Skip to content
Publicly Available Published by De Gruyter November 17, 2015

The Transformation of Medicare, 2015 to 2030

Henry J. Aaron and Robert Reischauer


Medicare today is a better program on almost every dimension than it was just after July 30, 1965 when Lyndon Johnson signed public law 89–97. Nonetheless, short-comings, limitations, and inadequacies remain. What should be done to make Medicare a better program? What should Medicare look like in 2030? In this paper we try to answer these questions. Three perspectives are relevant: that of beneficiaries, current and future; that of policymakers and administrators, the program’s stewards; and that of society at large. We posit certain objectives and goals that we believe – and that we think a broad swath of Americans would agree – should be pursued to improve the Medicare program. Those goals include (a) affordability for Medicare beneficiaries, (b) affordability for the working population that is paying and should continue to pay for much of the current cost of the program, (c) reduction in what we regard as needless complexity, and (d) stability and continuity in several different senses. We restrict ourselves to changes that we judge to be affordable and feasible – politically, technically, and administratively – if not today, then over the next decade or two. We believe that changes in Medicare will remain incremental, as they have been for the last 50 years. We shall assume that the ACA takes root and that the exchanges, whether managed by states or by the federal government on behalf of the states, continue to operate. We shall assume that federal and state officials eventually surmount the administrative challenges they still confront. In particular, we assume that the exchanges come to serve a growing share of the American population and that they increasingly exercise the rather considerable regulatory powers over insurance offerings that the ACA grants to them. We divide Medicare reforms into four categories: payment reform, benefit reform, quality reform and management, and the role of private insurance plans (Medicare Advantage [MA]).

Keywords: Medicare

1 Introduction

In the 50 years since Medicare became law, the US health care system has changed markedly in almost every way. The power of health care to improve and extend people’s lives has vastly increased. The cost of medical care as a share of national income has tripled. The business organization of hospitals and other providers has become increasingly concentrated and centralized. Medical knowledge has grown enormously, causing progressively narrower specialization. Specialization has led, in turn, to recognition that coordination of care is essential to assure high quality treatment, especially of complex conditions.

Medicare has been shaped by all of these developments. But, to a degree, it has also been an independent force for change. For many years, Congress blocked Medicare from using its clout as the largest single payer for medical care to shape the delivery system. Even so, Medicare pioneered with prospective payment for medical care, first for in-patient hospital care and later for other services. Now, Medicare is initiating what may be a sea change in the way that providers are paid by spurring a shift of the risk-bearing function from insurers (or, in the case of Medicare, the public) to providers through accountable care organizations and bundled payments and by attempting to vary payments by performance and value.

Additional forces promise continued change. The Affordable Care Act is replete with pilots, experiments, and demonstrations to slow growth of spending, encourage coordination of care, and improve quality. Not all are likely to succeed, but neither are all likely to fail. The digitization of medical data promises improved ability to compare the value of different therapies. Genomics and nano-technology promise revolutionary advances in care. These forces and the current fluidity of the institutional structure suggest that the pace of change affecting health care delivery may well accelerate.

Medicare today is a better program on almost every dimension than it was in the first years after July 30, 1965 when Lyndon Johnson signed public law 89–97. Nonetheless, short-comings, limitations, and inadequacies remain. What should be done to make Medicare a better program? What should Medicare look like in 2030?

In this paper we try to answer these questions. Three perspectives are relevant: that of beneficiaries, current and future; that of policymakers and administrators, the program’s stewards; and that of society at large. We posit certain objectives and goals that we believe – and that we think a broad swath of Americans would agree – should be pursued to improve the Medicare program. Those goals include (a) affordability for Medicare beneficiaries, (b) affordability for the working population that is paying and should continue to pay for much of the current cost of the program, (c) reduction in what we regard as needless complexity, and (d) stability and continuity in several different senses. Clearly, these goals may be in conflict – most obviously, the goal of affordability for current beneficiaries may be at odds with affordability for workers and others whose taxes cover costs of care not paid by current beneficiaries.

We restrict ourselves to changes that we judge to be affordable and feasible – politically, technically, and administratively – if not today, then over the next decade or two. We believe that changes in Medicare will remain incremental, as they have been for the last 50 years. In support of our assumption that incrementalism will continue to characterize changes in health care policy, we invoke the ACA. Criticized by some as radical, the ACA is actually stunningly incremental. Most of the ACA’s expanded coverage comes through extension of an existing public program, Medicaid. The rest comes through purchase of private insurance in “exchanges,” markets regulated to promote competition among private venders, and through regulations that extend the ability of adult children to remain covered under parental plans. The ACA left largely untouched the tax breaks that support group insurance coverage for most working age Americans and their families. Private nonprofit and for-profit hospitals, other vendors, and privately employed professionals will continue to deliver most care. In that spirit, the incremental changes that we shall propose continue current or emerging trends.

We acknowledge that past litigation and political disputes have cast great uncertainty over the Affordable Care Act. The Supreme Court’s June 2015 decision in the King v. Burwell case eliminated uncertainty over whether tax credits and cost-sharing subsidies may be paid on behalf of people in states that chose to allow the federal government to administer health care exchanges. Nonetheless, considerable uncertainty remains. Other legal challenges are working their way through the courts. The next president may seek changes in the ACA, broad or narrow depending on the outcome of the 2016 elections. Nonetheless, we shall assume that the ACA takes root and that the exchanges, whether managed by states or by the federal government on behalf of the states, continue to operate. We shall assume that federal and state officials eventually surmount the administrative challenges they still confront. In particular, we assume that the exchanges come to serve a growing share of the American population and that they increasingly exercise the broad regulatory powers over insurance offerings that the ACA grants to them.

We divide Medicare reforms into four categories: payment reform, benefit reform, quality reform and management, and the role of private insurance plans (Medicare Advantage [MA]).

Medicare is more parsimonious than the private insurance coverage typically offered by large employers. Medicare covers approximately 80 percent of the actuarial value of covered services, the standard for “gold” plans under the ACA.[1] Enrollees in traditional Medicare face daunting complexity, particularly if one takes account of the supplemental coverage that most Medicare enrollees have in order to reduce the risk of large out-of-pocket expenses. Supplemental coverage can come through Medicaid, employer-financed private retiree health coverage, or an individually-purchased Medigap plan. Those electing MA normally can choose from a number of plans, many of which offer supplemental cost sharing protection. The degree of choice they have dwarfs that available to most working Americans, who are normally restricted by their employers to one or a few options.

1.1 Complexity

Everyone age 65 or older who has paid payroll taxes for the requisite period now qualifies for Medicare Part A (Hospital Insurance or HI).[2] So too do those who are both age-eligible and have been or are married to such a person and those who have received a Social Security Disability Insurance pension and have received those benefits for at least two years. For two reasons, nearly all – 92 percent – of HI participants in 2014 enrolled voluntarily in Part B, which covers physician and other services.[3] First, premiums are highly subsidized. Second, most supplemental retiree policies (other than retiree benefits under the Federal Employees Health Benefit Program, TriCare and the VA) require Part B enrollment.[4] Medicare Part D offers prescription drug coverage through a bewildering variety of competing private sector drug plans to those in traditional Medicare. Each drug plan has its own formulary. Drug tiers differ. So do deductibles and cost sharing as long as their actuarial value at least equals the Part D standard. Many MA plans include drug coverage subsidized through Part D as do some private plans offered by employers to their retired workers. Like Part B, Part D coverage is highly subsidized from general revenues. Monthly premiums vary widely depending on each plan’s features.

Beneficiaries may enroll in a MA plan instead of receiving coverage under Part A, B and D, traditional Medicare, in which providers are usually paid on a fee-for service basis. MA plans come in several types – health maintenance organizations (HMOs), point of service plans that combine the features of HMOs with the option of choosing independent physicians or other service providers, preferred provider organizations, private fee-for-service plans, special needs plans, and medical savings account arrangements. MA plans receive a flat monthly payment per person adjusted for the risk characteristics of enrollees. In return MA plans are obliged to provide benefits similar to those available in traditional Medicare. By managing care, limiting networks of providers, redesigning cost sharing, and implementing other cost saving measures, MA plans may lower cost sharing or premiums or expand benefits.

1.2 Financial Risk

Enrollees in traditional Medicare may face large out-of-pocket expenses:

  • Monthly premiums for part B ($104.90 per month in 2015 for most enrollees, and from 40 percent to 220 percent higher for beneficiaries with incomes above annually adjusted thresholds) and part D (an average of $32 per month in 2015, with income-related surcharges similar to those for Part B);

  • Deductibles – $1260 for each hospital stay of 60 days or less under part A, and an annual deductible of $147 in 2015 under part B;

  • Diverse co-payments and coinsurance – a daily co-payment under part A ($315 in 2015 for hospital stays of 60 to 90 days); a fee for stays in Skilled Nursing Facilities of 20 to 100 days ($157.50 a day in 2015); up to 20 percent of most part B allowed charges (laboratory, home health and preventive services have no cost sharing).

Coverage is capped for some services, including hospital and skilled nursing facility stays beyond certain limits; and there is the infamous donut-hole (coverage gap) under some part D plans, now being phased out by the ACA. Part D, covering pharmaceutical products, is the only element of traditional Medicare that caps out-of-pocket expenses.

Because Medicare lacks a cap on health expenditures, other than in Part D, and because cost sharing for some services is steep, many Medicare beneficiaries risk incurring unaffordable medical costs if they rely solely on traditional Medicare for their health insurance. Those with very low incomes and few assets may qualify for Medicaid which covers premiums, deductibles, and other cost-sharing for dually-eligible enrollees and pays for some additional services not available in Medicare. Those with resources just above their state’s Medicaid eligibility standards but below the poverty threshold can have their premiums and cost-sharing paid under the Qualified Medicare Beneficiary (QMB) program. Those with incomes between the poverty level and 135 percent of poverty have their premiums paid by the Specified Low-Income Medicare Beneficiary (SLMB), and qualified individual (QI) programs. These three Medicare Savings Programs (MSPs), which are run through state Medicaid offices, have low asset as well as income tests. Although participation has increased significantly in recent years, many who are eligible have not availed themselves of this protection.

Because enrollees in traditional Medicare may incur large out-of-pocket expenses, over 90 percent of them seek additional protection – through Medicaid, employer sponsored retiree policies, MA plans or individually-purchased Medigap policies. This dual coverage has created a confusing jumble of payment systems for the elderly and disabled. It is unsurprising that many end up with coverage that is less than optimal given their circumstances. Dual coverage complicates billing by providers and therefore raises administrative costs.

2 Benefit Reform

These problems can be addressed in different ways; by cleaning up the separate components of traditional Medicare and supplemental coverage; by combining those parts; or by replacing traditional Medicare with something else modeled on the current MA program.

2.1 Fixing the Separate Components

The simplest, but least satisfactory, reform would keep the current structure, but cap out-of-pocket spending under all supplemental coverage – Medigap plans, employer sponsored retiree policies, and MA alike. Two of the 10 standardized Medigap plans that are available to new enrollees now cap out-of- pocket spending. But fewer than 2 percent of Medigap buyers chose such plans in 2012.

Rather than indirectly limiting exposure to financial risk through supplemental coverage, legislators could limit out-of-pocket spending directly in Medicare. To offset the resulting increase in program costs, cost sharing could be imposed for some Part B services that currently do not require it – home health and lab services but not preventive services, for example. In addition or alternatively, Part B premiums or coinsurance could be raised. Partly offsetting the added costs, premiums for Medigap or retiree policies would fall as some costs are shifted on to Medicare. Retaining the current framework has the virtue of simplicity, but would do nothing for beneficiaries who lack supplementary protection. Worse, it might not even work. Requiring Medigap plans to include catastrophic protection would raise their prices. Higher prices could lower demand for Medigap coverage. On net, exposure to risk could increase.

That most supplemental coverage plans lower a wide range of expenses and few enrollees elect plans offering stop-loss protection suggests that most beneficiaries, particularly those with modest incomes, want more and different protection from Medicare-related costs than a catastrophic spending cap would provide. To meet this demand, we suggest a third, and we believe, fairer approach: capping out-of-pocket Medicare spending as a fraction of family income. The cap could be a uniform fraction or one that is lower for people with low incomes. Until recently, the cost of administering such a cap would have been prohibitive.

Three recent developments have changed that situation. First, the IRS has acquired experience with income related Part B and D premiums. Second, CMS has garnered relevant experience under the Low-Income Subsidy program in Medicare Part D and the Medicare Savings Programs (QMB, SLMB, and QI). Third, experience under the ACA has taught the IRS how to administer income-related tax credits and cost-sharing subsidies and how to determine whether people are enrolled in Medicaid or an exchange plan. The ACA has birthed an administrative infrastructure able to handle such calculations.

These three developments mean that the introduction of income-related caps on out-of-pocket payments under Medicare should now be manageable. Those selling Medigap and retiree policies would have to let CMS know how much of Medicare’s cost sharing they had covered. Providers would have to report their beneficiary receipts. And the IRS would provide beneficiary income information as it does now for income-related Part B and D premiums. Medicare’s costs would rise but the need for and cost of supplemental protection would fall.

2.2 Consolidation

We believe that the desire of beneficiaries for protection from financial risk can be satisfied more simply. Most Americans, other than those enrolled in Medicare, can and do receive comprehensive coverage from a single health insurance plan purchased from a single insurer. Medicare enrollees should have the same opportunity. Current and looming fiscal challenges mean that such a plan should be financed by premiums and other charges so that it does not increase the net burden on public budgets.

The long-run goal should be to provide adequate coverage through a single insurance plan that encourages cooperation among the various health care providers with whom patients have contact. Such collaboration is particularly important for patients with numerous health conditions, especially if they are chronic. Such health problems are most prevalent in the elderly and in people with disabilities, the very groups that Medicare serves.

At a minimum, Parts A and B should be consolidated, with a single deductible and with coinsurance, co-payments, or cost sharing keyed to the value of the intervention. Any such consolidation that leaves overall financial exposure of enrollees unchanged will produce some financial gains and losses among those who are currently receiving benefits. If resistance to such cost shifts is strong, unification could be mandated only for new enrollees and made optional for those now on the rolls. Unifying only parts A and B is not really adequate, however, as it will do nothing to reduce the needless cost, complexity, and confusion generated by the double coverage through Medicare and supplemental coverage.

3 Program Structure

Since enactment of Medicare 50 years ago, the designs of Medicare and of private insurance have diverged. Exemplified by the Blue Cross/Blue Shield plans prevalent in the 1960s, Medicare was fashioned with two parts. Part A, Hospital Insurance, focused on hospitalization; Part B, Supplemental Medical Insurance, on physician services. Each part also covered additional services. In the ensuing decades, the two-part segmentation of private insurance has largely vanished. Whether covered by employer-sponsored insurance or non-group plans, one insurance plan typically covers medical and surgical services provided in hospitals, other institutions, or physicians’ offices.[5]

Traditional Medicare, in contrast, has gone from two parts to three – Congress in 2003 added Part D covering drugs. Since the 1970s Medicare beneficiaries have been able to opt out of the A/B structure and, after 2006, the A/B/D structure and buy coverage from private insurers which are paid a risk adjusted monthly fee to supply all of the services except hospice care offered through traditional Medicare. In 2003, this option was christened Part C, and plans paid on a capitated basis became known as MA plans.[6]

Nineteen years ago, we suggested that traditional Medicare should transition into a Part-C-like plan under which Medicare beneficiaries would receive a voucher (cash payment), indexed to a measure of health care costs, which could be used to buy health insurance in a well-regulated competitive health care market place. [Insert reference: Henry J. Aaron and Rober D. Reischauer, “The Medicare Reform Debate: What is the Next Step?” Health Affairs, 14, no. 4, (1995): 8–30] This option, which we christened premium support, has become the subject of considerable debate. The Affordable Care Act created such regulated markets (health exchanges) open to people buying individual insurance coverage and to employees in companies with up to 50 employees (increasing to 100 or fewer employees in 2016 and, at the state’s discretion, over 100 in 2017). The ACA’s income-linked premium assistance and cost-sharing subsidies are available only to people who buy coverage through these exchanges.

When we made our premium support proposal, Medicare private plans were not available everywhere. They enrolled only 9.2 percent of Medicare participants. Most were HMOs that managed care. Some were on a shaky financial footing. Today Part C of Medicare – MA – is operating fairly well. It serves a growing share of Medicare beneficiaries – 30.2 percent or 16.2 million as of 2014. In some places, over half of Medicare participants are enrolled in a MA plan. In the average county, beneficiaries can choose among 10 available plans offering several different organizational types (HMO, POS, PPO, etc.), as well as traditional fee for service Medicare. While the performance of some types of plans is spotty, many MA plans have an established record of delivering high quality care at costs no greater than, and often below, those of traditional Medicare. Most provide reduced cost sharing, additional services, or premium rebates. Furthermore, the organization of some MA plans facilitates the types of care coordination that provide the best settings for high quality health care.

Notwithstanding these positive developments, we no longer believe that Medicare should eventually transition into an all-premium support system. Choice of a health insurance plan is complex and important. When plans may freely exit a market, change their premiums and cost-sharing annually, and modify their provider networks at will, beneficiaries face great uncertainty from year-to-year. Even when information about health plans is presented as clearly as possible, well-educated and mentally alert consumers find it difficult to make informed choices among the many private plans on offer.[7]

To be sure, people buying insurance through the ACA exchanges face similar challenges. But, on average health care is not so important for them as it is for the elderly and disabled. Furthermore, cognitive impairments that that make informed decision making difficult or impossible is far more prevalent among Medicare beneficiaries than among current exchange enrollees. These problems would greatly expand the need for assistants similar to “representative payees,” who in 2013 served 9 million recipients of Supplemental Security Income and Social Security benefits who were found to need help managing their money. The even more challenging task of selecting among health insurance plans and paying health insurance bills would increase the need for similar assistance. To be sure, people in traditional Medicare who must select primary care doctors and specialists and choose among competing Part D plans face similar challenges. But except for choosing a part D drug plan, these decisions are generally simpler and more easily changed than is the decision to select a health plan for a full year when health care needs are likely but uncertain.

Those decisions are challenging for anyone. But, the stakes are far higher for the elderly and people with disabilities than for the young, with their generally lower needs for health care. And the challenge is greater still for those with cognitive impairments, the prevalence of which is higher among Medicare beneficiaries than among the general population.

For these reasons, we think it would be unwise at any time in the foreseeable future to require all Medicare enrollees to deal with ACA-like exchanges. It would certainly be premature to do so now before the ACA exchanges have met the implementation challenges they all face for their younger clients. And it would be rash in the extreme ever simply to turn all Medicare beneficiaries loose in an unregulated insurance market. We also think it vital to continue to offer the option of traditional Medicare and to do so in a way that makes it affordable both to enrollees and to the public.

The objective of traditional Medicare should be to offer adequate coverage at an affordable price through a single, simplified insurance structure to those who do not want to participate in MA. Having multiple insurers responsible for the same claim is expensive to administer and confusing for the patient. Providers or patients must first send claims to Medicare and, after Medicare has paid, to the beneficiary’s secondary insurer of which there are many each with its own procedures and payment cycles. Only when both claims have been paid can they send a correct bill to the patient. Sometimes after receiving the provider’s bill indicating Medicare’s payment, patients prematurely pay the balance only to discover later that their liability was less by the amount paid by their secondary insurance. Then they must wrangle with the provider to get a refund of the secondary insurer’s payment. Some choose to wait until all the payment dust has settled and they have received several repeat bills from the provider. This drives up the provider’s administrative and financial costs. For those who do not have supplemental insurance with first dollar coverage there is the confusion of meeting two different deductibles.

Furthermore, designing measures to improve health or create rational incentive structures for providers and beneficiaries is virtually impossible when the costs of a service are incurred by one insurance plan and the benefits accrue to another. For example, in some cases more expensive drug therapy might lower overall costs and provide better results, saving money for Part A or Part B but increasing costs for Part D. Under the current fragmented structure, the Part D plan has no incentive to use the more expensive drug.

We propose offering traditional Medicare indefinitely but in two variants. Unified Traditional Medicare – UTM – would bring together into a single unified insurance plan Parts A, B and D and some form of catastrophic protection. Unified Traditional Medicare Plus – UTM+ – would add enhanced cost sharing protection like that offered through Medigap and employer/union sponsored retiree policies. Each option would have deductible and rationalized coinsurance and copayments.[8]

A consolidation of the sort suggested here would cause dislocations for some beneficiaries and insurers. The new limits on out-of-pocket payments in UTM+ would raise Medicare costs; but administrative efficiencies would offset them, at least in part. Overall, the cost of UTM+ would be lower than the combined cost of traditional Medicare and supplemental coverage.

Medigap policies and employer retiree policies for those 65 and older could still be offered. They might be of interest to those who, for one reason or another opt to enroll in UTM. But a fee, equivalent to the added cost of the induced demand for Medicare services that such supplemental coverage generates should be added to the cost of supplemental insurance.[9]

Because of that fee and the greater complexity of independent insurance through UTM and a separate supplemental plan we believe that the market would shrink to the risk-averse segment of those enrolled in Unified Traditional Medicare but who, for one reason or another, did not choose UTM+.[10]

In addition to affecting Medigap plans and employer sponsored retiree policies, our proposal would have a significant impact on Part D plans serving those with traditional Medicare. Most insured Americans have drug coverage that is coordinated with and integrated into their broader health insurance plans. The insurer may contract with a pharmacy benefit manager (PBM) or it may provide these services itself. The insurer has an incentive to design the formulary, cost sharing, and other plan parameters to moderate overall health costs, not just drug costs, while maximizing health outcomes. This is how employer/union-sponsored insurance for workers and their families, plans offered on the exchanges, and MA plans operate. In contrast, the Part D plans available to those in traditional Medicare have little incentive to coordinate the design of their plans with Parts A and B of Medicare. Their incentive is to appeal directly to participants, which usually means emphasizing low premiums. To be sure, Part D has safeguards intended to ensure that all plans cover drugs in all therapeutic classes. But Part D plans have an incentive to discourage the use of costly drugs even when their use could lower the costs for Part A and B services.

Under our proposal to consolidate Parts A, B and D of traditional Medicare, existing Part D plans would transition into the pharmacy benefit manager (PBM) role that they play in the rest of the health care market. CMS would be responsible for constructing formularies and designing the cost sharing structure of which there could be several options. If there was a need to protect drug coverage decisions from political influence, these decisions could be vested in a commission of experts appointed for fixed terms and removable only for cause.

Competing private companies would submit bids to CMS on the standardized product or products for the UTM and UTM+ for each of the 34 PDP regions. CMS could choose more than one bidder in each region. Nevertheless the number of companies in this space would probably continue to fall, as it has for years. Currently, Part D is a highly concentrated market. The top three companies account for half of total enrollment, the top 10 account for 79 percent.[11] Companies can and do offer multiple plans. Traditional Medicare enrollees had at least 28 to choose among in 2014 (plus MA-PD plans). Under the new structure, the current confusing amount of choice would be greatly reduced and the 10 percent of Medicare beneficiaries without credible drug coverage would gain protection.[12]

4 Quality Reform and Management

Organized medical groups have long feared that program administrators would use their leverage as major payers to interfere in the practice of medicine. To win at least tacit acceptance of Medicare, the law creating the program expressly barred Medicare from trying to influence the practice of medicine. In recent years, analysts and, to an increasing degree, members of Congress, have recognized that Medicare can, and should, be used to improve the quality of care and to encourage the efficient delivery of care. CMS introduced payment incentives under HI to reward good care and penalize bad care. MA plans receive bonus payments based on the quality of care they provide. The ACA conditioned sharing savings generated by accountable care organizations on measures of health care quality. The recently enacted legislation ending the flawed Sustainable Growth Rate formula for physician fees replaced that formula with large payment variations based on measures of the quality of physicians’ performance. The pay-for-performance movement was based on hopes, perhaps overblown, that incentives could be fine-tuned to promote improved inpatient and outpatient care. The conversion of paper to electronic health records enhances the opportunities to marshal data in support of effective procedures.

Medicare also plays an important part in defining acceptable standards of care by approving or, more rarely, disapproving new medical procedures. Medicare’s decisions on whether to pay for new drugs, devices, and procedures influence decisions by private insurers. In general, one might think that coverage decisions should be based on scientific evidence. To a large extent they are. But there have been some deplorable exceptions, when a Congress responded to pressure from industry and professions and intervened to support coverage of ineffective services.[13] A complete divorce between coverage decisions and democratic politics is impossible. It is, in fact, undesirable. When budgets are limited and some drug, device, or procedure is medically superior to, but more costly than, another, someone must decide whether the added benefit is worth the added cost. Science should inform such decisions, but the decision also involves values, the political expression of which is a responsibility of democratically elected officials. Elected officials are loath to surrender these powers.[14] Nonetheless, it would be desirable if Congress vested decisions about Medicare coverage in an independent board consisting of members appointed for their scientific expertise and supported by sufficient staff to do authoritative studies of the value and cost of medical innovations.

The United States Preventive Services Task Force (USPSTF) offers a model for such a body. This independent, volunteer group consists of 16 nationally recognized experts in prevention and evidence-based medicine. It is appointed by the Agency for Healthcare Research and Quality (AHRQ) which also provides funding and staff support. The Task Force makes recommendations as to whether and for whom particular preventive services are efficacious. These recommendations, which are transmitted to Congress in an annual report, are based on the Task Force’s assessment of the peer-reviewed scientific evidence of the effectiveness (but not the cost) of various clinical preventive services such as screenings, counseling and preventive medications. Congress, some private insurers, and some physicians/providers have not always followed the Task Force’s recommendations. Such was the case with breast and prostate cancer screening. If such a board (or, preferably, separate boards each focusing on a class of disease) was created for Medicare, the decisions would take effect unless Congress voted to override them. Such an arrangement would affirm Congress’s power to express the values of its constituents, but set a presumption that scientific judgments should normally prevail in setting Medicare’s coverage policy.

5 Improving Medicare Advantage

MA, which has evolved significantly over the last four decades, has clearly been a success. It has provided Medicare beneficiaries with several alternative delivery systems as alternatives to the unmanaged, fee-for-service delivery system available in traditional Medicare. Three out of ten Medicare participants are now enrolled in a Part C plan. Participation is likely to grow as younger generations, who have experience with provider networks and managed care from their years with employer sponsored insurance, reach Medicare’s age of eligibility and as growing numbers of existing FFS participants join a MA plan after experiencing traditional Medicare’s high cost sharing requirements.

In the past, flawed methods of paying private plans have stimulated MA enrollment. In general, plans received capitated payments larger than the costs CMS would have incurred providing Parts A and B services to the plan’s participants under traditional Medicare. In some instances these excess payments were truly excessive.[15] In recent years, plans have been required to devote some or all of the excess payments to additional services (such as vision or dental care), reduced cost sharing or rebates on Part B and D premiums. These added benefits made the plans even more attractive and created pressure to ensure that plans were available everywhere, even in rural areas where population density and provider availability made coordinated care delivery systems and provider networks impractical. In response, new types of MA plans – private fee-for-service and medical savings account plans – were added to the Part C menu.

Over the years Congress has modified the payment mechanism several times in an effort to reduce excess payments while ensuring wide availability of the MA option. Under the current system, which will be fully phased in by 2017, plans submit bids for providing Part A and B services (plus administrative costs and profits) to a standardized Medicare participant in each county they want to serve. This bid is compared to a benchmark that is related to traditional Medicare costs for a standardized participant in the county compared to the national average. For counties where traditional Medicare costs are high relative to the national average the benchmark will be 95 percent of the average cost of traditional Medicare in that county; where the cost of traditional Medicare costs is well below the national average the benchmark can be as high as 115 percent of the local average. Plans with a quality measure score above 4.0 (37 percent of plans in 2014[16]) are given bonuses that can raise their benchmark. The difference between the adjusted benchmark and a plan’s bid, if positive, is split between CMS and the plan. High quality plans receive a larger fraction of difference, which must be used to provide additional benefits or to reduce participants’ costs and premiums. Plans bidding above the benchmark are paid the benchmark and must charge their participants additional premiums equal to the difference between their bid and the benchmark. All payments are risk-adjusted to reflect the health status of each plan’s enrollees. MedPAC has estimated that, on an enrollment weighted basis, plans’ bids were 98 percent of traditional Medicare costs, while they received payments equal to 112 percent of those cost. Analysts believe that this gap will narrow significantly when the new mechanism is fully phased in.

This payment mechanism is an improvement over previous payment methods. Nonetheless, it is likely to be too generous to plans offering coverage in areas where traditional Medicare costs are very low, and it may not provide Medicare with a fair share of the savings that could be realized from an efficient Part C structure. Setting benchmarks at the local average cost of traditional Medicare in areas where the cost of traditional Medicare is below the national average would reduce this problem. The rationale for this approach is that MA plans should be at least as efficient as traditional Medicare.[17] By 2030, however, it should be possible to encourage even stronger competitive forces by setting benchmarks so that they are linked to the bids made for the local area. Those benchmarks could be set at the enrollment-weighted average bid or at some other level, as is done in the ACA. The benchmark could include the average costs of Unified Traditional Medicare (UTM) where those costs were below the national average. To ensure that bidding reflect competitive forces, it is probably necessary to include safeguards such as a requirement that the benchmark is no higher than that calculated under the previous payment mechanism.

To facilitate comparisons among plans and between the private plan options and UTM and UTM+, all MA plans should be required to include Part D protection and cost sharing that is at least actuarially equivalent to that provided to UTM enrollees. To simplify CMS’s administrative responsibilities, it would make sense to eliminate plan types that, after a trial period of five or so years, attract few enrollees. If this limitation were in force now, private fee for service plans and high deductible/medical savings plans might be eliminated.[18]

6 A Change That Should Not Be Made

In 1983, Congress enacted an increase from 65 to 67 in the age at which so-called “full” Social Security benefits are paid. This change is often mischaracterized as an “increase in the retirement age.’ In fact, it has no impact on the age at which Social Security benefits are first available (age 62) or the age at which the maximum benefit is payable (age 70). Nor has it had much impact on the age at which most people claim benefits, then and now before age 65. The 1983 change was set to phase in with glacial slowness – it will not be complete for those reaching age 62 until 2022, four decades after enactment. Nevertheless, the 1983 legislation is commonly described as an increase in the “retirement – or “normal retirement” – age. This terminology has led to proposals to raise the age of eligibility for Medicare from 65 to some higher age.

We shall not here take a position on whether or not the Social Security benefit formula should be further reduced along the lines of the 1983 legislation. Rather we note that the 1983 legislation provides no basis for delaying the age of eligibility for Medicare. In our view, other considerations make any such delay unwise. The primary justification for further reducing the generosity of the Social Security benefit formula is that increasing longevity willy-nilly raises the lifetime discounted present value of benefits, a consequence that – according to some – justifies a reduction in annual benefits. However, increases in longevity have been confined largely to those who are well educated and well compensated. Yet any increase in the age of eligibility for Medicare would deny access to benefits equally to people at all educational and income levels. To be sure, the Affordable Care Act could be amended to provide premium assistance to people over age 65 – current law does not authorize it. Even if ACA premium subsidies were broadened, however, the financial burdens on people with modest incomes could be crushing unless assistance were considerably increased. For example, a couple at age 65 living in Florida with an income equal to four times the federal poverty level, $62,920, would face an insurance bill equal to roughly one-fourth of income. The premium for the same couple in San Francisco would be roughly 30 percent of income. For those who do not have access to employment-based coverage – either because they never had it or because they left jobs that provided such coverage – the delay in access to Medicare would extend a period of extreme financial vulnerability.

7 Conclusion

Nothing in the delivery of health care suggests that distinct payment arrangements for people who are under and over a certain age or who are in or out of the labor force is optimal. But the caprice of path-dependence is nowhere more evident than in US arrangements for the financing of health care. Employment-based insurance emerged from a Treasury decision during World War II to exclude the cost of health insurance from compensation subject to wage controls and from income subject to personal tax. Medicare coverage emerged half a century ago as a weird political compromise – something liberals could accept as the initial installment on what they really wanted (full national health insurance) and something conservatives could accept, partly because it came packaged with what, at the time, they thought they wanted (means-tested coverage of the poor, Medicaid).

However happenstance its creation was and however strained the arguments for its continuation are, this bifurcated system has endured and seems certain to survive for many years to come. But the structure of this system, determined as it was by accidents of history, can and should be changed. We have laid out several changes that we think would improve Medicare for beneficiaries and taxpayers alike. This list does not include one reform that we recommended two decades ago. Just as circumstances have changed over the 50-year life to Medicare, they have changed since the concept of premium support entered public discourse. And, as John Maynard Keynes is alleged to have remarked, “When the facts change, I change my mind. What do you do, sir?”

Corresponding author: Henry J. Aaron, The Bruce and Virginia MacLaury Chair, Senior Fellow, Brookings Institution – Economic Studies, 1775 Massachusetts Ave. NW, Washington, DC 200036, USA, e-mail: ;

Published Online: 2015-11-17
Published in Print: 2015-12-1

©2015 by De Gruyter