cost of those services.

This option would shorten the continuous rental period for oxygen equipment from 36 months to 13 months. As under current law, Medicare would continue to pay 80 percent of the costs for oxygen contents, supplies, and accessories and for maintenance and servicing after ownership of the equipment is transferred from the supplier to the beneficiary. By shortening the rental period for the oxygen equipment, this option would reduce Medicare outlays by $3.5 billion over five years.

Proponents of this option maintain that it would more closely align Medicare's payments for oxygen therapy and the associated equipment with their costs. They argue that Medicare's rental payments under current law far exceed the purchase price of oxygen equipment, resulting in unnecessary Medicare expenditures. Moreover, they contend that limiting the rental period to 13 months would reduce beneficiaries' cost sharing for oxygen equipment.

Opponents of this option point out that oxygen suppliers might reduce the quality and continuity of care for beneficiaries in response to reduced payments from Medicare. In particular, they are concerned that payments after the rental period ends would be insufficient to provide for adequate clinical support, equipment replacement, and 24-hour emergency service. Moreover, after the rental period ends, beneficiaries might face barriers in obtaining new equipment if new technology became available that would improve their quality of life.

Since 1973, patients diagnosed with permanent kidney failure, or end-stage renal disease (ESRD), have been automatically eligible for Medicare regardless of their age. Those individuals typically must undergo regular dialysis or receive a kidney transplant. If they are covered by private group health insurance when they are diagnosed with ESRD, that private coverage remains the primary payer for ESRD-related costs for a 30-month period beginning with the date of Medicare eligibility (which is typically three months following the diagnosis of ESRD). Subsequently, Medicare becomes the primary payer unless normal kidney function returns—for example, following a successful transplant.

This option would extend the period during which Medicare is the secondary payer from 30 months to 60 months. That change would reduce federal spending by $40 million in 2008 and by about $1 billion over five years.

In 2004, approximately 336,000 patients in the United States underwent dialysis. Those patients had an average age of 61 and an average life expectancy of 5.6 years. Roughly 136,000 patients received kidney transplants that year; those patients had an average age of 49 years and average life expectancy of 15.7 years. Medicare was the primary payer for about 80 percent of those dialysis patients and 50 percent of those transplant patients; it was the secondary payer for about 9 percent of the dialysis patients and about 16 percent of the transplant patients. Overall, Medicare spending for ESRD, which today averages more than $50,000 per patient per year, grew from $12 billion in 1998 to over $20 billion in 2004, while spending by private health insurance for ESRD grew from $800 million to $2.3 billion. That growth was driven more by increases in the number of patients with ESRD than by increases in costs per ESRD patient per year.

An argument in favor of this option is that it would bring Medicare's ESRD coverage more closely into line with Medicare's coverage for other chronic diseases. According to that view, individuals who remain covered under their employer's group health plan while they or their family members undergo dialysis or a kidney transplant should continue with that coverage as the primary payer. This option would not affect the majority of ESRD patients, most of whom do not have private coverage when they first experience renal failure, are not eligible for private group coverage beyond the 30-month period (perhaps because they have stopped working), have had a successful transplant, or do not survive beyond the current 30-month period. For beneficiaries affected by the option, the impact would most likely be small because Medicare and private coverage for ESRD are typically similar.

An argument against this option is that it would increase pressure on private health insurers to either raise premiums or drop coverage for ESRD. Also, the option could increase financial difficulties for some individuals whose private group coverage was limited. For example, individuals who reached lifetime dollar caps between their 30th and 60th months following renal failure might have to pay out of pocket for ESRD treatment or seek Medicaid or other sources of coverage.