Illusions of Cost Control in Public Health Care Plans

Report Health Care Reform

Illusions of Cost Control in Public Health Care Plans

July 24, 2009 About an hour read Download Report
robert book
Robert A. Book
Former Senior Research Fellow in Health Economics
Robert is a former Senior Research Fellow specializing in Health Economics.

One of the most hotly debated proposals in health care reform is the establishment of a new "public plan"--a health insurance program operated by the federal government and modeled on Medicare. In most variants of this idea, the public plan would "compete" with existing health plans currently offered by employers, in the individual insurance market, and/or in a new national health insurance exchange. President Barack Obama has recently expressed support for this idea:

I strongly believe that Americans should have the choice of a public health insurance option operating alongside private plans. This will give them a better range of choices, make the health care market more competitive, and keep insurance companies honest.[1]

One prominent articulation of the public plan idea was recently offered by Professor Jacob S. Hacker,[2] a political scientist at the University of California at Berkeley, and promoted by the Institute for America's Future (IAF).[3] Senator Max Baucus (D-MT),[4] Senator Edward Kennedy (D-MA),[5] and Representative Fortney H. "Pete" Stark, Jr. (D-CA)[6] have made similar proposals, and one has been included in draft versions of the forthcoming Kennedy-Dodd bill in the Senate and the recently published draft of the America's Affordable Health Choices Act in the House.

Proponents of the public health plan idea claim that the high cost of American health care is caused by private insurance companies' expenditures on marketing,[7] efforts to deny claims,[8] high executive salaries,[9] unrestrained pursuit of profit,[10] and unwillingness to drive "hard bargains for reduced prices" from hospitals and physicians.[11] They claim that the federal government has superior "bargaining power" that allows Medicare--the nation's largest existing public health plan--to achieve lower costs and slower cost growth, and that the government could achieve similar results with a public plan for the non-elderly.

Proponents also claim that competition from such a public plan would reduce private-sector health care costs by forcing private insurers to either reduce costs to the supposedly lower public plan level or go out of business. Many even claim that if the entire privately insured population were switched to a public plan, enough could be saved in administrative costs alone to pay for covering all Americans who are currently uninsured.[12]

This rationale for creating a new public program modeled on Medicare is based on four erroneous beliefs: (1) that Medicare, compared to private-sector health plans, provides comparable access to health care at costs that grow more slowly than those of the private sector; (2) that Medicare has lower administrative costs than private insurance; (3) that Medicare uses superior bargaining power to reduce health care costs without harm to patients; and (4) that public health plans are more innovative, whereas private health plans only follow the government's lead.

All of these beliefs are demonstrably false. Contrary to the claims of public plan advocates:

  • Total per-beneficiary health care costs are growing faster for Medicare patients than for private insurance patients. Medicare's per-beneficiary patient care costs appear to grow more slowlythan costs in the private sector only if one ignores the fact that Medicare is paying a rapidly shrinking share of its beneficiaries' total health care costs. Total per-beneficiary patient care costs for Medicare patients are growing faster than total costs for patients with private insurance. However, spending by the Medicare program is growing more slowly than private insurance because much of the growth in health care costs for Medicare beneficiaries is offset by increased out-of-pocket spending by beneficiaries and other sources of private-sector funding.
  • Medicare's per-beneficiary administrative costs are substantially higher than the administrative costs of private health plans. The illusion that Medicare's administrative costs are lower comes from expressing administrative costs as a percentage of total costs, including patient care. Medicare's average patient care costs are naturally higher because its beneficiaries are by definition elderly, disabled, or end-stage renal disease patients, so its per-person administrative costs are spread over a larger base of health care costs.
  • Medicare has no "bargaining power." To the extent that the prices that Medicare pays health care providers are lower than prices paid by private health plans, it is because of the government's regulatory power, not because it reduces the actual costs of providing care or has superior bargaining power. Furthermore, lobbyists for physicians have persuaded Congress in each of the past seven years to intervene to block scheduled reductions in the prices that Medicare pays for physician services--and in six of those seven years to replace the reduction with an increase. This experience suggests that Medicare does not in fact have any bargaining power that would enable it to lower prices further, or even to maintain prices at current levels.
  • Historically, public plans have more often been followers, not leaders, in health care delivery innovation. It is private-sector organizations that have introduced new quality-improvement methods and new customer services, as well as disease management and coverage of preventive care.

A public health care plan would not improve the current health care situation and would likely make matters worse. Far from saving enough to cover the uninsured, it would increase the cost of covering even the presently insured at the current standard of care. A public plan could reduce overall spending only at the cost of substantial harm to patients.

Public Plan Claims

Advocates of a new public health care plan modeled on Medicare justify their proposals based on a number of specific factual claims. On closer examination, all of these claims turn out to be false.

Claim #1: Medicare has controlled cost growth better than private-sector health plans.

Fact: When payments from all sources are considered, spending on Medicare beneficiaries is increasing faster than spending on the privately insured.

No one on any side of the health care debate claims that Medicare's per-beneficiary health care cost is or could be lower than the cost for the privately insured. Medicare beneficiaries are by definition over age 65, disabled (as defined by Social Security) for more than two years, or diagnosed with end-stage renal disease. Clearly, Medicare beneficiaries require more health care on average than those in the privately insured population.

However, advocates of the public plan idea do claim that Medicare's costs per person are growing more slowly over time than those of the private sector. For example, the Institute for America's Future confidently declares:

Medicare has controlled health care costs much better than have private health insurers over the last 25 years.…

Private health insurers' average annual spending per enrollee grew 29 percent faster than Medicare spending between 1983 and 2006, and it grew 59 percent faster than Medicare between 1997 and 2006.Centers for Medicare and Medicaid Services data shows that private health insurance spending per enrollee for comparable benefits grew an average of 7.6 percent a year between 1983 and 2006 compared with Medicare's growth of 5.9 percent, or 29 percent more. Between 1997 and 2006, private health insurance spending per enrollee grew an average of 7.3 percent compared with Medicare's growth of 4.6 percent, or 59 percent more.[13]

Each of the factual claims in this declaration is false, misleading, or irrelevant to the conclusion drawn. We will see that:

  1. Medicare benefits are not directly comparable to private insurance benefits, and the source for the figures cited above, the Centers for Medicare and Medicaid Services (CMS), makes no claim that the quoted figures are for comparable benefits.
  2. The figures cited above include only spending by Medicare and private insurance companies, excluding out-of-pocket spending and other sources of funds. They also fail to account for the fact that Medicare is paying a rapidly decreasing share of its enrollees' total health care costs, while private insurers are paying a stable, even slightly increasing, share of costs for the privately insured.
  3. When payments from all sources are considered, spending on Medicare beneficiaries is increasing faster than spending on the privately insured.

In other words, Medicare is not controlling costs. Rather, it is allowing costs to grow faster than costs for private insurance but shifting an increasing share of those costs onto other payers, including the beneficiaries themselves.

Medicare benefits are not comparable to private insurance benefits. The claim of comparable benefits is highly doubtful, and it is based on an obvious misreading of the source data. In fact, CMS, the source of the figures cited in the IAF's report, makes no claim that the per-enrollee figures are for comparable benefits. It lists figures for "common benefits," which refers to the types of services covered by both types of health plans. A note to the CMS data clearly states that "[c]ommon benefits refers to benefits commonly covered by Medicare and Private Health Insurance. These benefits are hospital services, physician and clinical services, other professional services and durable medical products."[14]

Providing common benefits does not necessarily mean providing comparable levels of those benefits. The fact that Medicare and private insurance both cover these types of services does not in any way imply that they cover them to the same extent, and indeed they do not. Medicare requires higher co-payments and deductibles than most private insurance plans and, unlike most employer-sponsored plans, has no catastrophic limit on out-of-pocket expenses.[15] For example, a comparison of Medicare, the Blue Cross Standard Option of the Federal Employees Health Benefits Program (FEHBP), and the median private-sector large employer plan found that the private employer plan was 12 percent to 35 percent more generous than Medicare, depending on health status, and that the FEHBP plan was 6 percent to 48 percent more generous.[16] A congressional study concluded that Medicare is the least generous of the leading forms of health insurance.[17]

Medicare beneficiaries evidently concur with this judgment, because about half of them purchase private-sector Medigap or other private supplemental insurance to help with the high cost of Medicare co-payments and non-covered services. Many others enroll in managed-care plans, or Medicaid if they are eligible. As of 2005, only 11 percent of Medicare beneficiaries relied exclusively on Medicare for their health coverage.[18]

Private insurance pays an increasing share of health care costs for Medicare beneficiaries. The claims of slower spending growth are based on the growth in Medicare spending divided by the number of enrollees, not the total cost of health care for those enrollees. It ignores not only Medicare's high co-payments, but also the growing percentage of Medicare enrollees who are covered by other primary insurance, including private insurance, and therefore cannot actually receive Medicare benefits except in very limited circumstances even if they have substantial health care needs. These individuals are enrolled in Medicare, and many pay Part B premiums ($1,156.80 to $3,699.60 per year, depending on income[19]) but find their eligibility for Medicare benefits restricted by the Medicare Secondary Payer (MSP) program because they are also covered by private insurance.

Under the MSP program, Medicare coverage is by law secondary to other insurance, which means that private insurance is legally obligated to pay before Medicare pays. For MSP beneficiaries, Medicare will pay only if the private insurance payment plus the patient's statutory Medicare co-pay is less than the amount Medicare would have paid in the absence of other insurance. Even then, Medicare pays only the difference (minus the co-pay).[20]

Because most employer-sponsored plans pay more than Medicare,[21] Medicare makes relatively few and typically small payments for MSP beneficiaries. For example, a worker over age 65 could be enrolled in Medicare, pay Medicare Part B premiums, and have substantial health care needs but receive little or no benefit from Medicare. Yet such individuals are still counted in the statistics as Medicare enrollees. Because Medicare pays only a very small portion of their costs, counting them reduces Medicare's average cost per enrollee.

As the Social Security retirement age increases and more people work longer and retain their employer-sponsored insurance longer, the number of MSP enrollees will continue to increase. By law, employer-sponsored health plans may not discriminate on the basis of Medicare eligibility. The law also imposes substantial penalties on Medicare-eligible individuals who do not enroll in Medicare (and begin paying the Medicare Part B premiums) within three months of reaching age 65.[22]

However, the current retirement age for full Social Security benefits is nearly 67.[23] As a result, a substantial number (almost 2.8 million in 2008) and an increasing percentage (6.07 percent in 2008) of Medicare enrollees have employer-sponsored health insurance[24] because either they or their spouses still work.

Number of Medicare Beneficiaries

Indeed, as Table 1 shows, the proportion of Medicare enrollees with employer-sponsored primary insurance has increased steadily, more than doubling from 2.66 percent in 1996 to 6.07 percent in 2008. In addition to these workers, several other categories of enrollees with other primary coverage cost Medicare either zero or substantially less than their actual health care costs due to MSP rules. These include military retirees with TRICARE For Life, certain military veterans who qualify for Veterans Health Administration benefits,[25] and some people covered by workers' compensation or other liability insurance.

The total percentage of Medicare enrollees subject to MSP has more than doubled from 4.0 percent (1.53 million) in 1996 to 8.7 percent (3.98 million) in 2008. As a result, reported growth in Medicare spending per enrollee has slowed down artificially because the increasing proportion of enrollees subject to MSP is partially offsetting the increased spending on those Medicare enrollees who are eligible for primary Medicare benefits.[26]

Medicare's Cost Control Illusion

Because of this and other factors, Medicare's share of total spending on health care services for non-institutionalized Medicare enrollees fell from 72.2 percent in 1997 to 50.8 percent in 2005.[27] The remaining 49.2 percent was covered by a combination of private insurance, including employer-sponsored insurance (for employees and retirees) and individually purchased Medigap plans (21.4 percent); beneficiaries' out-of-pocket spending (15.7 percent); and other public plans, including TRICARE, Veterans Health Administration, and Medicaid (12.0 percent). Total spending per capita in 2005 was $12,157, including $6,180 paid by Medicare and $2,603 paid by private insurance.[28]

Over time, Medicare's share of health care spending for Medicare beneficiaries has fallen, while private insurance's share has risen. In 1997, Medicare's share was 72.2 percent, and the private-insurance share was 12.2 percent. Total spending per capita in 1997 was $5,438, including $3,925 from Medicare and $662 from private insurance.[29]

While total per-capita health care spending for Medicare enrollees increased by 124 percent from 1997 to 2005 (an average of 10.6 percent per year), spending by the Medicare program increased by only 57.5 percent (5.8 percent per year). Meanwhile, private insurance spending on Medicare beneficiaries increased by 294 percent (18.7 percent per year), and out-of-pocket spending by Medicare beneficiaries increased by 205 percent (15.0 percent per year).

Per-Beneficiary Spending Growth

Total health care spending for Medicare beneficiaries is growing more than twice as fast as the part covered by Medicare, shifting larger shares to out-of-pocket spending (from 11.5 percent in 1997 to 15.7 percentin 2005) and nearly doubling the share paid by private insurance (from 12.2 percent in 1997 to 21.4 percent in 2005).

To make a valid "apples-to-apples" comparison between Medicare and private insurance, we can take the amounts spent on private insurance beneficiaries from each source of funds (from the National Health Expenditure tables) and subtract the amounts spent on those who are beneficiaries of both private insurance and Medicare. This will give us the amounts spent on those who are beneficiaries of private insurance but not Medicare. (To obtain per-beneficiary costs, we divide by the number covered only by private insurance). We can then compare cost growth for the non-Medicare population with that for the Medicare population without any overlap between the two groups.

The results of these calculations are shown in Table 2. During the same eight-year period, payments by private insurance for private-only beneficiaries grew 85.5 percent (8.0 percent per year); out-of-pocket costs grew 99.2 percent (9.0 percent per year); and the total cost of health care for private-only beneficiaries grew 81.7 percent (7.7 percent per year).

In other words, although per-enrollee spending by Medicare is growing slower (5.8 percent per year) than per-enrollee spending by private insurance (8.0 percent annually), total spending on health care for Medicare enrollees is growing faster (10.6 percent annually) than total spending on health care for the privately insured (7.7 percent annually).

Medicare spending is growing more slowly because Medicare has become much less generous, paying a rapidly declining share of its beneficiaries' health care costs. As noted above, Medicare's share of its beneficiaries' costs has dropped from 72.2 percent in 1997 to 50.8 percent in 2005, while the share of health care costs paid by private insurance for its beneficiaries has increased slightly from 63.2 percent to 64.5 percent. (See Chart 2.)

Furthermore, people who are enrolled in Medicare and have private insurance are counted as beneficiaries of both.[30] The numbers in Table 2 correct for this double-counting, but the numbers cited in the IAF report do not. Thus, Frank Clemente and Jacob S. Hacker's approach for the IAF counts health care costs for these beneficiaries in both the Medicare and private categories. In the case of Medicare enrollees with private insurance subject to MSP, their approach gives Medicare "credit" for having the enrollees, even though private insurance pays most of their health care costs.

One reason why Medicare's cost growth is underestimated is that the share of enrollees subject to MSP has increased. Another reason is that neither category includes out-of-pocket spending by the beneficiaries or spending by other sources, both of which have been increasing faster for Medicare beneficiaries than for the privately insured. (See Chart 3.)

In summary, to claim that Medicare spending is "controlling health care costs" more effectively than private health plans do, Clemente and Hacker count all of Medicare's enrollees but only some of their health care costs--the costs paid by the Medicare program. Because Medicare's share of the total costs has fallen significantly, Medicare appears to control costs better than private insurance does when in fact the opposite is true.

For an increasing percentage of Medicare beneficiaries, someone other than Medicare is paying more of those costs. When the full cost of health care for Medicare beneficiaries is considered, private health plans are clearly doing a better job of controlling spending growth.

Claim #2: Medicare's administrative costs are lower than the private sector's.

Fact: Per beneficiary, Medicare's administrative costs are substantially higher than the private sector's.

This claim has been repeated in the media so frequently that it is often mistaken as an indisputable fact. For example, Professor Jacob Hacker claims:

Perhaps the most obvious advantage of public insurance is that it is inexpensive to administer. The public Medicare plan's administrative overhead costs (in the range of 3 percent) are well below the overhead costs of large companies that are self-insured (5 to 10 percent of premiums), companies in the small group market (25 to 27 percent of premiums), and individual insurance (40 percent of premiums).[31]

Cost Growth: Medicare vs. Private Insurance

Some go even farther and claim that Medicare administrative costs are so much lower than those of private insurance that, as New York Times columnist Paul Krugman claims, "eliminating the excess administrative costs of private health insurers...would by itself more or less pay the cost of covering all the uninsured."[32]

Hacker does not cite a source for Medicare's administrative costs being "in the range of 3 percent," but this appears to match CMS budget figures, which ranged between 2.8 percent and 3.4 percent from 2000 to 2005.[33] The budget request for fiscal year (FY) 2009 estimated total administrative CMS spending at $18.6 billion, or 2.6 percent of the estimated $703.9 billion in total benefit payments under these programs.[34]

This approach to administrative costs has three major problems:

  • It includes only costs that appear in the CMS budget for Medicare, which does not account for all non-benefit costs of Medicare.
  • Expressing administrative costs as a percentage of total costs is highly problematic because most Medicare administrative costs are accounted for by activities not directly related to the level of health care expenditures. Medicare's beneficiaries need more health care on average than the privately insured, so this approach spreads out Medicare's administrative costs over a larger base. A better approach is to express administrative costs on a per-beneficiary basis, which is more closely related to how the costs are incurred.
  • For private insurance, the prevailing definition of administrative costs includes some actual health care services, such as disease management and on-call nurse consultation. These are inadvertently counted as administrative costs for private insurance. Medicare does not incur these costs because it does not provide these benefits.

Spending from Non-Medicare Sources

In the private sector, corresponding costs incurred by health insurance companies would be included in the administrative costs because administrative costs are calculated as the difference between premiums collected and health benefit claims paid. Therefore, all other costs are necessarily included.[35] This includes the costs of health care services that do not result in a benefit claim, such as disease management and on-call nurse consultations.

Furthermore, the figures for private insurance administrative costs include state taxes on health insurance premiums paid by private insurance plans.[36] The state taxes vary from state to state but historically have averaged just over 2 percent.[37] Medicare is exempt from these taxes because the Constitution bars states from taxing the federal government.

A recent study by Benjamin Zycher estimates the administrative costs that are not included in the Medicare budget, including overall policy management and the interest cost associated with operating capital.[38]

Management determines overall policies for Medicare's goals, coverage, benefit levels, and pricing. In the private sector, a corporate board of directors and top-level executive management make these decisions. Congress, the President, and their staffs make these decisions for Medicare, and the associated expenses are reflected in the operating budgets of Congress and the Executive Office of the President.[39]

Private health insurers must raise capital to cover fixed assets, such as buildings and information systems, and working capital to cover possible variations over time between premium collections and paid claims. This is reflected in interest payments and dividends. The U.S. Treasury handles this function for Medicare.[40] The Medicare Trust Funds reimburse the Treasury for the operational costs of this function, but not for the portion of the interest on the national debt that is attributable to Medicare. In the federal budget, this interest is reported as a lump sum for the entire federal government. The portion attributable to Medicare is not included in the CMS budget, but it is a necessary cost of operating Medicare.

Moreover, these measures of administrative costs ignore the large drag that Medicare taxes impose on the economy.[41] This reduced economic output is caused by the increasingly higher taxes imposed to pay for most of Medicare.[42] No private health plan can incur this type of cost, but it is nonetheless a real cost to the economy.

Medicare Administrative Costs per Beneficiary. The real problem, however, is not so much what is included in administrative costs, but how that figure is interpreted. The figures cited by Clemente, Hacker, Krugman, and nearly all of the policymakers and analysts who discuss health care administrative costs express these costs as a percentage of total program outlays (administrative costs divided by the total of administrative costs and paid benefit claims).

This is not the most informative approach. To understand why, we first note that administrative costs can be divided broadly into three categories:

  • Some costs, such as setting rates and benefit policies, are incurred regardless of the number of beneficiaries or their level of health care utilization. These may be regarded as fixed costs, since they must be incurred regardless of the size of the program.
  • Other costs--such as enrollment, record keeping, and premium collection--depend on the number of beneficiaries, regardless of their level of medical utilization.
  • Claims processing depends primarily on the number of claims for benefits submitted.

Claims processing accounts for only a small share of administrative costs. Claims processing is the only category that is at all sensitive to the level of health care utilization. However, it is correlated with the number of claims paid, not their dollar value or the intensity of service provided, because the cost of processing a $100 claim is generally the same as the cost of processing a $1,000 claim. In the case of Medicare, this category represents only a very small share of administrative costs. In FY 2005, Medicare spent $805.3 million processing claims.[43] This was 4.04 percent of Medicare's administrative costs--which is, in turn, only 0.234 percent (that is, less than 24 cents for every $100) of total Medicare outlays.[44]

Clearly, only an extremely small portion of administrative costs is related to the level of health care benefit claims, and even that is related only tangentially to the dollar value of those claims. Therefore, expressing all administrative costs as a percentage of benefit claims gives a misleading picture of the relative efficiency of government and private health plans.

This is especially the case when comparing plans with populations that have different levels of health care needs. Medicare beneficiaries are by definition elderly, disabled, or patients with end-stage renal disease. Private insurance beneficiaries may include a small percentage of people in these categories, but they consist primarily of people under age 65 who are not disabled. Naturally, the average Medicare beneficiary needs more health care services than the average person with private insurance. Yet the bulk of administrative costs are incurred on a fixed program-level or a per-beneficiary basis. As a result, expressing administrative costs as a percentage of total costs makes Medicare's administrative costs appear lower, not because Medicare is necessarily more efficient but merely because its administrative costs are spread over a larger base of actual health care costs.

In short Medicare's administrative costs are a lower percentage of the total not because Medicare has cheaper administration, but because it has more expensive patients.

Therefore, comparing the administrative costs of Medicare and private health insurance as a percentage of total costs, including health care expenses, is misleading. Indeed, comparing the administrative costs of any health plans with vastly different demographic characteristics on this basis would be misleading. It would be more informative--and a better measure of administrative efficiency--to compare programs on the basis of administrative cost per beneficiary.

To make this comparison, we take the total Medicare administrative costs calculated by Zycher and divide by the number of beneficiaries instead of by total outlays. We make the same calculations for private health insurance. For Medicare, we include only primary Medicare beneficiaries. This excludes enrollees who are subject to the Medicare Secondary Payer rules, because they have private primary health insurance and therefore receive Medicare benefits only in extremely limited circumstances.[45] These individuals are included in the figures for private health insurance, so this approach avoids the problem of double-counting individuals and avoids counting those whose Medicare enrollment costs (and Part B premiums) are largely a waste from the point of view of patient care.

The results are shown in Table 3. On a per-beneficiary basis, Medicare's administrative costs are substantially higher than the administrative costs of private insurance. For 2000-2005, administrative costs per patient averaged about 20 percent higher for Medicare than for private insurance. In 2005, the most recent year for which all necessary data are available, Medicare administrative costs were $509 per person, compared to $453 per person for private insurance.

Actual Health Care Counted as Administrative Costs. The per-beneficiary approach still puts private insurance at a disadvantage because it fails to account for some health care services that are inadvertently counted as administrative costs for the private sector. For example, many private insurers provide disease-management services for patients with chronic conditions and/or on-call nurses for patients to consult by phone. Because insurance companies provide these services directly, they do not result in paid benefit claims. Therefore, they are reported as administrative costs because private-sector administrative costs are reported as the difference between premiums collected and claims paid.

Medicare does not provide these types of services and therefore does not pay for them. Private insurers provide these services to attract customers, but perhaps also in an attempt to reduce health care utilization. Disease management is intended to increase the level of preventive care for those with chronic conditions, benefiting patients and reducing costs in the long run by reducing the need for and costs of associated adverse health events.[46] On-call nurses are intended to direct patients to appropriate care when they cannot contact their primary care physicians. This is believed by some to reduce costs in the long run, for example, by reducing unnecessary emergency room visits and the incidence of more severe adverse health events caused by forgoing appropriate emergency room visits.

Adminstrative Costs of Medicare and Private Health Insurance

In both cases, providing these additional services increases the apparent administrative costs of private insurance and may reduce claims costs. Both effects further increase the apparent percentage of administrative costs in private health plans.

Administrative Costs of Providers. In addition, both Medicare and private insurance impose substantial administrative burdens on doctors, hospitals, and other health care providers. Both our calculations and those of the public plan advocates account only for the administrative costs of the payers, not those of the payees.

Some proponents of public plans argue that the providers' cost of dealing with a multitude of private insurers imposes substantial costs that are not imposed by Medicare. Different insurers have different billing procedures, and many private insurers impose requirements such as precertification for elective hospitalization, prescription drug formularies,[47] utilization review, and coverage limitations for some services. This is certainly true, but Medicare also imposes other costs that are not imposed by private insurers, such as placing the responsibility for verifying eligibility (e.g., whether Medicare is the primary or secondary insurer) on the provider and employing unusual and complex date-of-service restrictions for some services.[48]

Recent studies have estimated physicians' cost of dealing with billing and insurance companies at 27 percent,[49] 14 percent,[50] 10 percent,[51] and 6.9 percent[52] of revenue, but none of them differentiate systematically between the administrative burdens of Medicare and those of private insurance. Nor do they account for all types of costs imposed by Medicare and private insurance. Without such a study, it is impossible to say anything definitive about the relative levels of administrative costs imposed on providers by Medicare and private insurance.

The Medicare Advantage Comparison. Through Medicare Advantage (MA),[53] Medicare beneficiaries can enroll in a privately operated health plan, usually a health maintenance organization (HMO), as an alternative to the traditional fee-for-service (FFS) Medicare plan. The MA plans bid a specific per-capita amount to provide benefits equivalent to Part A and Part B FFS Medicare payments, after which Medicare pays the MA plan a risk-adjusted amount (which is a function of the bid and other factors) for each patient enrolled.

MA plans may also provide additional benefits beyond those provided by the Medicare FFS plan, funded by cost savings achieved relative to per-capita payments and/or by charging enrollees an additional premium. In essence, Medicare Advantage is somewhat like a voucher system--private insurance for Medicare enrollees paid for, at least in part, by Medicare.[54]

Some argue that comparing Medicare's administrative costs with the administrative costs of private insurance plans is inappropriate because, compared to the privately insured population, Medicare patients have higher levels of health care utilization. As noted above, we maintain that making the comparison on a per-beneficiary basis solves this problem because administrative costs are not primarily a function of the level of utilization.

However, others argue that a better way to compare private and public plans is to compare FFS Medicare with Medicare Advantage, because both draw enrollees from the population of elderly, disabled, and end-stage renal disease patients. The idea is that Medicare Advantage is operationally comparable to private insurance because it is operated by private companies and that it has a beneficiary population that consists of people eligible for Medicare.

This claim deserves a closer look to assess its validity. Hacker states this argument as follows:

These administrative spending numbers have been challenged on the grounds that they exclude some aspects of Medicare's administrative costs, such as the expenses of collecting Medicare premiums and payroll taxes, and because Medicare's larger average claims because of its older enrollees make its administrative costs look smaller relative to private plan costs than they really are. However, the Congressional Budget Office (CBO) has found that administrative costs under the public Medicare plan are less than 2 percent of expenditures, compared with approximately 11 percent of spending by private plans under Medicare Advantage. This is a near perfect "apples to apples" comparison of administrative costs, because the public Medicare plan and Medicare Advantage plans are operating under similar rules and treating the same population.[55]

However, Medicare Advantage is not "a near perfect 'apples to apples' comparison" for two main reasons:

First, they are not "operating under similar rules and treating the same population." MA is a different system serving a different population that operates under different rules. Indeed, one major purpose of the MA program is to allow beneficiaries to opt for a program with different rules. Nor does it treat "the same population," even though all Medicare beneficiaries are eligible for MA and 80 percent live in an area with at least one qualified MA plan.[56] MA enrollment exhibits substantial self-selection. Medicare beneficiaries who rate their health status as "fair," "good," "very good," or "excellent" are twice as likely to chose Medicare Advantage as those who rate their health status as "poor." Those who qualify for Medicare based on age are twice as likely to choose Medicare Advantage as those who qualify based on disability or end-stage renal disease. In short, Medicare Advantage tends to attract the healthiest Medicare patients.[57]

This factor would not be a problem in comparing administrative costs on a per-patient basis, but when comparing administrative costs on a percentage-of-costs basis, it presents precisely the same sort of problem--although perhaps to a lesser degree-- as is presented by comparing Medicare to private insurance for the non-Medicare population.

Second, and much more serious, Medicare Advantage is primarily an HMO program. As of 2006, 84.1 percent of MA enrollees were enrolled in HMO plans.[58] Some administrative costs incurred by MA plans, especially HMO plans, include not just the cost of running the health plan, but also costs of administration incurred by providers (e.g., doctors and hospitals). This introduces a serious discrepancy between what costs are included as administrative costs for Medicare FFS and what costs are included for MA plans.

For the Medicare FFS system, reported administrative costs include only costs incurred at the level of administering the health plan. Administrative costs incurred by doctors and hospitals are reflected in the payments made to doctors and hospitals and are counted in the benefit claims paid. This means that administrative costs incurred by physician practices, hospitals, and other providers are not included in the standard measure of Medicare FFS administrative costs. In other words, some costs that are counted as administrative in some MA plans are counted as patient care in FFS.

Yet this does not mean that these costs are entirely ignored and unknown. Congress explicitly recognizes the administrative burden that it places on providers in the complex rule structure used to determine FFS payments to physicians and other professional providers, such as podiatrists and physical therapists.[59]

Clerical and other indirect administrative costs are recognized as part of the practice expense (PE) component of the Resource-Based Relative Value System (RBRVS) that Medicare uses to determine the prices it pays for physician (and other professional) services through the physician fee schedule. The PE component is intended to account for the operational costs of a medical practice beyond the cost of the physician's time and effort. These costs include both clinical expenses, such as medical supplies and pay for nurses and other non-physician clinical staff, and provider-level administrative costs. In 2005, total practice expense accounted for 45 percent of total physician FFS payments (averaged over all physician specialties), of which 38.4 percent was for cost categories that might reasonably be classified as administrative in nature.[60]

This means that 17.3 percent[61] of payments to physicians represents physician-level administrative costs. This amount is not included in the administrative costs for traditional FFS Medicare, but corresponding costs, as well as similar costs for hospital payments, are included in the CBO figures for some, if not all, Medicare Advantage plans.

An administrative cost figure for Medicare FFS that would be comparable to figures reported for Medicare Advantage would be the portion of the payments to providers that are related to their administrative expenses plus the program-level administrative expenses reflected in the CMS budget (3.1 percent for 2005). This would be approximately 19.4 percent for 2005, for example. (Other costs incurred by government agencies in support of Medicare are not included in either figure.)

Administrative costs for MA plans have been reported at different levels. For example, the CBO reported 11 percent for 2005,[62] and the GAO reported 16.7 percent for 2006.[63] Both figures are below the administrative portion of the PE component of physician fee schedule payment, even before adding in the administrative costs of Medicare as a whole.

In short, the claim that private health plans have higher administrative costs than public plans because Medicare Advantage plans have higher costs than the Medicare FFS plan is not a valid claim. When the same cost categories are included for both, Medicare Advantage plans do not have higher administrative costs than the Medicare FFS plan.

The "Bigger Is Always Better" Argument. While only a small portion of administrative costs is related to the level of health care claims, it is difficult to determine from the available data how much of the administrative costs is a function of the number of beneficiaries and how much is incurred at the level of the entire program. This latter category would include policymaking, benefit design, and determination of premiums and payment rates.

If a substantial portion of the administrative costs is fixed relative to both the level of claims and the number of beneficiaries, it is tempting to argue that these costs ought to be spread over as many beneficiaries as possible. Taken to the logical extreme, one might be tempted to argue that incurring these costs for more than one plan is wasteful and, therefore, that the most efficient solution is to include the entire population in a single plan.[64]

Despite this enticing line of reasoning, the evidence indicates that health insurance provision is much more complicated and that this argument fails to account for other important factors and therefore does not hold up in practice.

First, traditional Medicare is only one plan. In the private sector, thousands of health plans cover a total of more than 200 million people.[65] Medicare covers about 42 million Americans and is much larger than the next-largest plan. Each of the thousands of other plans spreads its fixed costs over many fewer people than Medicare does. If Medicare and each private plan had identical program-level fixed costs, the per-person costs for the entire private insurance sector would be substantially higher than the per-person costs for Medicare.

For example, if there were 1,500 private plans and each plan incurred fixed costs equal to Medicare's fixed costs, then the entire private sector would have fixed costs 1,500 times greater than Medicare's fixed costs but could spread the costs over only five times as many people. If this were true, the fixed-cost component of per-beneficiary administrative costs for the private sector would be 300 times Medicare's fixed costs. But even the private sector's most ardent detractors do not make this claim. Rather, the evidence indicates that per-beneficiary administrative costs are lower in the private sector than in Medicare.

It might be argued that in the private sector, either most of the costs are not truly fixed, or private plans are phenomenally more efficient than Medicare at fixed-cost administrative functions, or there is a lot of "free riding" on the activities of others--or perhaps it is a combination of the three. Free riding sometimes leads to economic inefficiencies, but in this case, free riding (if any) would limit any gains from consolidation because those gains have already been realized.

Second, free riding can--and does--go both ways. Medicare's pricing system depends largely on surveys asking physicians to estimate their costs and on cost reports filed by hospitals. When CMS decides to cover a new service under Medicare, it must determine a price to pay. Normally, it tries to obtain data from providers on the cost of providing that service. This sort of data is available only if the service is already being performed and, if it is, usually because at least some private insurers already cover it. In those rare cases in which data from providers are unavailable, either CMS or one of its regional claims-processing contractors essentially has to "make up" a price, based on cost estimates uninformed by appropriate data.

If Medicare were the single payer, this inaccurate and expensive procedure would need to be used for all new services rather than just for a few unusual cases. Without the ability to free-ride off the data obtained from private plans, there would be two adverse outcomes: Medicare's pricing would become even more arbitrary than it already is, and its program-level fixed costs would increase.

Claim #3: "Bargaining power of public health insurance plans significantly reduces provider costs."[66]

Fact: Public plans do not bargain with providers, and bargaining power cannot affect provider costs. When providers use the political process to seek payments higher than those offered by public plans, they often succeed.

Public plans do not actually bargain with providers, at least not in the usual sense of the term. Public plans issue regulations specifying prices that amount to a take-it-or-leave-it offer to providers. Providers cannot bargain with Medicare or other public plans. However, they can lobby Congress, which establishes public plans' payment rules by statute, and Members can often influence CMS decisions in a more informal matter.

When it comes to "bargaining power," the providers almost always win. To the extent that lobbying constitutes bargaining, CMS has demonstrated little if any bargaining power when compared to, for example, physician and hospital groups seeking to influence Congress, the final arbiter of payment rates for federal public plans.

Beginning in 1999, Congress began requiring CMS to adjust the Part B conversion factor each year according to a statutory formula known as the Sustainable Growth Rate (SGR).[67] The conversion factor is a multiplier used to convert "relative value units" for each physician service into the dollar payment for that service. The SGR formula specifies an adjustment of the conversion factor--and, by extension, all payments under the Medicare Physician Fee Schedule--by a uniform percentage. The formula is designed to prevent total Medicare physician spending from growing faster than the Medicare population, the economy, and the amount necessary to implement changes in benefit coverage.

Every year starting in 2002, the SGR formula has called for reducing physician fees--a so-called negative update to the conversion factor. However, every year since 2002, physician groups have lobbied Congress to set aside the SGR and provide a positive or zero update instead. Every year since 2003, they have succeeded. Since 1999, when the SGR formula was supposed to take effect, it has called for a negative update for every year except for 2000 and 2001. Only in 1999 and 2002 did Congress allow the statutory reduction to go into effect. (See Table 4.)

Medicare's "bargaining power," so lauded by public plan advocates, is clearly no match for the lobbying and voting power of thousands of well-organized physicians.

Bargaining power cannot affect provider costs. Notwithstanding the claim in the IAF proposal, no level of bargaining power wielded by a public plan--or any plan--could affect the costs that providers incur in delivering health care services. The costs depend solely on the inputs (e.g., labor, time, equipment, and supplies) needed to produce the services and the prices paid for those inputs.

No amount of bargaining power wielded by a public health insurance plan or anyone else, for example, can enable a provider to stitch a wound with three sutures when four are required. However, payments that are too low might force a provider to provide less than the optimal level of care in order to survive financially.

History of Physician Fee Schedule

Perhaps the IAF meant to say that the Medicare-like bargaining power could reduce not the costs incurred by providers, but rather the prices paid to providers. Indeed, the IAF points out that Medicare pays about 19 percent less to physicians and 25 percent less to hospitals than is paid by private health plans.

Yet this is not so much the result of bargaining power as it is the result of Medicare's position as an effectively monopsonist[68] purchaser of health care for the elderly. Because Medicare controls a substantial percentage of patients, it can cut prices, exploiting providers much as a monopolist exploits consumers. Providers must either accept Medicare's low payment rates or cut themselves off from a sizable percentage of current and prospective patients.[69] As The Washington Post observed, Medicare uses "its 800-pound-gorilla capacity to dictate prices."[70] Of course, the SGR experience shows that, even as a monopsonist, Medicare's ability to reduce prices is limited by the ability of providers to influence Congress.

Both the IAF and Hacker use the term "bargaining power," but no actual bargaining takes place. Rather than utilizing bargaining power with providers, it is more accurate to say that Medicare dictates prices, subject to the lobbying power of providers and the whims of Congress.

Rather than bargaining, Medicare has two bureaucratic procedures for determining relative prices. Payments for hospital services (Part A) are determined according to the Inpatient Prospective Payment System (IPPS) and the Outpatient Prospective Payment System (OPPS). The IPPS is based on the Diagnostic Related Group (DRG) relative weights, a system adopted by Congress in 1983. The OPPS is based on ambulatory payment classifications (APCs), each of which has a payment rate. The payment for each DRG and APC is based on hospital cost reports submitted to CMS, data on wages in the hospital's area, and various "adjustments" for variables such as age distribution and socioeconomic conditions in the hospital's county and the ratio of interns and residents (if any) to the number of beds.

Payments for physicians and other professionals (Part B) are based on the Resource-Based Relative Value System and forecasts of service utilization levels. Each service is assigned a number of relative value units (RVUs) for each cost component of the service. The total number of RVUs is then adjusted for geography and other factors and multiplied by the conversion factor to determine the dollar amount of the payment.

Medicare does not bargain with providers; rather, providers "bargain" with each other for a nominally fixed payment pool. Under the RBRVS, costs are "measured" in an explicitly scheduled lobbying process, not by financial audits. Every year, certain services are scheduled for review by the RBRVS Update Committee (RUC), a committee convened by the American Medical Association (AMA) and made up of representatives of physician specialty societies. Three times each year, the associations representing the relevant physician specialties present their cases that the costs for their services have increased more than the costs of other specialties since the last review.

In effect, the RUC allocates a nominally fixed total amount of Medicare payments among the various physician specialties by assigning each service a number of RVUs. Medicare physician payments for each service are determined by multiplying the RVU value for that service by that year's conversion factor.

As a committee convened by the AMA, which is a private organization, the RUC technically has no legal authority and serves the CMS only in an advisory capacity through the public comment process. However, CMS has accepted the RUC's recommendations for 95 percent to 99 percent of the services under review since 2005 and an average of 94.5 percent since the RUC was established in 1993.[71]

In determining RVUs, lobbying skill, political clout, and provider self-interest substitute for the patients' interests and preferences, which would otherwise be expressed through market mechanisms. This process frequently misallocates health care resources because pricing errors discourage physicians from providing services that are priced too low and encourage them to seek patients who need services that are priced too high.[72]

These complex administrative payment systems focus entirely on crude estimates of the resources expended (costs) in providing each service without any consideration of the relative benefits to patients or the quality of service provided. In other words, the DRG and RBRVS systems specify a higher payment for a high-cost, low-value service than for a low-cost, high-value service. Except for geographic adjustments, every provider receives the same payment regardless of quality, outcome, or patient satisfaction. Not surprisingly, these systems are more often referred to as determining "reimbursements" rather than "prices."

In both cases, CMS simply sets a relative price for each service. The price might vary from region to region or from hospital to hospital, but no one from CMS negotiates with providers on prices for services. The Medicare price is essentially a take-it-or-leave-it offer, published in the Federal Register as a regulation. The closest the process comes to negotiation is when an association of providers hires lobbyists to influence CMS decisions during the public comment period before the regulation is finalized or to lobby Congress to change or override the formulas to their members' advantage. Neither of these is normally considered bargaining in the context of a mutually consensual business transaction.[73]

This phenomenon could not be generalized to "a public health insurance plan that competes with private plans on a level playing field."[74] Medicare does not compete with private insurance on a level playing field, either for providers or for patients. Medicare uses an "any willing provider" principle, meaning that any licensed provider willing to accept Medicare's payment rates and conditions is accepted, regardless of quality or any other factors. Notwithstanding the small but growing number of working Medicare-age individuals, retirees and the non-working disabled still make up more than 90 percent of Medicare's patients, and most have few other options for primary health insurance.

If a public plan tried to compete with private payers on a level playing field for the business of working Americans and offered providers payments 19 percent to 25 percent lower than those offered by private payers, few providers would participate in the public plan. Those who did participate would probably be lower-quality providers who have difficulty attracting patients or even being admitted to private health plans.

If all providers were compelled to accept the public plan and its much lower payment rates to solve this non-participation problem, the competition would not be on a level playing field. Private health plans cannot force providers to participate and accept whatever rate they offer. However, by charging lower premiums, forcing providers to participate in the plan, and paying providers less, the public plan could attract all of the patients and put the private plans out of business.

Some providers would no doubt drop out. For example, financially marginal hospitals would close, some doctors would retire early, and perhaps fewer people would pursue medicine as a career. Those who wanted to remain in business would have no choice but to accept the public plan's lower payments. The result would be a government monopoly--a single-payer government health care--with lower payments, fewer providers, longer wait times, providers with less ability to invest in new technology, and other manifestations of reduced access to care.[75]

This would inevitably start a "race to the bottom" in quality and access to specialized care, adversely affecting the health care of those in the public plan or any surviving private insurance plan with which it would "compete."

For all of the status quo's deficiencies, this race to the bottom would yield considerably worse results. If the public plan compels providers to participate and establishes lower payment rates and concomitant lower premiums, physicians could not drop out and make their living from other patients because there would not be enough other patients. Physicians would be forced to choose between submitting to the plan's terms or ceasing to practice medicine and either retiring or finding another profession. The inevitable result would be much lower payment rates and lower income for physicians.[76]

The only exceptions would be for the few patients who were wealthy enough to pay out-of-pocket for health care services at prices higher than the public plan and for the physicians who could attract their business. The result would be a two-tiered system, as is common in many countries with dominant or "universal" public health plans. The only way to prevent this two-tier system from developing is to outlaw the individual purchase of heath care services.[77]

Patients would also suffer, especially in the long run. Fewer highly talented people would willingly undergo the years of training under difficult working conditions and low pay needed to become physicians, and this would lead inevitably to reduced access to health care and longer wait times. Lower payments would also force physicians to invest less in advanced medical equipment, and physicians would likely spend less time with each patient. In addition, with fewer people undergoing the training necessary to conduct medical research, new treatments and cures would be developed at a slower rate, resulting in slower decreases--and perhaps even increases--in morbidity and mortality.

Claim #4: "Public insurance has pioneered new payment and quality-improvement methods that have frequently set the standard for private plans."[78]

Fact: Private-sector organizations have introduced new quality-improvement methods, new customer services, disease management, and coverage of preventive care.

Given the history of American health care over the past several decades, this is an astonishing claim, supported only by enticing speculation but no examples or evidence.[79]

In fact, in the past several decades, the private sector has produced many new ideas in health care provision. Government health care programs have adopted a few of these, have considered others in recent years, and ought to consider still others. Private-sector innovations include HMOs, new drug treatments, disease management, and on-call nurses.

It is important here to differentiate between traditional fee-for-service Medicare, which is operated by the federal government and serves as the model for new public plan proposals, and Medicare Advantage, which is essentially private health insurance that is paid for in part with Medicare funds. Medicare Advantage has the potential to allow for innovation within certain constraints, but it is not a public plan. On the contrary, as noted above, Medicare Advantage has been criticized by public plan advocates for having characteristics that make it "too similar" to the private health plans that those advocates oppose.

Examples of private sector innovations include:

  • Staff-Model HMO. Kaiser Permanente, a private-sector (not-for-profit) company, pioneered the integrated staff-model health maintenance organization, and the (public) Veterans Health Administration adopted a similar organizational structure. Integrated care is not available in traditional Medicare, nor has this approach been adopted by any other civilian federal agency.
  • Managed-Care HMO. Private insurance companies first developed the managed-care HMO to coordinate care and reduce spending. Medicare has allowed patients access to private-sector HMOs and other private-sector managed-care options through Medicare Advantage, but the traditional Medicare public plan does not use any managed-care practices.
  • Drug Treatments. Over the past 40 years, many drug treatments have been developed, replacing much more costly surgical treatments. Perhaps the most famous example is histamine-2 receptor antagonists such as Tagamet (cimetidine), which greatly reduced the number of patients needing surgery for peptic ulcers, saving money and providing much better results for patients.[80] Nearly every private insurance plan has covered prescription drugs for decades. Medicare began to cover prescription drugs only three years ago and undoubtedly paid for many more expensive non-drug treatments because it did not cover drugs earlier.
  • Disease Management. Private health plans have taken the lead in implementing disease-management programs for chronic illnesses, resulting in better clinical results and, according to some studies, lower costs. Medicare has begun to offer disease management in a very limited fashion years after the private sector, and only as part of the Medicare Prescription Drug Program (Part D), which is run primarily through private-sector drug plans.[81] Privately run Medicare Advantage programs often offer disease-management programs, but the traditional Medicare program does not.
  • On-Call Nurses. Many private health plans have on-call nurses, available 24 hours a day through a toll-free telephone number. This can improve outcomes for patients and may reduce costs in urgent care situations outside of office hours by enabling patients to make better decisions about seeking emergency care. Providing medical advice by phone can reduce the need for emergency care in circumstances when the advice is sufficient for the patient. Medicare has no such program and does not cover any such program. It has no way to pay nurses or other professionals for providing this type of service unless a private plan competing in Medicare Advantage chooses to provide it as part of its benefit package.
  • Transparency and Quality of Health Care Providers. Several private health plans have begun to provide patients with detailed information about the quality of health care providers, based on clinical outcomes achieved by those providers. For example, UnitedHealth's Premium Physician Designation Program is a leader in this field.[82] Medicare has sought to imitate this type of program with the misnamed "pay-for-performance" (P4P) concept, in which physicians are paid small bonuses for strictly complying with treatment protocols that are determined to be appropriate for the average patient, regardless of the needs of any particular patient the physician might be treating. However, Medicare is prohibited by law from excluding low-performing providers from participating in Medicare, and it has no system to identify such providers to patients.[83]

Medicare's contribution to the "innovation" stream has consisted primarily of the RBRVS and DRG payments systems, which encourage payers to focus entirely on providers' costs and explicitly ignore any value provided to patients.

No Cost Advantage to the Public Plan Concept

Despite the claims of "public plan" proponents, the available evidence from the nation's largest and oldest public plan does not indicate that a new or expanded public plan modeled on Medicare could provide Americans with health care that is comparable to that offered by existing private plans, much less at a lower cost. Contrary to their claims:

  • Medicare has not achieved slower cost growth than private plans. It has merely created the illusion of slowing cost growth by shifting an increasing fraction of its enrollees' health care costs onto other payers, including the beneficiaries themselves.
  • Compared to private health plans, Medicare has higher, not lower, per-beneficiary administrative costs.
  • It does not have "bargaining power" to reduce prices, just an administrative pricing system that uses crude measures of provider costs and explicitly ignores the value provided to patients while being subject to intense, often successful lobbying by providers to increase prices.
  • Nor has Medicare been a leader in innovation, quality improvement, or cost control.

The current Medicare program, which covers one-fifth of the American population, has unfunded future liabilities of over $36 trillion.[84] A public plan with Medicare's essential characteristics that covered the entire population--or a significantly larger fraction of it--would not reduce costs and would be even more financially unsustainable.

Furthermore, by its nature, any public plan would be driven by congressional interventions, bureaucratic processes, and lobbying rather than by incentives to innovate in the financing and delivery of quality, efficient health care. This same phenomenon was evident with Fannie Mae and Freddie Mac, "public plan" mortgage companies that were established to compete with private lenders to "keep them honest" and increase levels of home ownership. Driven by congressional interventions, an implicit government guarantee, and lending policies at odds with economic reality, these public mortgage companies collapsed and threw the entire financial system into chaos. A "Freddie Doc" would eventually produce similarly disastrous results.

A public plan "cure" would be no better than the status quo and might well prove to be much worse than the "disease" it is intended to cure. Americans clearly need health care reform, but a public plan is the wrong kind of reform. Contrary to the claims of President Obama, Professor Hacker, Representative Stark, Senator Baucus, and Senator Kennedy, a public plan could not achieve cost savings or substantially reduce the number of uninsured without substantially reducing the quality and access to health care that Americans currently enjoy.

Robert A. Book, Ph.D., is Senior Research Fellow in Health Economics in the Center for Data Analysis at The Heritage Foundation. The author thanks Joseph R. Antos and Walton J. Francis for helpful discussions and comments on earlier drafts, Benjamin Zycher for answering copious questions about administrative cost data, Paul L. Winfree and Tim Carr for help with private insurance enrollment data, and John W. Fleming for designing creative graphic representations of complicated quantitative concepts.

[1]Barack H. Obama, letter to Senator Edward M. Kennedy and Senator Max Baucus, June 2, 2009, at
(June 18, 2009).

[2]Jacob S. Hacker, "The Case for Public Plan Choice in National Health Reform," Institute for America's Future, December 16, 2008, at /static/reportimages/BA335F1ABFD738FF86DD9DAA604DDA87.pdf (June 18, 2009).

[3]Frank Clemente, "A Public Health Insurance Plan: Reducing Costs and Improving Quality," Institute for America's Future, February 2009, at /static/reportimages/E9225BA9CF3B2B184F6B10E4383B6C76.pdf (June 18, 2009).

[4]Max Baucus, "Call to Action: Health Reform 2009," November 12, 2008, at /static/reportimages/3E733E2C52AF16FDC0A31D8C2916547A.pdf (June 18, 2009).

[5]Edward M. Kennedy, "A Democratic Blueprint for America's Future," speech at the National Press Club, Washington, D.C., January 12, 2005, at (June 18, 2009).

[6]Pete Stark, "Medicare for All," The Nation, January 19, 2006, at (June 18, 2009).

[7]Hacker, "The Case for Public Plan Choice in National Health Reform," p. 6.

[8]Clemente, "A Public Health Insurance Plan," p. 25.

[9]Ibid., p. 6.

[10]Ibid., p. 15; Kennedy, "A Democratic Blueprint for America's Future"; Stark, "Medicare for All"; Baucus, "Call to Action," p. 77; and Hacker, "The Case for Public Plan Choice in National Health Reform," pp. 6-8.

[11]Clemente, "A Public Health Insurance Plan." No one seems to notice the contradiction between the two claims that private health plans put a high priority on profits but are unwilling to negotiate lower prices that would increase their profits.

[12]Paul Krugman, "The Health Care Racket," The New York Times, February 16, 2007, at (July 15, 2009), and David U. Himmelstein, Steffie Woolhandler, and Sidney M. Wolfe, "Administrative Waste in the U.S. Health Care System in 2003: The Cost to the Nation, the States, and the District of Columbia, with State-Specific Estimates of Potential Savings," International Journal of Health Services, Vol. 34, No. 1 (2004), pp. 79-86, at
(July 15, 2009).

[13]Clemente, "A Public Health Insurance Plan," p. 11 (original italics; original bolding and a bullet point omitted). Clemente's figures are based on Hacker, "The Case for Public Plan Choice in National Health Reform."

[14]Centers for Medicare and Medicaid Services, National Health Expenditures Web Tables, Table 13, at
(July 8, 2009).

[15]Walton Francis, Putting Medicare Consumers in Charge: Lessons from the FEHBP (Washington, D.C.: AEI Press, forthcoming).

[16]Dale Yamamoto, Tricia Neuman, and Michelle Kitchman Strollo, "How Does the Benefit Value of Medicare Compare to the Benefit Value of Typical Large Employer Plans?" Henry J. Kaiser Family Foundation Medicare Issue Brief, September 2008, at
(July 8, 2009).

[17]"Follow-Up Memorandum to the Distribution of Total Expenses by Source of Payments for Two Groups," technical memorandum to the Joint Economic Committee, May 22, 2003, cited in Michael J. O'Grady, "Health Insurance Spending Growth--How Does Medicare Compare?" Joint Economic Committee, U.S. Congress, June 10, 2003.

[18]Medicare Payment Advisory Commission, A Data Book: Healthcare Spending and the Medicare Program, June 2008, p. 25, at
(June 18, 2009). This is the most recent year for which figures are readily available.

[19]Centers for Medicare and Medicaid Services, "CMS Announces Medicare Premiums, Deductibles for 2009," September 19, 2008, at (June 18, 2009).

[20]Centers for Medicare and Medicaid Services, "Medicare Second Payer and You," modified December 14, 2005, at
(June 18, 2009); "Medicare Secondary Payer Fact Sheet," April 2009, at
(June 18, 2009); and Medicare Secondary Payer Manual, March 20, 2009, at
(June 18, 2009). The rules differ for certain types of employers with less than 20 employees.

[21]On average, Medicare payments to physicians for each service average about 20 percent lower than private insurance payments for the same service. Medicare Payment Advisory Commission, Medicare Payment Policy, March 2009, p. 96, at /static/reportimages/A077B8100F3D56E2158F92C7D14D6F8E.pdf (June 18, 2009).

[22]Medicare late-enrollment penalties are waived for those with employer-sponsored insurance, but this may not be widely known to affected individuals.

[23]Social Security Administration, "Retirement benefits by Year of Birth," at (June 18, 2009).

[24]Susan Y. Hu, extract of CMS Medicare Denominator file and Medicare Enrollment Database, Centers for Medicare and Medicaid Services, Office of Research, Development, and Information, May 14, 2009. The extract was prepared at the author's request and is reproduced in Table 1. These figures include the working disabled and working end-stage renal disease patients. Excluding these individuals, the figure would be just over 2.0 million. The share of the more restricted group has also more than doubled. For full details, see Table 1.

[25]However, TRICARE and the Veterans Health Administration have begun to bill Medicare in certain instances.

[26]Government documents and scholarly studies generally refer to these individuals as "beneficiaries" even though they have extremely limited eligibility for Medicare benefits until they drop or lose their private insurance. In this paper, they are referred to as "enrollees" because they pay Medicare Part B premiums and are technically enrolled in the program.

[27]The 2005 data are the most recent for which this breakdown is available. Figures are for non-institutionalized beneficiaries in traditional fee-for-service (FFS) Medicare. For Medicare, comparable data on non-FFS (Medicare Advantage) beneficiaries are not readily available, and data on institutionalized beneficiaries are not comparable to data on the privately insured population.

[28]Medicare Payment Advisory Commission, A Data Book: Healthcare Spending and the Medicare Program, June 2008, p. 63, at
(June 18, 2009).

[29]Calculated from figures in Centers for Medicare and Medicaid Services, Health and Health Care of the Medicare Population, 1997, pp. 98-117, Tables 4.2-4.6, at /static/reportimages/CD5AC49FD0E2126BDAF5D131C2D36BC7.pdf (July 2, 2009). In addition to the increased proportion of beneficiaries subject to MSP, the increasing use of prescription drugs over time may account for some of the reduced share of costs covered by Medicare. Prescription drugs were not in general covered by Medicare during the years covered by the data above. The Medicare prescription drug program, "Part D," did not exist in 1994. It went into effect on January 1, 2006. To make a valid "apples-to-apples" comparison, prescription drug spending is not included in these figures.

[30]For example, the Census Bureau's Current Population Survey, the only source of population-level data on private health insurance enrollment, double-counts the beneficiaries who have both Medicare and private insurance.

[31]Hacker, "The Case for Public Plan Choice in National Health Reform," p. 5. See also Cathy Schoen, Karen Davis, and Sara R. Collins, "Building Blocks for Reform: Achieving Universal Coverage with Private and Public Group Health Insurance," Health Affairs, Vol. 27, No. 3 (May/June 2008), p. 647.

[32]Krugman, "The Health Care Racket."

[33]Centers for Medicare and Medicaid Services, National Health Expenditures Data, at
(June 18, 2009). See also Benjamin Zycher, "Comparing Public and Private Health Insurance: Would a Single-Payer System Save Enough to Cover the Uninsured?" Manhattan Institute for Policy Research Medical Progress Report No. 5, October 2007, p. 8, at /static/reportimages/B779CF9F63049ED7EEFE7D174E3346E9.pdf (June 18, 2009).

[34]Centers for Medicare and Medicaid Services, Fiscal Year 2009 Justification of Estimates for Appropriations Committees, at
(June 18, 2009).

[35]This difference, which is generally taken to be the administrative costs of private insurance, is actually referred to as the "Net Cost of Private Insurance" in Centers for Medicare and Medicaid Services, National Health Expenditures Web tables, Table 12.

[36]However, self-insured group plans do not pay state taxes. William G. Schiffbauer, "The Level Playing Field Myth: Comparing Administrative Costs for Public, Private Health Insurance," Bureau of National Affairs Medicare Report, April 24, 2009.

[37]State taxes on health insurance premiums have averaged just over 2 percent. Congressional Research Service, "Cost and Effects of Extending Health Insurance Coverage," October 1988.

[38]Zycher, "Comparing Public and Private Health Insurance."

[39]The Medicare Trust Funds do pay for the operation of the Medicare Payment Advisory Commission (MedPAC), which advises Congress on Medicare policy, but not for other expenses, such as congressional staff. Likewise, the trust funds contribute to the operational expenses of the Office of the Secretary of Health and Human Services.

[40]The Secretary of the Treasury is ex officio the managing trustee of the Medicare Trust Funds.

[41]This type of drag is common to nearly all taxes. It is an example of what economists call "deadweight loss."

[42]Michael F. Cannon, "Medicare," in David Boaz, ed., Cato Handbook for Policymakers (Washington, D.C.: Cato Institute, 2009), p. 128, at /static/reportimages/EA22A7E16ADB31FD081A37CB473EECAA.pdf (June 18, 2009).

[43]Centers for Medicare and Medicaid Services, "Justification of Estimates for Appropriations Committees, Fiscal Year 2009," February 2008, p. 27, at /static/reportimages/542AA7F3C79882DB4513785E91472515.pdf (June 25, 2009).

[44]Author's calculations based on ibid.

[45]In general, a Medicare enrollee who has group insurance through an employer or spouse's employer is subject to MSP rules. As discussed above, in general, MSP patients are eligible for Medicare payments only to the extent that their other insurance is less generous than Medicare; even then, Medicare's payment is subject to co-payments and deductibles. Medicare payments for enrollees subject to MSP are therefore few and small.

[46]Our main point here is to note that for private health plans, the costs of operating a disease-management program are attributed to administrative costs, and for Medicare, these costs are not incurred at all. Whether and to what extent disease-management programs actually reduce costs and/or benefit patients is still not well established and is in fact somewhat controversial. Some studies find cost reductions, some find outcome benefits, some find both, and some find neither. The effectiveness and advisability of disease management is a topic beyond the scope of this paper.

[47]Since 2006, this has been a feature of Medicare through Part D (the prescription drug program).

[48]For example, when billing Medicare, a provider must specify whether Medicare is the primary or secondary payer for that patient. If the provider specifies this incorrectly, the claims processor will reject the bill entirely rather than pay the correct amount. One provider reported to the author that the rejection process often takes so long that resubmitting the claim within the time allowed after the date of service is difficult or impossible. Furthermore, the same provider, a specialty pharmacist who provides immunosuppressive (anti-rejection) drugs for transplant patients, explained that because these drugs are included in the payment to the hospital, Medicare will not pay an outside pharmacy for dispensing the drugs when the patient is hospitalized, except for the day when the patient is discharged. However, immunosuppressive drugs are so critical that hospitals will typically not discharge transplant patients without the drugs in hand. Usually, the pharmacy delivers the drugs to the hospital. However, if the doctor keeps the patient hospitalized for an additional day or more, Medicare will not pay the pharmacist because the drugs were dispensed while the patient was still hospitalized. The pharmacist cannot bill the patient because Medicare rules in general forbid billing Medicare patients for Medicare-covered services by a Medicare provider after payment is denied by Medicare. To protect the integrity of the drug supply, state laws prohibit the pharmacist from taking back drugs after delivery. The result is that the pharmacy may dispense drugs for which it can never be paid due to events beyond its control that occur after the drugs are dispensed. In such circumstances, the pharmacy can lose up to several thousand dollars, depending on the drugs prescribed. None of these restrictions apply to private insurance plans, which typically adjudicate pharmacy claims online and initiate the payment process before the drugs leave the pharmacy. See also Joan E. DaVanzo, Allen Dobson, and Ted Kirby, "Assessing the Cost of Dispensing Immunosuppressive Drugs to Medicare Transplant Recipients--An Update," The Lewin Group, April 15, 2007, at
(June 18, 2009). This study concerns the administrative costs associated with dispensing anti-rejection drugs for transplant patients, which are covered under the Medicare Part B Physician Fee Schedule, not under Part D.

[49]Steffie Woolhandler, Terry Campbell, and David U. Himmelstein, "Costs of Health Care Administration in the United States and Canada," New England Journal of Medicine, Vol.349, No. 8 (August 21, 2003), pp. 768-775.

[50]James G. Kahn, Richard Kronick, Mary Kreger, and David N. Gans, "The Cost of Health Insurance Administration in California: Estimates for Insurers, Physicians, and Hospitals," Health Affairs, Vol.24, No. 6 (November/December 2005), pp. 1629-1639, at http://content.healthaffairs
(June 18, 2009).

[51]Julie Ann Sakowski, Jeffrey M. Newman, James G. Kahn, Richard G. Kronick, and Harold S. Luft, "Peering into the Black Box: Billing and Insurance Activities in a Medical Group," Health Affairs, Vol. 28, No. 4 (July/August 2009), pp. w544- w554, at
(June 18, 2009).

[52]Lawrence P. Casalino, Sean Nicholson, David N. Gans, Terry Hammons, Dante Morra, Theodore Karrison, and Wendy Levinson, "What Does It Cost Physician Practices to Interact with Health Insurance Plans?" Health Affairs, Vol. 28, No. 4 (July/August 2009), pp. w533-w543 (May 14, 2009).

[53]Medicare Advantage was previously known as Medicare+Choice or Medicare Part C.

[54]For more details about the payment system for Medicare Advantage, see Lyle Nelson, "Designing a Premium Support System for Medicare," Congressional Budget Office Pub. No. 2596, November 2006, pp. 8-14, at /static/reportimages/2FA0F63A9D5E01BF57C97CA3276944E5.pdf (July 8, 2009).

[55]Hacker, "The Case for Public Plan Choice in National Health Reform," p. 6. See also Nelson, "Designing a Premium Support System for Medicare," p. 12.

[56]For more details about the payment system for Medicare Advantage, see Nelson, "Designing a Premium Support System for Medicare," p. 8.

[57]Medicare Payment Advisory Commission, A Data Book.

[58]Author's calculation based on James C. Cosgrove, "Medicare Advantage Organizations: Actual Expenses and Profits Compared to Projections for 2006," letter to The Honorable Pete Stark, December 8, 2008, U.S. Government Accountability Office Publication GAO-09-132R, p. 8, at /static/reportimages/0D993CC8EA8A67CA4AEA9EEC1AA2D5EE.pdf (July 7, 2009).

[59]The payment methodology for hospitals is even more complex, but it also accounts for hospital administrative costs incurred to comply with Medicare regulations and procedures. This paper focuses on the rules for the physician fee schedule because they are easier to explain. An analysis of the hospital payment system would make the same point.

[60]Author's calculations based on Centers for Medicare and Medicaid Services, table of practice expenses per hour, at
(July 6, 2009).

[61]17.3 percent is 38.4 percent of 45 percent.

[62]Nelson, "Designing a Premium Support System for Medicare," p. 12.

[63]Cosgrove, "Medicare Advantage Organizations," p. 8.

[64]In principle, this argument could be applied not just to health care, but to any product or service that incurs fixed costs for some resource that has no known capacity limit. In fact, AT&T raised this argument in an attempt to keep its monopoly on long-distance telephone service. AT&T argued that, because carrying calls costs virtually nothing once the lines are laid, every line laid by a competitor was a waste of resources. Subsequent events have not borne out this argument.

[65]For example, America's Health Insurance Plans (AHIP), a trade organization, claims to represent "nearly 1,300 member companies." America's Health Insurance Plans, "About AHIP," at
(July 9, 2009). Member companies may offer multiple plans, and thousands of additional plans are offered by self-insuring employers.

[66]Clemente, "A Public Health Insurance Plan," p. 7.

[67]Balanced Budget Act of 1997, Public Law 105-33.

[68]Just as a monopolist is the only seller of something, a monopsonist is the only buyer of something.

[69]Some anecdotal evidence indicates that a significant number of physicians are choosing not to accept Medicare patients, at least in certain geographic regions. For example, see Julie Connelly, "Doctors Are Opting Out of Medicare," The New York Times, April 1, 2009, and American Medical Association, "Member Connect® Survey: Physicians' Reactions to the Medicare Physician Payment Cuts," February 2008, at http://www.ama-
(April 3, 2009).

[70]Editorial, "Reforming Health Care," The Washington Post, April 27, 2009, at
(June 18, 2009).

[71]American Medical Association, "RVS Update Process," 2007, at /static/reportimages/4AB19D60A662C3531F3D7E49298888FD.pdf (June 18, 2009).

[72]Kerry N. Weems and Benjamin E. Sasse, "Is Government Health Insurance Cheap?" The Wall Street Journal, April 14, 2009, at (June 18, 2009).

[73]See Walton J. Francis, "Why a New Public Plan Will Not Improve American Health Care," Heritage Foundation Backgrounder No. 2267, May 5, 2009, at

[74]Clemente, "A Public Health Insurance Plan," p. 5.

[75]For a more detailed explanation of the inevitable effects of single-payer government health care, see Robert A. Book, "Single Payer: Why Government-Run Health Care Will Harm Both Patients and Doctors," Heritage Foundation WebMemo No. 2381, April 3, 2009, at

[76]Some advocates view this is as one of the benefits because it would reduce total health care spending. Of course, delivering lower quality care in smaller quantities is an easy way to reduce health care spending.

[77]Among democratic countries, only Canada has attempted this approach. Private payment of health care services covered by the public plan was completely illegal until a Canadian Supreme Court decision in 2005 overturned the ban. Privately purchased health care in Canada is still unusual. However, Canada never succeeded in preventing a two-tier system from developing, because those who could afford to pay out of pocket often sought (and still seek) health care in the United States.

[78]Hacker, "The Case for Public Plan Choice in National Health Reform," p. i.

[79]Ibid., p. 15-16.

[80]Alan J. McKay and C. S. McArdle, "Cimetidine and Perforated Peptic Ulcer," British Journal of Surgery, Vol. 69, Issue 6 (June 1982), pp. 319-320.

[81]Joan DaVanzo, Allen Dobson, Lane Koenig, and Robert Book, "Medication Therapy Management Services: A Critical Review," The Lewin Group, May 17, 2005, at
(June 18, 2009).

[82]United HealthCare Services, "UnitedHealth Premium Designation Program from UnitedHealthcare," 2008, at
(June 18, 2009).

[83]Weems and Sasse, "Is Government Health Insurance Cheap?"

[84]Centers for Medicare and Medicaid Services, 2009 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplemental Medical Insurance Trust Funds, March 25, 2008, p. 69, at /static/reportimages/36EEA7CB19274AB71936FF847DD439A1.pdf (July 10, 2009).


robert book
Robert A. Book

Former Senior Research Fellow in Health Economics