The FEHBP as a Model for Medicare Reform: Separating Fact fromFiction

Report Health Care Reform

The FEHBP as a Model for Medicare Reform: Separating Fact fromFiction

August 7, 2003 About an hour read Download Report
Walton Francis
Distinguished Fellow

Members of the House-Senate conference on Medicare legislation are deciding the future of the Medicare program. Section 241 of the Medicare Prescription Drug and Modernization Act, passed by the House of Representatives, attempts to create a reformed Medicare system in 2010. The Senate version does not seriously attempt to provide for a consumer-driven version of Medicare reform.

The best model for serious Medicare reform is the Federal Employees Health Benefits Program (FEHBP), the working program that has covered federal workers and retirees for over four decades. Significantly, the authors of the House bill have as an explicit objective the creation of an "FEHBP-style" competitive system for the next generation of retirees. In recent years, however, the FEHBP's performance has been increasingly misrepresented, either directly or by implication, by ardent defenders of the statist Medicare model.

Congressional Opposition

In the meantime, many Members of Congress are strongly opposed to the creation of a Medicare program based specifically on the FEHBP. Recently, Representative Lincoln Davis (D-TN) offered a motion to instruct the House conferees to reject the provisions of Section 241, Subtitle C of Title II of the Medicare Prescription Drug and Modernization Act, which would provide for the application of an FEHBP-style competitive reform beginning in 2010. Although the motion failed by a vote of 221 to 191, the strength of congressional opposition to the FEHBP model is curious, since Members of Congress are invariably enrolled in this popular program. It is also mysterious since the House of Representatives recently passed legislation to protect federal and congressional retirees from the future impact of the Medicare legislation in the House-Senate conference.1

While many Members of Congress somehow believe that the FEHBP is an inferior option for America's retirees, many leading health policy analysts who have examined both the FEHBP and Medicare are convinced otherwise. For example, Harry Cain, former vice president of the Blue Cross/Blue Shield Association and a careful student of both the FEHBP and Medicare, opined several years ago that "the FEHBP has outperformed Medicare every which way--in containment of costs both to consumers and the government, in benefit and product innovation and modernization, and in consumer satisfaction."2

There are ample data to support these conclusions, dispel misconceptions about the FEHBP, and guide Medicare reform. For example:

  • The FEHBP is superior to Medicare in providing access to physicians, health plans, and rural health coverage
    Based on recent data, 99 percent of physicians accept national FEHBP plans; FEHBP enrollees always get a choice of between 12 and 20 plans; and FEHBP enrollees in 87 percent of rural counties in America have chosen from among six or more health plans.
  • The FEHBP is superior in providing innovative benefits and satisfying consumers
    Beyond providing prescription drugs and catastrophic protection, FEHBP plans routinely and rapidly upgrade their benefit offerings. Not surprisingly, 78 percent of FEHBP enrollees in fee-for-service (FFS) or preferred provider organization (PPO) plans and 63 percent of enrollees in health maintenance organizations (HMO) rate their plans at 8 or higher on a scale of 1 to 10.
  • The FEHBP is superior in controlling costs
    Based on data comparisons over 28 years, the FEHBP ties Medicare in cost control without regard to benefit changes over time. Taking into consideration these benefit adjustments, FEHBP costs, as with private-sector insurance generally, have increased less than Medicare's costs over most or all of the life of the Medicare program.

A recent Joint Economic Committee report reached similar conclusions. That study found that over the period 1983 to 2002, the average annual increase in Medicare spending was 6.7 percent, the average annual increase in FEHBP spending was 6.5 percent, and the average annual increase in FEHBP spending without drugs was 5.8 percent.3

Misrepresenting FEHBP Performance

Careful analysis of the FEHBP model is particularly important, as previously noted, because of the extent to which its performance has been subject to misrepresentation. For example, analysts at the Commonwealth Fund recently published an analysis described in its abstract as an answer to those who

seek to remake the federal health insurance program for the elderly...on the model of the...Federal Employees Health Benefits Program. This paper...rebuts those argu-ments by showing that Medicare bene-ficiaries are more satisfied with, have better access to, and have greater confidence about their access to health care, and they report having fewer financial problems as a result of medical bills than those enrolled in private employer plans.4

Likewise, the Kaiser Family Foundation recently published an analysis claiming that the "FEHBP has done...slightly worse than Medicare, on average, [in controlling costs] since 1996."5 The selection of 1996 as a base year leads to a conclusion contradicting the author's own analysis in a prior Kaiser Family Foundation report showing that the FEHBP outperformed Medicare very substantially from 1992 to 1997.6 This same author persistently states that there are only six plans available to all federal employees and retirees when there are in fact 12 plans available to all, and 15 to 20 available to most.7

Yet another report comparing the FEHBP to Medicare, published by Public Citizen, is subtitled "Limiting Choice of Doctors" and is accompanied by a press release saying that "this study shows that the Bush administration's [Medicare reform] plan would really place beneficiaries between a rock and a hard place."8

Upon analysis, all three of these reports contain misrepresentations. Interestingly, the actual data and analysis presented in all three analyses are also far less negative than claimed.

Nine Reasons Why the FEHBP Is Superior to Medicare

REASON #1: Health Benefits
Medicare serves as a lifeline to the elderly of America. Its coverage of hospital and doctor costs is vital to the economic well-being and survival of millions.

Yet Medicare is infamous for its obsolete, vintage 1960 design. It does not provide a catastrophic ceiling on costs even for those costs it covers. It does not cover prescription drugs, except in rare instances. It does not cover many preventive services. It does not cover dental services. By failing to cover health care costs incurred abroad (except in Canada and Mexico), it forces the elderly either to forgo retirement travel outside of North America or to obtain other coverage. Indeed, Medicare coverage is so deficient that over 90 percent of its enrollees purchase supplementary insurance or have it purchased for them.9

The FEHBP has none of these deficiencies. It has painlessly evolved over time through the competitive, consumer-driven process that is its central feature. The Medicare plan can be rated for its benefit coverage in 2003, compared to average FEHBP plans. For a retired person without dual coverage, for example, cost comparisons demonstrate that FEHBP retirement benefit coverage is far superior to Medicare's. (See Table 1.)


There is another significant dimension of benefit superiority. In both programs, the great majority of common hospital and physician procedures are covered routinely. However, at the margin, Medicare coverage choices are dictated either by statutory law or by administrative law dictated through the Medicare coverage processes. Although there is some variation by area because of carrier discretion, this tends to be minimal. Further, all Medicare HMOs are required to offer benefit coverage identical to that in traditional Medicare.

In the FEHBP, coverage choices are made by individual plans. This means that consumers can seek out plans that have better coverage for particular services of importance to them. Acupuncture, cardiac rehabilitation, expensive dental procedures, and other services are usually available, at a price, in some available plan. Medically proven procedures, such as pancreas-only transplants and the latest advances in pacemakers, are covered in all or almost all FEHBP plans but are often covered by Medicare only after years of delay, if ever. And FEHBP plans are free to cover, and often do cover, services that they would not ordinarily cover if the services are approved as part of a case management package tailored to a particular enrollee's needs.

During the 10-year period ending in 1992, out-of-pocket costs in the FEHBP for a market basket of hospital, medical, drug, and dental costs decreased from about 32 percent of total costs per enrollee to about 20 percent of total costs.10 This improvement resulted from benefit improvements in both FFS and HMO plans and from a significant shift in enrollment from the former (higher cost) to the latter (lower cost).

Both sources of improvement have largely halted in the past decade, primarily because of rising prescription drug costs and increases in copayments aimed at restraining these costs. Furthermore, copayments play a significant role in restraining FEHBP costs, and plans have very little room left for copayment reduction without facing untenable cost and premium increases.

Finally, as plans approach complete coverage, the margin for further improvements necessarily decreases. However, no such improve-ment has ever occurred in Medicare, whose benefits on a market basket basis have deteriorated over this entire period.

REASON #2: Provider Choice and Access
In a sense, Medicare is one of America's relatively few remaining FFS medical plans. Most private plans either limit provider choices substantially or, as is quite common, provide differential cost sharing depending on whether or not the provider is "preferred." Of course, Medicare is not really a fee-for-service plan since it regulates prices and, indeed, makes it illegal for providers to negotiate higher prices with enrollees and still be reimbursed.11

Almost all of the FEHBP national plans allow enrollees to go "out of plan" and pay only one-fourth of a reasonable charge. These plans' reimbursements are more favorable for "preferred" physicians, but some payment is available whether or not the physician has an arrangement with the insurance company. At worst, the patient pays the bill and then gets reimbursed directly by the insurance company. Every federal retiree can join any of a dozen health plans that reimburse costs for virtually any physician who accepts private patients.

More physicians are available through the FEHBP than through Medicare. The Medicare Payment Advisory Commission, which advises Congress on Medicare physician payment, also surveys physicians. In its most recent report, the panel found that physicians are significantly less willing to accept Medicare patients than private plan patients.12

Specifically, in 2002, over 99 percent of physicians accepted private FFS and PPO patients, but only 96 percent accepted Medicare patients. This is a seemingly small difference, but if it includes a person's doctor or the best specialist in town, it can have a major effect on that person's health care. Until the recently enacted increases in Medicare payments, the proportion of physicians unwilling to accept Medicare patients was apparently about to rise substantially.

In this context, the FEHBP has a significant advantage over Medicare because of its multiplicity of plans. In 2003, every federal employee or retiree, no matter where he or she lives, anywhere in America or anywhere in the world, has no fewer than 12 plan options from which to choose.13

Federal retirees in areas covered by participating HMOs have additional plans from which to choose. Thus, while a retiree in North Dakota or Wyoming may have "only" 12 plan choices, a retiree living in a medium- and large-size city will typically have several more plan options. In the larger metropolitan areas, where the great majority of both Medicare and FEHBP retirees reside, about 20 plan choices are often available to federal retirees.

Public Citizen, a self-styled consumer interest organization, has published a particularly misleading analysis of access.14 It purports to examine the effect of Medicare "beneficiaries' choice of doctor if they were enticed to join private PPO plans." The analysis focuses on four counties in each of five states and compares FEHBP PPOs to Medicare, focusing on primary care and two specialties, cardiology and oncology.

In a sense, the study's findings are unexceptionable. In general, it finds that only one-third to one-half of Medicare participating physicians are preferred providers. Of course, this is true. The purpose of preferred provider panels is to selectively enroll a fraction of all physicians.

The Public Citizen report, however, makes this appear like a sinister result, neglecting to mention that federal employees in these same PPO plans are remarkably satisfied with their access to care, with about 90 percent giving these plans the two highest ratings on "getting needed care," "getting care quickly," and "getting referrals to specialists."

REASON #3: Rural Access
Of particular concern to rural Americans is the absence of plan choices in the areas in which they live. One recent analysis by the Rural Policy Research Institute (RUPRI) shows that under Medicare+Choice (M+C), only 7 percent of rural counties offer Medicare beneficiaries any choice of plan beyond traditional Medicare.15 In contrast, using an overly conservative methodology that significantly understates choice in the FEHBP by using enrollment levels rather than actual plan availability, the RUPRI study finds that 87 percent of rural counties enroll federal employees and retirees in six or more plans and that 98 percent enroll them in three or more plans.

Another study, published by the Kaiser Family Foundation, attempts to minimize the FEHBP's strong rural access by claiming that, "unless participants in more isolated areas are willing to travel long distances or pay extra amounts for care, they may find that only one or two plans offer meaningful access to services."16 This analysis focuses on Lebanon, Kansas, and states that only two plans (actually, four plans because the analyst erroneously ignores dual plan options) offer preferred primary care providers within 25 miles.

This is true, but three of the dismissed FEHBP plans (including the National Association of Letter Carriers plan and the two Mailhandlers plans) use the FirstHealth network and offer 694 preferred physicians and clinics within 50 miles, a large total for a town that is not within one hundred miles of a metropolitan area. Further, why should the study cavalierly dismiss these and other plans when they pay 70 percent or 75 percent of the charge for any physician in or near Lebanon simply because this requires participants to "pay extra amounts for care" compared to the less costly preferred provider rate? Why is 75 percent reimbursement of reasonable physician charges characterized as less than "meaningful access" to services?

Regardless of how one characterizes access in Lebanon, Kansas, a proper comparison would cover a larger number of rural areas because preferred provider networks vary from place to place and no network is equally comprehensive everywhere. Furthermore, the fundamental access problem for rural Americans encompasses specialist care and hospitals, not just primary care.

One of the interesting areas in the RUPRI analysis is Kenedy County in southwest Texas. This county has fewer than 500 residents, and RUPRI scores it as one of the 2 percent most underserved areas in the FEHBP because federal employees and annuitants among these residents have signed up for no more than two plans.17 Residents of Sarita, the primary town in this Texas county, have no physicians or hospitals that are preferred providers within 20 miles under the FirstHealth network.

But Kenedy County is only one county removed from Corpus Christi, Texas. Using a 50-mile-radius search that reaches that metropolitan area, Sarita residents have 13 hospitals and 694 physicians and clinics available under the supposedly inferior FirstHealth network.18

FirstHealth may not be quite as comprehensive as Blue Cross or the other FEHBP networks, but it does contract as preferred provider with over 4,000 hospitals and almost 400,000 ambulatory providers. With this kind of reach, it obviously provides substantial preferred provider access to virtually all rural residents of the United States. The same can be said for all of the provider networks used by the national FEHBP plans.

Public Citizen charges that Medicare beneficiaries would have highly limited access to care in PPOs. Table after table shows only zero, one, or two preferred specialists available in rural counties.19 Much of this alleged scarcity of physician access results from a cleverly misleading analytic approach.

The Public Citizen report counts the number of participating physicians in each plan by county, but ordinary Americans do not get health care by county. Neither hospitals nor physicians are randomly scattered throughout counties. Instead, they tend to cluster together in cities, often in cities just across county lines. People in rural areas around those cities do not look across the road to the farming village for specialist health care; instead, they travel 10, 25, or 50 miles to the city for the specialized care they need.

As an example, Public Citizen presents tables showing that no FEHBP plan has more than one preferred cardiologist or oncologist--and most have none--in Franklin County, Maine. In fact, taking Farmington (the largest town in Franklin County) as the starting point, several FEHBP plans have 50 or more preferred provider cardiologists and 10 or more oncologists within 50 miles.

REASON #4: Innovation and Reform
The importance of health plan choices, of course, goes far beyond serving patient needs for provider choice and benefit options. The FEHBP, like most other services in the U.S. economy, relies on competition for consumers to improve quality and restrain costs. For example, plans are free to add, drop, increase, or decrease deductibles.

These are not trivial decisions. Deductibles have substantial effects on consumer acceptance, premiums, and health care utilization. Plans that strike the right balance do best over time. Persistent wide variations in deductibles over time suggest that there is more than one "right" model.

In fact, most plan benefits are quite stable. Deductibles are infrequently changed. But some benefits do change rapidly in most plans.

Notable for experimentation and change are plan payments for prescription drugs. Ten or 15 years ago, most plans charged either a nominal copayment or a modest coinsurance percentage for all drugs. Enrollees were free to go to the drug store of their choice. Mail order and formularies were almost nonexistent.

In the past decade, with ever increasing spending on drugs--reflecting mainly new drugs with major new therapeutic benefits--plans have vigorously changed their approaches. Today, most plans have a six-tier benefit structure for drugs. There is one set of copayments for mail order, and another somewhat higher set is for preferred pharmacies. Generic drugs cost the enrollee the lowest copayment, preferred name-brand drugs on the formulary cost somewhat more, and other name-brand drugs cost the most.

One can only imagine the political turmoil and potential for unnecessarily costly or constraining decisions if price controls and formularies were proposed as features of a Medicare drug benefit. It is inconceivable that such a benefit, once enacted into law under the standard Medicare approach, would receive the kind of nimble evolutionary adjustments used in the FEHBP as plans jockey for the best mix of generosity and cost control in order to attract customers.

Current FEHBP drug benefit structures place both the burden and the opportunity for decision-making on the enrollee. They encourage frugality but allow for medical necessity. They have evolved virtually without political controversy and without legislative and bureaucratic fiat. This has proven effective in restraining drug spending and saving both the payer and the enrollees a great deal in premium costs.20

Based on RAND research, the annual savings to the FEHBP from current tiered payment systems is approximately $500 million annually: about 2 percent to 3 percent of program-wide premium costs, shared by the government and enrollees.21 Additional savings from the use of pharmacy benefit managers (PBMs) may equal or exceed those from tiered copayments. Adoption and continuing reform of prescription drug and other benefits in the FEHBP has been politically and programmatically painless while saving billions of dollars over time.

"Open season" is the annual opportunity, each fall, for federal employees and annuitants to "vote with their feet" by switching plans. Although only about 5 percent change plans each year, this annual switching generates relentless and continuing pressure on all plans to adapt and improve services while controlling costs. In contrast, most private employers frequently attempt to lower costs by changing their single plan from one insurance company to another. This imposes major disruptions on their employees and their families, who are forced to change physicians when involuntarily transferred from Plan A to Plan B.

Paradoxically, the seemingly radical FEHBP system of continuous competition is far more stable. This stability benefits enrollees not only directly and immediately, but also over time, since plans retain incentives to invest in preventive care today to avoid higher expense years in the future.

REASON #5: Consumer Satisfaction
Consumer satisfaction is difficult to measure fairly, particularly in comparing Medicare to the FEHBP. However, the innovative use of quality information in the FEHBP program by the Office of Personnel Management (OPM) has led to the adoption of participant surveys. These surveys measure overall satisfaction as well as specific dimensions of plan performance, such as getting needed care, how well doctors communicate, and claims processing. By providing this information to enrollees, the OPM has significantly aided them in plan selection.

The most recent survey shows that about 79 percent of FFS and PPO enrollees and 63 percent of HMO enrollees rate their plans 8 or higher on a scale of 1 to 10.22 Taking into account both open season movement and survey results, the overall level of enrollee satisfaction with the FEHBP is clearly very high.

A recent Commonwealth Fund Survey of Health Insurance compared Medicare and private insurance.23 It found that 85 percent of Medicare elderly rated their plan as good, very good, or excellent. In contrast, "only" 81 percent of those privately insured and of working age rated their plans as highly. However, these results really prove nothing. It is well-known that plan satisfaction increases with age of respondent. Younger enrollees are far more critical.

This largely explains the differential between FFS and PPO ratings in the FEHBP, since HMOs disproportionately attract younger enrollees. HMOs enroll 40 percent of federal employees but only 10 percent of retirees. In the Commonwealth survey, an 81 percent favorable rating by those aged 19 to 64, compared to 85 percent favorable among those aged 65 or more, arguably shows that private health plans would actually be rated far higher by consumers than Medicare if available equally to each age group.

Finally, the reported results failed to distinguish between the elderly enrolled in Medicare alone, without any supplementary benefits, and the roughly 90 percent who have supplemental plans, including retirees simultaneously enrolled in Medicare and the FEHBP--an extraordinarily rich benefit combination. Thus, the results say nothing at all about satisfaction with traditional Medicare standing alone. Indeed, the survey shows that the Medicare disabled--a younger group much less likely to have a supplemental benefit--give Medicare only a 66 percent favorable rating. Thus, among the respondents who are below age 65, Medicare scores far worse than private health plans.

Another recent survey, sponsored by the American Association of Health Plans, offers additional evidence on seniors' views of health plans.24 This survey, in which the respondents were exclusively elderly, found that 72 percent of seniors enrolled in traditional Medicare (88 percent among M+C enrollees) believed that a choice of plans was important.

On a variety of measures of plan satisfaction, enrollees in traditional Medicare and M+C showed essentially identical satisfaction levels. For example, 82 percent of the traditional Medicare enrollees and 79 percent of M+C enrollees were very or somewhat satisfied with the benefits they received. This is expected since the overwhelming majority of the former group has supplemental benefits and presumably re-sponded based on their total benefit package. Just as for the Commonwealth survey, one can reasonably assume that those enrolled in traditional Medicare alone, without either supplemental benefits or an M+C option, would have registered far lower satisfaction levels.

REASON #6: Guaranteed Solid Benefits
The FEHBP and Medicare programs differ fundamentally in several ways, one of which is the difference between a "premium support" structure and a "defined benefit" structure. A study by the American Association of Retired Persons argues that the Medicare approach is better because the benefits are "entitlements" that are "protected" by law.25 This line of argument is fundamentally flawed in three ways.

First, statutorily defined benefits can be taken away whether or not they are defined as legal entitlements. The Medicare deductible was set by law at $50 but is now $100. Congress once enacted prescription drug benefits and then repealed them. Indeed, Congress amends the Medicare statute every year. As the program steadily progresses toward insolvency, maintenance of current benefit levels hardly seems assured.

The FEHBP is also an "entitlement," but it is handled differently. The FEHBP premium level is "protected" by law, and the "entitlement" formula that defines it provides a substantially better level of insurance benefits than Medicare. The entitlement says, in essence, that the government pays 75 percent of the average cost of plans that enrollees voluntarily choose. Indeed, unlike Medicare, the FEHBP statute has never been amended to reduce enrollee benefits.

Second, FEHBP benefits have been superior to Medicare benefits for decades. The "defined benefit" has become a guarantee of second-rate benefits, and the allegedly weaker "premium support" guarantee has proven a superior guarantor of benefits.

Third, premiums and benefits can be guaranteed in statute without having every benefit enumerated in excruciating, micromanaged detail as is done with Medicare. Enrollees can be guaranteed by law an actuarially reasonable value of benefits, both overall and in broad categories such as hospital or drugs. Within such constraints, plans can make the decisions, for example, as to which deductibles (if any) to use, where to set deductible levels, where to set copayment and coinsurance levels, whether or not to tier benefits, which treatments to accept as medically proven, and where to set the catastrophic guarantee level.

In fact, this is essentially how the OPM operates the FEHBP. The FEHBP statute could be amended to explicitly guarantee actuarial fairness and soundness tests better than those of Medicare without changing the program in any way.

The "premium support" model used by the FEHBP has proven to be both better and safer as an entitlement than the "defined benefit" Medicare model.

REASON #7: Promoting Consumer Understanding
It is often alleged that consumers, particularly elderly consumers, cannot handle the complications of a competitive plan system. After all, many consumers do not understand traditional Medicare itself.26

However, while choice certainly is more complicated than no choice, no evidence shows that consumer choice poses any more of a problem for health insurance than for any other product or service. The elderly choose their own doctors, automobiles, foods, and living arrangements. Any of these is as--or more--complicated than choosing health insurance. What is truly bizarre about these academic discussions is that they often contain no--or minimal--references to the rich informational resources available to federal retirees.27

Medicare's Communications Problem. Furthermore, criticisms of plan choice implicitly assume that traditional Medicare poses little or no information burden. In fact, traditional Medicare creates difficult informational problems and choices.28

For example, upon turning age 65, most people have a choice among various Medigap plans, but they receive little or no information from Medicare or any other source as to the comparative value of such plans. Low-income beneficiaries may be eligible for Medicare supplement and premium payments, but they are rarely informed of these benefits. If they attempt to explore the benefits, they are faced with the daunting Medicaid bureaucracies. Hundreds of thousands of older workers (e.g., state employees hired before 1986) are not even eligible for Medicare but do not know it.29 Errors in Medicare decision-making expose the elderly to financially disastrous mistakes.

These problems are so serious that one analyst calls for new informational campaigns and for reforming state application processes, arguing for the need to "act now to fix the programs that we already have in place" before modernizing Medicare.30

In contrast, the FEHBP program poses few "gotchas" and is essentially free of complex decision issues. The worst potential financial error arises from the requirement that enrollees participate continuously in the program for five years before retirement to retain benefits after retirement. The most complex decision is the choice at age 65 as to whether or not to enroll in Medicare Part B to supplement the FEHBP benefit.31 In the FEHBP, unlike traditional Medicare, errors in plan enrollment decisions and changing circumstances can be remedied or accommodated each year in the annual open season.

Federal employees and retirees are, on average, better educated than Medicare beneficiaries. The average working American is also better educated than the elderly and far less likely to suffer mental impairments. But no system of choices in our society--whether choices of friends, spouses, foods, automobiles, or anything else--depends on every single consumer's being smart and well-informed. Errors inevitably occur, but that is the price of individual autonomy in decision-making.

Most fundamentally, criticisms of choice based on decision complexity create a ridiculous standard. How many consumers of any age or educational level understand the innate workings of automobiles--the physics of and technology used in engine, transmission, braking, and other systems? Yet, somehow, through magazine ratings, recommendations of friends, test drives, modest government oversight and regulation, past experience, and above all the pressures of a competitive marketplace, the elderly overwhelmingly select and use cars that are effective, durable, safe, comfortable, and economical. Banning competition in the automobile industry because some consumers are ignorant or uninformed, or even incapable of understanding certain complexities, and a few therefore make bad choices would be absurd.

The entire economy rests on consumers making choices among tens of thousands of competing goods and services, choices that are analytically complex beyond even the abilities of Consumer Reports to simplify in its relative handful of comparative analyses. Somehow, despite all these complexities, some critics identify health insurance as the one service that will overwhelm normal cognitive abilities and choice among plans as the one decision that consumers cannot be trusted to make.

Comparative Information. Competitive choice among health plans is certainly facilitated by careful oversight and information dissemination. The OPM has done this, and the private market has provided additional information to consumers and those family and friends who advise them.32 However, most consumers do not rely primarily on these formal and organized information sources. Instead, they use their own experience, the experience of friends and neighbors, and--above all--the market-driven menu of good options to make annual decisions among plans.

Since most FEHBP plans are excellent choices, overwhelmingly satisfying enrollee preferences for benefits, provider choices, responsiveness, and cost, 95 percent or so make the simplest possible choice each year: remaining in the same plan. In contrast to federal employees, the elderly do not have coworkers to advise them on plan selection.33 But seniors have information networks of their own, including an extensive system of counselors located in area aging agencies.

Confusion in choosing among competing products has simply not been a problem for the millions of federal annuitants who, over the years, have benefited from their plan selection decisions. If Medicare is reformed into a pro-consumer choice system, assuring adequate information will not be difficult if the OPM approach is emulated and the private sector is encouraged to supplement government information.

REASON #8: Controlling Adverse Selection
Some argue that any form of multiple plan choice will necessarily lead to destructive risk selection and unpredictable exit and entrance of plans--the dreaded "death spiral." The FEHBP has no system of any kind for managing risk selection.34 In contrast, Medicare ceaselessly searches for improved methods of fine-tuning its risk-management features. Reform of the Adjusted Average Per Capita Cost formula was delayed for a decade or more because no one could devise a perfect system. The long-delayed reform failed again to correct the fundamental design error: that well-managed health care does not in fact cost 50 percent more in one place than another.

There is even a respectable argument that some risk selection is desirable. For example, if people with dental problems tend to join plans with better dental benefits and willingly pay the full marginal cost of their decision, what ethical or managerial principle is violated?

The FEHBP has survived for four decades with no management of risk selection other than the stability inherently produced by its insurance subsidy. Curtis Florence and Ken Thorpe, analysts at Emory University, recently concluded that the program has almost no measurable adverse risk selection.35 An earlier Kaiser Family Foundation study, while critical, nonetheless concluded that the "FEHBP's stability may amount to stable biased selection."36

Whatever circumstances may lead to the "death spiral," they do not obtain in the FEHBP. Amusingly, program critics like to cite the 1990 departure of the Aetna FFS plan from the FEHBP. However, Aetna was dropping all of its FFS products at that time and simply found its increasingly marginal FEHBP position a convenient excuse to leave the program.

A related and often repeated accusation is that FEHBP plans will attempt to "cherry pick" lower-risk enrollees through benefit design, selective coverage of geographic areas, and selective advertising. This is the principal argument advanced for requiring all plans to provide identical benefits. However, no such behavior has ever been observed: The accusation is pure hyperbole.

REASON #9: Achieving Superior Cost Control
The most recent comprehensive examination of cost control found that the FEHBP had actually controlled costs slightly better than Medicare.37 This author's updated analysis of those data now shows that the two programs roughly tie when costs are analyzed without regard to benefit changes. (See Figure 1.) However, when benefit improvements are taken into account, the FEHBP maintains its superiority in cost control.


Each program has good years and bad years, and these do not correspond in any simple way. By careful selection of the base year, it is easy to "prove" that one program outperforms the other. Depending on the length of the comparison (e.g., one, three, five, or 10 years), the answer can vary dramatically.

To get around these problems, one good method is to use multiple rolling averages covering 10 years. This shows long-term performance without the noise that affects shorter comparisons. One needs multiple 10-year comparisons because the latest one can be (and usually is) unduly influenced by a particular good or bad base year in one program or the other. Table 2 shows the latest results. (See the Appendix for the raw data.)


In recent years, both programs have had a 10-year average cost increase of around 5 percent or 6 percent per year. Over the full set of comparisons, the programs have differed by more than 1 percentage point only a few times. Measured this way, the cumulative difference over 28 years is a 1 percent advantage for Medicare.

Another way to view relative performance over time is to chart the average cost per enrollee, using the same data. As shown above, the FEHBP and Medicare both started and ended at almost exactly the same levels over the 28-year period. However, during this period, the FEHBP was consistently below the Medicare level, often by substantial amounts. Hence, the FEHBP cumulatively saved substantial amounts compared to Medicare. And this comparison does not include adjustments for improvements in FEHBP benefits over time.38

In summary, the FEHBP and Medicare programs have virtually identical records over time on controlling costs, ignoring substantial and costly benefit improvements in the FEHBP. Put another way, after accounting for benefit improvements, the FEHBP clearly outperforms Medicare in cost control.

In recent years, however, Medicare has had an advantage, and the future performance of these programs is almost impossible to predict. One substantial problem facing the FEHBP is that, with recent increases in government cost sharing, enrollees pay only about 17 percent of after-tax premium costs (as compared to 25 percent before tax), and incentives to attenuate cost and premium differences are greatly attenuated from those of past years.39

It should not be surprising that the records are broadly similar, since both programs operate in the context of the American health care system, with the same underlying structure of hospitals, doctors, costs, technological changes, and a myriad of other commonalities.

However, viewed another way, it is a surprise. The Medicare Administrator operates a system of price controls. As Congress has so amply demonstrated in its recent flip-flop attempts to set physician, hospital, and M+C reimbursements at the "right" levels (determined in large part by the decibel level of the political outcry), price controls can be set arbitrarily within a fairly broad range. Medicare, therefore, could outperform the FEHBP in reducing premium costs through cutbacks in provider prices and income, benefit reductions, and other government-mandated reductions--health care resources, both human and bricks and mortar, are not perfectly mobile in the short run. Thus, the Medicare budget is set ultimately by what the political system tolerates, not by the market or any objective method.

There is also the question of how Medicare compares to the private sector's cost-control experience generally. One recent and prominent study by Cristina Boccuti and Marilyn Moon, researchers at the Urban Institute, claims that "Medicare can be counted on to control per enrollee spending growth over time, more than private insurers can."40 This study relies on a comparison of Medicare and private insurance payment data derived from National Health Accounts data provided by the agency that administers Medicare. The data purport to show that since the mid-1980s, Medicare has consistently outperformed the private sector in controlling spending on comparable services (e.g., excluding prescription drugs because these are not covered by Medicare).

A competing analysis published by The Heritage Foundation uses the National Health Accounts data together with data from the National Medical Care Expenditure Survey (MEPS) and other sources. It demonstrates that when cost increases are adjusted for benefit improvements, the private sector at large has outperformed Medicare over the last 30 years.41 In other words, whether looking at private spending in general or the FEHBP in particular, benefit-adjusted private-sector costs have increased less than Medicare costs over most or all of the life of the Medicare program.

This cost-control performance has come despite (or because of) higher administrative costs for the FEHBP, paying physicians and other providers more than Medicare,42 and the near absence of direct managerial controls. One reason, of course, is that Medicare lurches from one crisis to another as both consumers and providers find ways to game the system. In the FEHBP, plans are ceaselessly looking for ways to control unnecessary spending, relying on a wide range of techniques. The OPM can urge plans to adopt useful innovations by simple requests, unencumbered by the Federal Register process used by the Centers for Medicare and Medicaid Services, which on average requires years from inception to final publication of binding rules.

For example, it took years of regulatory indecision, and ultimately an act of Congress, to stop Medicare payment for unnecessarily expensive seat-lift chairs, once routinely prescribed by doctors for patients who saw beautiful and expensive lounge chairs advertised on television as covered by Medicare. In the FEHBP, the OPM was not involved, and plans simply agreed to pay for only the most austere models of seat lifts, relying on "reasonableness" clauses in their policies.

The Components of Real Reform

Members of Congress can reform Medicare based on the FEHBP model, but they must build on the best features of the program.43 Specifically:

  • Ensure that the government is a good business partner with private plans
    This means providing a reasonable and predictable level of payment to private plans while allowing them to make changes in the details of their benefits packages to cope with consumer demands and changes in medicine.
  • Promote flexibility
    Just like the FEHBP, health plans should be allowed to decide coverage details. Congress should ensure that service areas are flexible, and exempt competing plans from state mandates and regulations.
  • Encourage existing employer-based plans and FEHBP plans to participate in the new Medicare system
    Individuals should be able to keep their existing coverage and take it with them into retirement if they wish to do so, and that should include both public-sector and private-sector retiree coverage.


The choice before Congress ultimately is between these two models--consumer choice or detailed legislative and bureaucratic control of benefit design, prices, and operational decisions. The food stamp program has long demonstrated that it is possible to have a government entitlement that leaves purchasing decisions almost entirely with consumers rather than legislators or bureaucrats.

By good fortune, Congress has a successful example of the consumer choice model in the FEHBP, which meets the health care needs of 9 million federal employees, retirees, and family members. Surely, Congress can use this model to aid in reforming the Medicare program.

Walton Francis is a self-employed economist and policy analyst and has authored the annual CHECKBOOK's Guide to Health Insurance Plans for Federal Employees for the past two decades. This paper is based largely on the author's testimony before the Senate Special Committee on Aging on May 6, 2003, and before the Senate Finance Committee on June 6, 2003.





1. H.R. 2631 provides that federal retirees must be provided with a drug benefit equal in actuarial value to that provided to federal employees. The effect of the legislation would be to insulate federal retirees from the impact of the House and Senate Medicare legislation, which would result in the dumping of millions of senior citizens out of their private drug coverage. The bill quickly passed the House of Representatives by voice vote on July 8, 2003. For an analysis of Section 241, Subtitle C, Title II of H.R. 1, see Lanhee Chen et al., "An Analysis of House Medicare Legislation," Heritage Foundation Web Memo No. 302, June 25, 2003, at

2. Harry P. Cain, "Moving Medicare to the FEHBP Model, Or How to Make an Elephant Fly," Health Affairs, July/August 1999, at

3. Michael J. O'Grady, "Health Insurance Spending Growth--How Does Medicare Compare?" U.S. Congress, Joint Economic Committee, June 10, 2003, at

4. Karen Davis et al., "Medicare Versus Private Insurance: Rhetoric and Reality," Health Affairs Web Exclusive, abstract in November/December 2002 issue, at

5. Mark Merlis, The Federal Employees Health Benefits Program: Program Design, Recent Performance, and Implications for Medicare Reform, Henry J. Kaiser Family Foundation, May 2003, p. 9, Table 3, at

6. Mark Merlis, Medicare Restructuring: The FEHBP Model, Henry J. Kaiser Family Foundation, February 1999, p. 36, Table 13, at

7. Merlis repeatedly states that only six plans are available when his own Table 2 shows 11 of the 12 plans available to all federal employees and retirees (it would have shown 12 but for a factual error). See Merlis, The Federal Employees Health Benefits Program, especially pp. 2 and 4.

8. Benjamin Peck, Bush Plan to Privatize Medicare: Limiting Patient Choice of Doctors in Five States, Public Citizen Congress Watch, June 2003, at

9. Medicare Current Beneficiary Cost and Use Survey, 2000 Access to Care, at

10. Walton Francis, "The Political Economy of the Federal Employees Health Benefits Program," in Health Policy Reform: Competition and Controls, ed. Robert B. Helms (Washington, D.C.: American Enterprise Institute, 1993).

11. John Hoff, Medicare Private Contracting: Paternalism or Autonomy? American Enterprise Institute, 1998.

12. Medicare Payment Advisory Commission, 2002 Survey of Physicians About the Medicare Program, 2003, at

13. This includes both "high" and "standard" options offered by the same carrier, since these options always differ significantly in both benefits and premium.

14. Peck, Bush Plan to Privatize Medicare.

15. Timothy McBride et al., "An Analysis of Availability of Medicare+Choice, Commercial HMO, and FEHBP Plans in Rural Areas: Implications for Medicare Reform," Rural Policy Research Institute Rural Policy Brief, March 2003, at

16. Merlis, The Federal Employees Health Benefits Program, p. 13.

17. McBride et al., "An Analysis of Availability," p. 6, map 3.

18. All unpublished data on provider access in Lebanon, Kansas, and Sarita, Kenedy County, Texas, are based on Web searches at FirstHealth, conducted on June 4, 2003. See

19. Peck, Bush Plan to Privatize Medicare, Figure 3 through Figure 7.

20. Geoffrey Joyce, et al., "Employer Drug Benefit Plans and Spending on Prescription Drugs," Journal of the American Medical Association, Vol. 288, No. 14 (October 9, 2002).

21. Walton Francis, testimony before the Subcommittee on the Civil Service, Census, and Agency Reorganization, Committee on Government Reform, U.S. House of Representatives, December 11, 2002, at

22. Walton Francis, CHECKBOOK's Guide to Health Plans for Federal Employees (Washington, D.C.: Washington Consumers Checkbook, 2002), p. 80, retiree version with full text at

23. Davis et al., "Medicare Versus Private Insurance: Rhetoric and Reality."

24. American Association of Health Plans, "Seniors Rally in Support of Medicare+Choice," press release, May 14, 2003, at

25. Craig F. Caplan and Lisa A. Foley, "Structuring Health Care Benefits: A Comparison of Medicare and the FEHBP," AARP Public Policy Institute Issue Paper, May 2000, at

26. Nora Super Jones, Communicating to Beneficiaries About Medicare+Choice: Opportunities and Pitfalls, National Health Policy Forum Issue Brief, July 24, 1998, at

27. Jones, Communicating to Beneficiaries About Medicare+Choice, and Medicare Payment Advisory Commission, "Structuring Informed Beneficiary Choice," Chapter 4 of Report to the Congress on Selected Medicare Issues, June 1999, at

28. David Carliner, "Getting the Elderly Their Due," Health Affairs, November/December 2002, at

29. Gordon Schiff, "An Unsuspecting American with No Medicare Coverage--Me!" Health Affairs, November/December 2002, at

30. Carliner, "Getting the Elderly Their Due."

31. Medicare Part B is a bad financial buy for federal retirees turning age 65, but one virtually forced on them by unnecessary financial penalties and the uncertainty of future political decisions. See Francis, testimony before the Subcommittee on the Civil Service, Census, and Agency Reorganization.

32. For thorough and user-friendly displays of information, see the latest CHECKBOOK's Guide to Health Plans, at, and the OPM Web site, at

33. Robert Reischauer, "Medicare Reform and the Federal Employees Health Benefits Program," testimony before the Senate Committee on Finance, May 21, 1997, at

34. Francis, testimony before the Subcommittee on the Civil Service, Census, and Agency Reorganization.

35. Curtis Florence and Ken Thorpe, "Marketwatch: How Does the Employer Contribution for the Federal Employees Health Benefits Program Influence Plan Selection," Health Affairs, March/April 2003, at

36. Merlis, Medicare Restructuring, p. 44.

37. Francis, "The Political Economy of the Federal Employees Health Benefits Program."

38. The data end in FY 2003 because the current budgetary projections for 2004 are unreliable for both programs. However, Medicare officials have recently announced an unexpected increase of 12 percent in Medicare Part B costs for 2004. Using later estimates for both programs, the FEHBP would likely have outperformed Medicare in the cumulative comparison.

39. Francis, testimony before the Subcommittee on the Civil Service, Census, and Agency Reorganization.

40. Cristina Boccuti and Marilyn Moon, "Comparing Medicare and Private Insurers: Growth Rates in Spending over Three Decades." Health Affairs, March/April 2003, at

41. Joseph Antos and Alfredo Goyburu, "Comparing Medicare and Private Health Insurance Spending," Heritage Foundation Web Memo, April 8, 2003, at

42. Robert Pear, "Critics Say Proposal for Medicare Could Increase Costs," The New York Times, May 6, 2003.

43. For a detailed discussion of the necessary elements of Medicare reform, see Walton Francis, "Nine Tests for Rational Medicare Reform," Heritage Foundation Backgrounder, forthcoming.


Walton Francis

Distinguished Fellow