November 13, 2013

Mandatory SpendingOption 6

Function 570 - Medicare

Convert Medicare to a Premium Support System

(Billions of dollars) 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2014-2018 2014-2023
  Without a Grandfathering Provision
Change in Mandatory Outlays                        
  Second-lowest-bid alternative 0 0 0 0 -31 -43 -45 -48 -52 -56 -31 -275
  Average-bid alternative 0 0 0 0 2 -8 -14 -15 -16 -17 2 -69
 
  With a Grandfathering Provision
Change in Mandatory Outlays                        
  Second-lowest-bid alternative 0 0 0 0 -1 -4 -8 -12 -16 -20 -1 -61
  Average-bid alternative 0 0 0 0 * -1 -3 -4 -6 -7 * -22

Notes: This option would take effect in January 2018. It would not apply to dual-eligible beneficiaries (people who are jointly enrolled in Medicare and Medicaid).

* = between zero and $500 million.

Overview of the Issue

Nearly 30 percent of Medicare beneficiaries are enrolled in the Medicare Advantage program, or Part C, under which private health insurers assume the responsibility for, and the financial risk of, providing Medicare benefits. Almost all other Medicare beneficiaries receive care in the traditional fee-for-service (FFS) program, which pays providers a separate amount for each service or related set of services covered by Part A (Hospital Insurance) or Part B (Medical Insurance). Federal payments to Medicare Advantage plans depend in part on the bids that the plans submit (indicating the per capita payment they will accept for providing the benefits covered by Parts A and B) and in part on how those bids compare with predetermined “benchmarks.” Under a method that will be fully phased in by 2017, Medicare Advantage benchmarks depend on per capita spending in the FFS program at the county level. (Private insurers also participate in a separate bidding process that is used to determine payments under Part D, Medicare’s prescription drug benefit program.)

The current system ties federal payments for Medicare Advantage enrollees to spending in the FFS program, limits the degree of competition among plans, and does not require the FFS program and private insurers to compete on the same terms. Some policymakers and analysts have proposed replacing the current Medicare system with a premium support system, in which Medicare beneficiaries would buy insurance coverage from one of a number of competing plans—potentially including the FFS program—and the federal government would pay part of the cost of the coverage.

Key Design Choices That Would Affect Savings

The effects of a premium support system on federal spending and on beneficiaries’ total payments (premiums and out-of-pocket costs for medical care) would depend crucially on how the system was designed. Important choices include setting the formula for the federal contribution, determining whether the traditional FFS program would be included as a competing plan, setting eligibility rules for the premium support system, and designing the features of the system that would influence beneficiaries’ choices among plans.

This discussion assumes that a premium support system would retain certain features of the current Medicare program—namely, insurers could not refuse to enroll a Medicare beneficiary because of the person’s health, age, or other characteristics; federal payments to insurers would be adjusted to account for differences in enrollees’ health; and all enrollees in a given plan and geographic area would pay the same premium for the same coverage (except that, as under current law, higher-income beneficiaries would pay more to enroll in Part B). Changes to any of those features could have major consequences for federal spending and for beneficiaries’ total payments under a premium support system.

The Federal Contribution. Two general approaches are possible for determining how much of the cost of health insurance coverage the federal government would pay for under a premium support system: The amount could be derived either from the bids of participating health plans or through a mechanism designed to achieve a specified path for federal spending on Medicare. Either approach could be applied in many different ways. Some recent proposals would base the federal contribution on the second-lowest bid or on the average bid in a region, although many other possibilities exist. Setting the federal contribution to achieve a specific path for Medicare spending would require setting an initial amount per person and increasing it over time based on the growth of some particular economic or budgetary measure, such as per capita gross domestic product. In some cases, a hybrid of those two general approaches has been proposed: The federal contribution would be set on the basis of insurers’ bids but its growth would be capped on the basis of some broader economic measure.

If the federal contribution was based on insurers’ bids (but its growth was not capped), the contribution could be set to keep pace with insurers’ costs of providing the benefits covered by Medicare. The contribution would therefore be sufficient in future years for beneficiaries to buy coverage from at least one health plan in each region at a premium that represented the same percentage of the total cost of coverage that was chosen at the outset. Setting the federal contribution to achieve a specific path for federal Medicare spending would give the government greater control over its spending, but beneficiaries might face much higher premiums if insurers’ costs grew faster than the federal contribution did. The same issue could arise if the federal contribution was determined from insurers’ bids but its growth was capped.

The Fee-for-Service Program. A key choice in designing a premium support system is whether Medicare’s FFS program would be eliminated or retained as an option for beneficiaries, competing alongside private insurers. In the Congressional Budget Office’s assessment, eliminating the FFS program and the rates that it would pay health care providers under current law would cause the rates that private insurers paid providers for their premium support enrollees to be much higher—with a concomitant increase in the costs of providing Medicare coverage—than if a premium support system included the FFS program as a competing plan. That assessment is based on the observation that although Medicare Advantage plans generally pay providers about the same rates as Medicare’s FFS program, private insurers generally pay substantially higher rates for services provided to enrollees with private coverage. CBO expects that the presence of the FFS program as a competing plan would constrain the rates that private insurers paid for services provided to premium support enrollees, whereas eliminating the FFS program would cause those rates to rise toward the rates paid for enrollees with private coverage.

In a system in which the federal contribution was based on insurers’ bids, eliminating the FFS program would result in higher bids, which would reduce federal savings and could even cause federal spending to be higher under a premium support system than under current law. CBO also expects that in some regions, the FFS program’s bid would be among the lower bids, so getting rid of that program could directly reduce federal savings by raising the federal contribution in those regions. By contrast, in a premium support system in which the federal contribution was set to achieve a specific path for federal spending, eliminating the FFS program would not affect that spending, although the resulting increase in the cost of coverage for private plans would lead to higher premiums for beneficiaries.

Eligibility. Federal savings from a premium support system would depend partly on which beneficiaries were included in the new system. Some proposals include a “grandfathering” provision, under which all beneficiaries who became eligible for Medicare before the premium support system took effect would remain in the current-law Medicare program and only people who became eligible after that time would enroll in the new system. Although a grandfathering provision would keep current beneficiaries from having to adjust to a premium support system, it would reduce federal savings greatly, because only a small portion of the Medicare population would be covered by the new system initially, and that portion would increase only gradually over many years. Savings would be even more limited because average health care costs for newly eligible people entering the premium support system would be lower than the average for Medicare beneficiaries as a group (since those new entrants would be younger and, therefore, generally in better health).

Another key choice is whether and how dual-eligible beneficiaries—people who are jointly enrolled in Medicare and Medicaid—would be included in a premium support system. (CBO estimates that in 2009, those beneficiaries made up 19 percent of the Medicare population and accounted for 29 percent of total spending for Part A and Part B benefits.) Medicare covers some services for dual-eligible beneficiaries and Medicaid covers others, thus creating conflicting financial incentives for the federal and state governments (which jointly fund Medicaid) and for health care providers. Recent federal and state efforts have focused on integrating the two programs’ funding streams and coordinating the often-complex care that many dual-eligible beneficiaries receive. Including that group in a premium support system would pose substantial additional challenges. For instance, it would be difficult to give dual-eligible beneficiaries incentives to choose low-bidding plans in a premium support system while also minimizing their total payments for medical care. Nevertheless, excluding such beneficiaries would reduce the potential savings that a premium support system could achieve.

Features of the System That Could Affect Enrollment Choices. Many features of a premium support system would influence beneficiaries’ sensitivity to differences in plans’ premiums, thus affecting insurers’ incentives to reduce their bids. Two features of particular importance are how enrollees would initially select a plan and how much standardization would be required of the various plans.

One possible approach to structuring enrollment would be to have beneficiaries affirmatively choose a plan (possibly including the FFS program) when they entered the premium support system or else be assigned to a plan whose bid was at or below the benchmark. A second approach would be to allow beneficiaries who did not choose a plan when they entered the new system to remain in their current plan—or the FFS program, if that was their current source of coverage—or be assigned to a similar plan or to the FFS program if their current plan was unavailable. (An option for beneficiaries who were just entering Medicare and did not choose a plan would be to assign them to the FFS program.) The first approach would probably give insurers a greater incentive to lower their bids because they would anticipate that enrollments would rise more as a result. Under the second approach, beneficiaries would generally have less risk of being assigned to a plan that excluded their current providers from its network, but, depending on the region, some beneficiaries could unwittingly remain in plans that would require much higher premiums than they had paid before.

Another key question concerns the degree of standardization that would be required for benefit packages. Possible approaches include making all plans cover the same services and impose identical cost-sharing requirements; requiring all plans to cover the same services but allowing them to vary their cost-sharing requirements, as long as the benefit packages were actuarially equivalent (that is, each package covered the same percentage of total expenses for a given population); or letting plans vary both their covered services and cost-sharing requirements, as long as the benefits were actuarially equivalent. Federal costs under any of those approaches would depend crucially on whether the standard package had the same actuarial value as Medicare’s current benefits or some different value. In general, greater standardization of benefits would make it easier for people to compare plans on the basis of price, thus enhancing competition and lowering bids. However, standardization would prevent plans from offering benefit packages that some people might prefer to a standard package specified by the federal government. It could also limit the extent to which insurers developed innovative cost-sharing arrangements that might result in lower costs, higher-quality care, or both.

Specific Alternatives and Estimates

CBO examined four alternatives for converting Medicare to a premium support system. In all of the alternatives, the federal government’s contribution would be determined from insurers’ bids, and Medicare’s FFS program would be a competing plan. The nation would be divided into regions within which competing private insurers would submit bids indicating the amounts they would accept to provide Medicare benefits to a beneficiary of average health. The FFS program’s bid would be based on projected FFS spending in a given region for a beneficiary of average health. Insurers would bid on a benefit package that would cover the same services as Parts A and B of Medicare (with a few exceptions, as noted below) and that would have the same actuarial value as Parts A and B combined. (Medicare’s prescription drug benefit, which is delivered through a competitive system under Part D, would be administered separately.)

The four alternatives would differ by whether they included a grandfathering provision and by which of two approaches they used to determine the benchmarks for setting the federal contribution:

  • Under the second-lowest-bid approach, the benchmark in a region would be the lower of a pair of bids—the region’s second-lowest bid submitted by a private insurer and Medicare’s FFS bid.
  • Under the average-bid approach, the benchmark in a region would be the weighted average of all bids, including the FFS bid. Each bid would be weighted by the proportion of beneficiaries enrolled in that plan in the preceding year.

For each enrollee of average health, the federal government would pay insurers an amount equal to the benchmark for the region minus the standard premium paid by enrollees (explained below); insurers would receive larger or smaller government payments for beneficiaries whose health was worse or better than average. Neither the amount nor the growth rate of the federal payment would be capped.

Beneficiaries who enrolled in a plan with a bid that equaled the benchmark would pay a standard premium directly to the insurer; the standard premium would equal one-quarter of the estimated per capita cost of providing Part B benefits for all Medicare beneficiaries and would be the same across the nation (which corresponds to the formula used under current law for Part B premiums). Beneficiaries who chose a plan with a bid less than the benchmark would pay the insurer a premium that was lower by the full amount of the difference between the bid and the benchmark, and those who chose a plan with a bid greater than the benchmark would pay a premium that was correspondingly higher. The income-related Part B premiums specified in current law for higher-income beneficiaries would continue and would be withheld from Social Security benefits.

Beneficiaries would choose a plan during an annual enrollment period and would be required to remain in that plan for a year. They would automatically continue to be enrolled in the plan in subsequent years unless they chose a different one. (When the premium support system went into effect, however, Medicare beneficiaries would not remain in their previous plan automatically.) Beneficiaries who did not select a plan when they entered the premium support system would be assigned (with equal probability) to a limited number of plans that presented bids at or below the benchmark, including the FFS program if it met that criterion.

CBO assumed that the premium support system would not affect certain portions of federal spending for Medicare. For example, dual-eligible beneficiaries would be excluded from the system under these alternatives, and CBO assumed that Medicare’s spending for those beneficiaries would continue at the amounts projected under current law—as would spending for Part D (which would operate separately) and spending for certain items and services that are not covered by the bids that Medicare Advantage plans submit under current law. Those items and services include Medicare’s additional payments to hospitals whose share of low-income patients exceeds a specified threshold and spending for medical education, hospice benefits, and certain benefits for patients with end-stage renal disease. CBO excluded those categories of spending from the premium support system to simplify the analysis. In 2012, those excluded categories made up about 35 percent of net federal spending for Medicare (total Medicare spending, including spending on dual-eligible beneficiaries and on prescription drugs under Part D, minus beneficiaries’ premiums and other offsetting receipts).

For this option, CBO assumed that legislation establishing a premium support system would be enacted early in fiscal year 2014. To allow time for the federal government to develop the necessary administrative structures and for beneficiaries and insurers to learn about and prepare for the new system, CBO assumed that the system would be implemented in calendar year 2018. CBO also made many other detailed assumptions for these alternatives, which are described in Congressional Budget Office, A Premium Support System for Medicare: Analysis of Illustrative Options (September 2013). Some specifications were chosen to illustrate the potential for savings from a highly competitive system; others were chosen for feasibility of implementation or to simplify the analysis.

Unlike the other options in this report, whose budgetary effects are measured against CBO’s May 2013 current-law baseline projections, estimates of the effects of these alternatives over the next 10 years are based on analyses that were largely conducted using CBO’s March 2012 baseline projections of Medicare spending. Analysis of the longer-term effects of the alternatives is based on CBO’s June 2012 long-term projections of Medicare spending. (Those two sets of projections were the most recent ones available when much of the analysis was performed.) To estimate the budgetary effects of the alternatives over the next 10 years in dollar terms, CBO applied the estimated percentage changes in federal spending derived from the analyses based on its March 2012 baseline projections to its most recent projections of Medicare spending, which were released in May 2013.

Estimates of the budgetary impact of these alternatives over the next 10 years are highly uncertain, given the substantial changes to the Medicare program that a premium support system would entail, the government’s lack of experience with similar systems, the rapid evolution of health care and health insurance, and the significant changes occurring in the Medicare program under current law. Estimates are even more uncertain for the period after 2023.

Budgetary Effects Without a Grandfathering Provision. If the premium support system covered people who were already eligible for Medicare as well as future beneficiaries (but excluded dual-eligible beneficiaries), the second-lowest-bid alternative would reduce net federal spending for Medicare by $275 billion between 2018 and 2023, CBO estimates, and the average-bid alternative would reduce net federal spending over that period by $69 billion. By 2020 (an illustrative year shortly after the premium support system would be implemented), the second-lowest-bid alternative would reduce net federal spending for Medicare by 6 percent, compared with projected spending under current law, and the average-bid alternative would reduce that spending by 2 percent.

Another way to measure the effects of these alternatives is to examine their impact on the federal government’s net spending for affected beneficiaries—everyone, other than dual-eligible beneficiaries, who would have enrolled in Medicare under current law—for the benefits that would be included in the premium support system. (That measure consists of federal spending for affected beneficiaries—excluding spending for Part D benefits and the items and services noted above that are not covered by the bids of Medicare Advantage plans under current law—minus beneficiaries’ premiums and other offsetting receipts.) With no grandfathering provision, the second-lowest-bid alternative would reduce net spending for affected beneficiaries in 2020 by 11 percent, and the average-bid alternative would reduce such spending by 4 percent, CBO estimates. Those percentages are larger than are the percentage reductions in total Medicare spending because these savings are measured relative to the portion of Medicare spending that would be covered under the premium support system, rather than relative to total Medicare spending.

Under either alternative, the savings to Medicare between 2018 and 2023 would be similar in percentage terms to the savings estimated for 2020, with one main exception. Under the average-bid alternative, federal spending for 2018 would be higher than under current law, CBO estimates. The main reason for that difference is that the FFS program’s bid would receive a greater weight in constructing benchmarks in the first year of the new system than it would in later years (because CBO assumed that the weight would equal the proportion of beneficiaries enrolled in the FFS program under current law in 2017). Thus, under the average-bid option, most regions would have higher benchmarks in 2018 than they would later.

Looking beyond the next 10 years, CBO expects that, under either alternative, annual federal savings in percentage terms would remain roughly stable from 2023 through 2032, although the dollar amount of the savings would increase. Over the long term, the increase in price competition from the premium support system specified here would probably reduce the growth of Medicare spending by decreasing the demand for expensive new technologies and treatments and by increasing the demand for cost-reducing technologies. However, the potential for a premium support system to produce additional savings would be limited by provisions of current law that are designed to restrain the growth of Medicare spending. In particular, CBO anticipates, private insurers would not be able to hold down payments to health care providers to the extent required in the FFS program by the sustainable growth rate mechanism for physicians and by other current-law provisions that will limit payment increases for other providers.

Budgetary Effects With a Grandfathering Provision. Federal savings would be much smaller under a premium support system that excluded people already eligible for Medicare. CBO estimates that if the system applied only to people who turned 65 (or qualified for Medicare before age 65) in 2018 or later, and all other beneficiaries (including dual-eligible beneficiaries) remained in the current-law Medicare program, the system would cover only about 15 percent of the Medicare spending from 2018 through 2023 that it would cover if it did not have a grandfathering provision. With that system, the second-lowest-bid alternative would reduce net federal spending for Medicare by $61 billion through 2023, and the average-bid alternative would reduce such spending by $22 billion, CBO estimates.

Thus, modifying the second-lowest bid alternative to include a grandfathering provision would yield savings between 2018 and 2023 that are 22 percent of the savings that would be achieved without grandfathering. Under the average-bid alternative, the estimated savings over that period with a grandfathering provision are 32 percent of the savings that would be achieved without grandfathering. Those percentages are greater than the percentage of Medicare spending that would be covered by the premium support system because of a number of factors. Both with and without grandfathering, some factors would cause private insurers’ bids under a premium support system to be lower than their bids under the Medicare Advantage program, and other factors would cause those bids to be higher (see CBO’s September 2013 report for details). However, the factors that would cause bids to be higher would be relatively weaker with a grandfathering provision.

Grandfathering would also reduce, for an extended period, the incentives created by a premium support system to modify the development and adoption of new medical technologies. Thus, the restraints on the growth of Medicare spending that would probably occur under a premium support system would be substantially smaller for many years.

Other Considerations

The premium support alternatives would affect the premiums that Medicare beneficiaries paid for Part A and Part B benefits, their total payments for those benefits (premiums plus out-of-pocket spending), and the combined payments of the federal government and beneficiaries. CBO analyzed those effects in 2020, focusing on affected beneficiaries in the two alternatives without grandfathering—that is, on everyone enrolled in Medicare other than dual-eligible beneficiaries. (The agency has not yet completed such an analysis for the two alternatives with grandfathering.) The alternatives could also affect beneficiaries’ access to care and the quality of care they receive; CBO does not have the tools to study such effects, however, and does not anticipate having them in the near future.

Effects on Beneficiaries’ Premiums. CBO estimates that the premiums paid by affected beneficiaries for Medicare Part A and B benefits under the second-lowest-bid alternative in 2020 would be about 30 percent higher, on average, than the current-law Part B premium projected for that year. (Medicare beneficiaries generally do not pay premiums for Part A under current law.) In contrast, under the average-bid alternative, affected beneficiaries would pay premiums that were about 6 percent lower, on average, than the current-law Part B premium in 2020. The premiums paid by beneficiaries under each alternative would depend on the premiums charged by the available plans (which would vary by region) and on beneficiaries’ choices of plans.

Under either of the alternatives without grandfathering, beneficiaries in each region would be offered at least one plan with a premium at or below the standard premium (given the manner in which benchmarks would be calculated), and in most cases, at least one plan with a premium below the standard premium would be offered. CBO expects that, depending on how bidding regions were defined, there might be some regions in which no private insurers would participate in the premium support system. In those places, the FFS program would be the only plan available, and enrollees would pay the standard premium.

The standard premium under either of those alternatives would be lower than the current-law Part B premium, CBO estimates, because both alternatives would reduce total Medicare spending, and the standard premium would equal the same share of spending that the Part B premium equals under current law. That reduction in the standard premium is the main reason that the average premium paid by beneficiaries under the average-bid alternative would be lower than the projected current-law Part B premium; the additional premiums paid by beneficiaries who enrolled in plans with bids above the benchmark would roughly offset the premium reductions for beneficiaries who enrolled in plans with bids below the benchmark. Under the second-lowest-bid alternative, however, the regional benchmarks would generally be lower than they would be under the average-bid alternative, so CBO expects that many beneficiaries would enroll in plans with bids above the relevant benchmark, resulting in a much higher average premium than under current law.

Most beneficiaries who wished to remain in the FFS program would pay much higher premiums, on average, under either alternative than they would for Part B under current law. The difference would be greatest in regions where FFS spending per beneficiary was highest. Beneficiaries in regions where such spending was lowest would pay a premium for the FFS program that was, on average, close to the projected current-law Part B premium.

Effects on Beneficiaries’ Total Payments. CBO estimates that affected beneficiaries’ total payments for benefits from Parts A and B in 2020 would be about 11 percent higher, on average, under the second-lowest-bid alternative without grandfathering than under current law. In general, the premiums paid by beneficiaries would increase under that option, but out-of-pocket costs for medical care would decline (because more beneficiaries would enroll in lower-bidding private plans, which would tend to reduce the total costs of care while maintaining the required actuarial value). The reduction in out-of-pocket costs would offset part, though not all, of the increase in premiums.

Under the average-bid alternative without grandfathering, beneficiaries’ total payments for Part A and B benefits in 2020 would be about 6 percent lower, on average, than under current law. That reduction results from both lower average premiums and lower out-of-pocket costs for medical care. As in the previous alternative, the difference in out-of-pocket costs would be attributable primarily to increased enrollment in lower-bidding private plans.

The change in total payments for particular beneficiaries could differ markedly from the national average under either alternative. For example, people who chose to remain in the FFS program would generally face much higher premiums and would not see a reduction in their out-of-pocket costs.

Effects on Combined Spending by the Government and by Beneficiaries. The sum of net federal spending for Medicare and beneficiaries’ total payments would be about 5 percent lower in 2020 under the second-lowest-bid alternative than under current law, CBO estimates, and about 4 percent lower under the average-bid alternative than under current law. (Those effects are measured as a percentage of projected net federal spending and beneficiaries’ total payments, in each case focusing on affected beneficiaries and spending for benefits that would be covered by the premium support system.) The estimated reduction in total spending is slightly greater under the second-lowest-bid alternative because the federal contribution would be smaller under that alternative, which would increase competitive pressure, resulting in lower bids by private plans and causing a larger share of beneficiaries to enroll in low-bidding plans. The federal savings would be much larger under that alternative than under the average-bid alternative, but beneficiaries’ payments would be higher.