Law and Contemporary Problems: Richard L. Schmalbeck, The Impact Of Tax-Exempt Status: The Supply-Side Subsidies, 69 Law & Contemp. Probs. 121 (Autumn 2006)

Cited: 69 Law & Contemp. Probs. 121
[*pg 121]

THE IMPACT OF TAX-EXEMPT STATUS: THE SUPPLY-SIDE SUBSIDIES

RICHARD L. SCHMALBECK^*

I. INTRODUCTION

II. BACKGROUND: HISTORY OF THE EXEMPTION OF HEALTH CARE INSTITUTIONS
   A. The Section 501(c)(3) Exemption
   B. Treasury Guidance on Exemption Standards
   C. Uneven Application of Treasury Principles

III. THE IMPACT OF THE SUBSIDIES
   A. How Do the Subsidies Work?
   B. The Exemption
   C. Charitable Contributions Deductions
   D. Tax-exempt Financing

IV. CONCLUSION

FOOTNOTES

I. INTRODUCTION

In their impressive article, "Distributive Injustice(s) in American Health Care," Clark Havighurst and Barak Richman offer several observations about the ways in which American tax law subsidizes nonprofit health care providers, particularly hospitals. They argue that, because qualification for exempt status for a nonprofit hospital requires a demonstration that the hospital will cross-subsidize some services by excess charges imposed on others, only hospitals enjoying monopoly profits can qualify.¹ One of several consequences is a perverse tendency to impose unjustified charges, which fall largely on those middle-income employees who have health insurance, in order to provide the wherewithal for the subsidies of lower-income patients.² Another consequence is that nonprofit hospitals, being monopolists that are barred from distributing profits to shareholders, will tend to accumulate large, and largely unnecessary, surpluses.³ Because their charters typically prohibit the use of these surpluses for any purpose other than provision of health care, there is a tendency for the most successful of these hospitals to become bloated, trapping resources in ways that lead to, and reflect, inefficient investments in health care.⁴

Some of these things are certainly true, and all of them may be so. It is certainly true that tax rules are a critically important part of the economic framework that underlies health economics in the United States. However, the features of greatest importance are not the ones mentioned in the preceding paragraph. Rather, the primary tax rules driving American health care economics are the exclusion of employer-provided health insurance from the gross income of employees⁵ and the deductibility of extraordinary medical expenses by the individuals who bear them.⁶ These features have been [*pg 122] described and analyzed in Lawrence Zelenak's insightful article in this issue.⁷ They constitute something of a demand-side subsidy, enabling consumers to purchase more health care than they often would purchase, or would want to purchase, otherwise.

Havighurst and Richman appear to agree on the primacy of the health-insurance and expense-deduction rules, but they go beyond that to argue that some of what might be called the supply-side tax features, which provide subsidies to certain qualifying nonprofit providers of health care, are also important.⁸ However, these tax rules may well have less significance than Havighurst and Richman seem to accord them. In particular, while some element of cross-subsidy remains a benchmark for exemption, the requirement has become so attenuated over time that its economic importance would now appear to be minimal. And, although nonprofit hospitals may indeed have become vessels of inefficient creation and storage of huge surpluses, this effect is difficult to prove and in any case may be ambiguous in its effects on the economics of American health care.

This article first provides some background and history of the tax rules governing nonprofit health care institutions, then assesses the significance of the subsidies these tax rules create. Such significance is, in short, negligible: the subsidies do not bring any very impressive forces to bear on the market for health care.

II. BACKGROUND: HISTORY OF THE EXEMPTION OF HEALTH CARE INSTITUTIONS

A. The Section 501(c)(3) Exemption

The centerpiece of the supply-side subsidies in the American health care system is the exemption, under Internal Revenue Code (I.R.C.) § 501(c)(3),⁹ from the corporate income tax enjoyed directly by the institutions in question. Some additional ancillary benefits flow indirectly from the tax exemption, including the deductibility of contributions made to those organizations by donors,¹⁰ which undoubtedly benefits the exempt institutions by stimulating such contributions; the ability to issue bonds whose interest payments are [*pg 123] exempt from the federal income tax,¹¹ a feature that lowers the cost of capital available to institutions permitted to issue such bonds; and exemption from certain state and local taxes.¹² These features are generally of lesser and more ambiguous effects than their demand-side counterparts, but may still influence the shape of health care in this country in important ways.

The availability in the health care industry of the subsidy provided by these supply-side tax features is at first puzzling. When one reads the list of categories of organizations entitled to these benefits, it is not immediately evident that there is a category into which a modern nonprofit hospital easily falls. I.R.C. § 501(c)(3) allows exemption for corporations and similar organizations that are "organized and operated exclusively for religious, charitable, scientific, testing for public safety, literary, or educational purposes, or to foster national or international amateur sports competition . . . or for the prevention of cruelty to children or animals . . . ."¹³ Some hospitals do, of course, perform research, and those same hospitals often provide medical education to students, as well as to recently credentialed doctors and nurses. And some hospitals are operated by religious orders. However, it would only rarely be true that educational, scientific, or religious activities would be the primary focus of a medical center. The primary focus is instead upon the treatment of injuries and diseases. How does this focus fit in the matrix of exemption-eligible categories listed above?

It really does not fit; but nonprofit hospitals are nevertheless regarded as eligible for exempt status on the grounds that they are within the more general "charitable" category of § 501(c)(3). That view, however, is more than a little anachronistic as applied to a medical center in the twenty-first century. Although the "charitable" category is something of a catch-all -- a residual description of organizations that seem somehow deserving of this status -- its primary usage is intended to connote organizations whose mission is "relief of the poor and distressed or of the underprivileged."¹⁴ It was substantially accurate to view hospitals in that way a century ago, when the range of medical treatments available was much more limited than it is today and hospitals were frequently little more than dormitories for those who were too ill or infirm to provide for their own sustenance and who had no wealth or family resources on which to draw for support.

Today's hospital is quite a different enterprise, and a dormitory is one of the things it least resembles. The modern hospital houses patients only reluctantly, and then only those in need of the most acute care. Rather than house patients [*pg 124] (or the "sick-poor," as the hospital population was formerly called¹⁵), today's hospitals primarily house expensive diagnostic and treatment equipment and a highly skilled labor force, which together provide very specialized services to patients across a broad range of economic circumstances. Although practices vary widely, the bias in selection of patients is generally not in the direction of serving the poor, but precisely the opposite: the doors are always open to the wealthy and the well-insured, but more grudgingly, if at all, to others.

Nevertheless, the availability of exempt status persists. Those who make and interpret the tax laws have not been completely oblivious to the gradually changing nature of hospitals, but have apparently preferred a series of awkward accommodations of reality to the more difficult task of fundamental tax reform. A brief sketch of those awkward accommodations will help inform our sense of how far the law has strayed from a coherent policy view of the supply-side subsidy embodied in the exemption and charitable contribution rules relating to hospitals.

B. Treasury Guidance on Exemption Standards

Although the exemption and its accompanying charitable deduction have been a part of the tax landscape from the beginning,¹⁶ very little official guidance from the Treasury or Internal Revenue Service ("IRS") appeared before the mid-1950s. A landmark ruling, Revenue Ruling 56-185, was issued in 1956, providing a list of "requirements" for exemption of a nonprofit hospital.¹⁷ In addition to conditions of little relevance here, the ruling explained that such a hospital "must be operated to the extent of its financial ability for those not able to pay for the services rendered . . . ."¹⁸ The clear implication of the paragraph was that an exempt hospital was expected to engage in more or less explicit cross-subsidization among patient groups, with those who could afford treatment paying for the total costs of operating the hospital, including costs attributable to care for those who could not afford to pay the full costs, if they could indeed afford to pay anything at all. A willingness to treat the poor either at diminished rates or without charge was clearly a paramount consideration: "It [an exempt hospital] must not, however, refuse to accept patients in need of hospital care who cannot pay for such services."¹⁹

[*pg 125]

This view was also reflected in a more or less contemporaneous expression of policy embodied in regulations proposed later the same year.²⁰ These regulations reflected a relatively narrow view of the "charitable" category, confined exclusively to relief of the effects of poverty.²¹ This position does not appear to have been offered as a departure from prior law, but rather as a codification of existing practice.²² These regulations were never finalized, but neither were they supplanted by any contradictory guidance over the next thirteen years.

In 1969, however, another ruling and regulation did signal an abrupt change in policy. In Revenue Ruling 69-545, the IRS considered two hypothetical hospitals that were seeking exemptions.²³ As in many revenue rulings, the two cases were presented as polar pairs, on opposite sides of the line dividing the good from the bad.²⁴ The surprise was in the liberality with which the IRS viewed the hypothetical good case. The good case was a community hospital that had an "open staff" -- meaning simply that its facilities were open to any doctor in the community. It was governed by a board whose membership consisted of "prominent citizens in the community."²⁵ The hospital operated an emergency room that had a policy requiring the care of all who needed emergency services, but the general patient policy of the hospital was to "limit[] admissions to those who can pay the cost of their hospitalization [by means of government assistance, private insurance, or personal resources.]"²⁶ Even though any cross-subsidy would be quite limited -- being confined only to the emergency room services -- the IRS concluded that a hospital of this sort was entitled to exemption from tax.²⁷

This ruling was quite controversial at the time. Indeed, it was challenged in Eastern Kentucky Welfare Rights Organization v. Simon²⁸ by a public-interest law firm, primarily on the grounds that by so narrowing the range of the provision of services to indigents, the ruling constituted an impermissible departure from the language of the Internal Revenue Code.²⁹ In effect, the argument was that such a hospital could not be considered "charitable" as [*pg 126] required by the statute and was therefore ineligible for exemption under I.R.C. § 501(c)(3). This argument succeeded at the district court level³⁰ but was reversed on substantive grounds by the court of appeals.³¹ Ultimately, an appeal to the Supreme Court resulted in denial of the claim (vacating the order of the court of appeals) on grounds that the plaintiffs lacked standing to challenge the ruling.³²

In the same year that it issued Revenue Ruling 69-545, the Treasury/IRS promulgated new regulations under § 501(c)(3) that expanded the range of "charitable" far beyond the relief-of-poverty rationale that had prevailed until that time.³³ Together with Revenue Ruling 69-545, these regulations introduced what came to be known as the "community benefit" standard and recognized "promotion of health" for the first time as a legitimate basis for tax exemption, regardless of the financial need of the patient population served.³⁴ Although hospitals seem to enjoy an especially favorable status under this standard,³⁵ it is also the case that other categories of nonprofit organizations enjoyed a more relaxed set of exemption qualifications under these regulations. For example, organizations designed to "lessen neighborhood tensions," "defend human and civil rights secured by law," and "combat community deterioration and juvenile delinquency" also found explicit endorsement of their purposes as ones that could qualify them for exemption under § 501(c)(3).³⁶

In light of the fact that the open emergency room seemed to be the last remaining link to the relief-of-poverty rationale, it is somewhat surprising that subsequent rulings eroded even that. In Revenue Ruling 83-157, the IRS ruled that in the case of specialized hospitals, such as cancer or eye hospitals, in which emergency rooms are clearly needed less and hence are usually not part of such hospitals' facilities, even the open-emergency-room requirement was waived.³⁷ Thus, in the span of less than fifteen years, the standard for exemption went from one that depended heavily on the general availability of care to the poor [*pg 127] to one in which only emergency care needed to be provided to the poor, and not even that if special circumstances made that impractical.³⁸

C. Uneven Application of Treasury Principles

Despite the much-liberalized policy on tax exemption of hospitals and the apparent breadth of the "promotion of health" concept, not all nonprofit organizations that might seem to qualify under this standard have actually managed to achieve tax exemption. For example, in Federation Pharmacy Services, Inc. v. Commissioner,³⁹ a nonprofit firm sold drugs to elderly and handicapped customers at prices intended to cover only the firm's costs.⁴⁰ The tax court was unimpressed that the drugs were sold at a discount from normal retail prices, observing that commercial enterprises frequently did the same.⁴¹ In sum, the tax court simply thought that this entity was, in fact, engaged in a commercial activity, even though it was structured in a way that provided health-related products at below-market prices to a population -- the elderly and handicapped -- that could reasonably have been presumed to be disproportionately poor. Judge Sterrett's summary says it all:

It is clear that [Federation's] exclusive purpose . . . is to sell drugs, an activity that is normally carried on by a commercial profitmaking enterprise[] . . . . We fail to see how the fact that it happens to deal in drugs [converts] it to a section 501(c)(3) organization. If it could be so converted, then so could a store selling orthopedic shoes, crutches, health foods, or any other product beneficial to health.⁴²

Precisely the problem with this interpretation is that once the IRS had expressed its sense that hospitals could qualify for exemption largely on grounds that they served to promote health, one might have thought that vending drugs, orthopedic shoes, crutches, et cetera, if operated on a nonprofit basis, would qualify as well. But apparently that is not so, at least in the view of the tax court.

The story is much the same in the case of health-maintenance organizations (HMOs). The very name of this category of entities suggests that promotion of health is what they are about, and at least some of them also have elements of cross-subsidization that should put them in harmony with the traditional ideas of charity. But they have not generally fared well in their efforts to achieve tax [*pg 128] exemption.⁴³ The IRS position with respect to HMOs has been consistently hostile, and in the few cases that have been tried, that position has ordinarily been sustained by the courts.⁴⁴ The prominent exception, Sound Health Ass'n. v. Commissioner,⁴⁵ involved an HMO that directly operated health care facilities, and thus to some degree resembled a hospital.

HMOs that merely arrange for the provision of care or for covering its cost, rather than providing care directly, have not achieved similar success under the "community benefit" standard. The most recent major decision in this area, IHC Health Plans, Inc. v. Commissioner,⁴⁶ is useful for its summary of the community-benefit standard. That standard, as viewed by the Tenth Circuit panel, requires a health care institution to make its services available to a broad range of the population within its community, but also requires "some additional 'plus.'"⁴⁷ The amorphous "plus" factor can vary, but the Tenth Circuit suggested that devoting surpluses to research or teaching, or providing free or below-cost services, would normally qualify.⁴⁸ One might argue that precisely because any need to provide shareholders with a return on equity investments is absent, a well-managed nonprofit HMO would be able to cover the cost of medical care at prices that would be at least modestly below market prices established by commercial insurance companies. Indeed, the tax court below in this very case found that the HMO's policies and practices "likely allowed its enrollees to obtain medical care at a lower cost than might otherwise have been available."⁴⁹ But providing care at below-market cost was not, apparently, the sort of "plus" factor that the court of appeals was looking for, and exempt status was denied.⁵⁰ A similar outcome, with a slightly different explanation, was reached in the one other case involving an HMO that did not engage directly in the provision of health care.⁵¹

[*pg 129]

Thus the current state of the law in this area can only be characterized as incoherent. Beginning from a historical position in which subsidized health care was the touchstone for exemption, we have moved to a position in which subsidized health care was neither necessary (as in the case of the cancer hospital) nor sufficient (as in the cases of HMOs and the nonprofit drug store). None of this movement was based on statutory change, though of course the promulgation of regulations consistent with the Internal Revenue Code and the decisions of our courts carry the full force of law. But that law makes little sense from a policy perspective. It is law that provides tax subsidies to organizations that institutionalize some degree of cross-subsidization of care and favors organizations that promote health, but not in a consistent manner.

Further, to the degree that some shred of the cross-subsidy flavor remains in the law, it is worth asking whether even that still makes sense in a health care system in which federal and state governments have largely assumed the burden of financing the cost of medical care for the poorest segment of our population (through Medicaid⁵²) and the segment of our population that generally requires the greatest amount of medical care, the elderly (through Medicare⁵³). Under such circumstances, cross-subsidies may involve, for example, situations in which the modestly compensated hourly employees of large employers are subsidizing the health care costs of better-compensated independent contractors who may not carry health insurance, such as real-estate sales workers, smaller construction contractors, or even lawyers in private practice.⁵⁴ Whether this sort of upside-down cross-subsidy is the norm or not, it is clearly among the possibilities under current rules and interpretations, casting serious doubt on the validity of the arguments for exempt status of those health care institutions that succeed in achieving that status.

III. THE IMPACT OF THE SUBSIDIES

But does any of this matter? To put it another way, are the tax subsidies available to certain health care institutions significant enough to make much of a difference? A shorthand answer, to be detailed a bit more below, is that the subsidies are not trivial and could make a difference at the margin in some cases. But the subsidies are probably not of a magnitude that would make them a dominant force in the economics of health care in the United States.

A. How Do the Subsidies Work?

A brief description of the tax features that provide subsidies to those institutions that achieve exemption will be helpful in assessing the importance [*pg 130] of these features. Considered here will be the effects of (1) the exemption from federal income taxes of qualifying nonprofit health care institutions;⁵⁵ (2) the opportunity for donors to such institutions to deduct their donations from their income for federal income-tax purposes;⁵⁶ and (3) the opportunity of exempt health care institutions to issue bonds whose interest payments are excluded from federal income taxes.⁵⁷ These features are ordinarily replicated in various state statutes providing similar exemptions, deductions, and exclusions from state income taxes, but not uniformly so.⁵⁸ A number of other ancillary features associated with exempt status are likely to be of benefit to exempt health care institutions, such as exemption from state and local sales and property taxes,⁵⁹ favorable postal rates, etc.⁶⁰ Although these are no doubt of considerable value to the institutions that qualify, they are too diffuse to be considered comprehensively in a paper of this scope. Instead, only the supply-side federal income-tax elements of the nonprofit medical-services industry will be examined here.

B. The Exemption

The first element of subsidy is of course the exemption itself, under I.R.C. § 501(c)(3). This provision exempts the qualifying institution from whatever income tax that would otherwise apply, which in almost all cases is the corporate income tax under I.R.C. § 11.⁶¹ It seems as though this would be a benefit of great significance, and in some cases it may be. But it is useful to remember that precisely because the institutions in question are organized as nonprofit corporations, they are under no pressure to produce any net income for shareholders that would be subject to tax. And many probably do not. The latest statistics available indicate that health organizations filing information [*pg 131] returns with the IRS for the 2002 tax year reported total revenue of about $550 billion, and total expenses of $541 billion.⁶² Although this yields a (nontrivial) net-income figure of $9 billion, the total revenue figure included some $41 billion of contributions, grants, and gifts, most of which would not be considered as part of gross income under the usual income tax rules.⁶³ Thus, allowing for the exclusion of most of the latter figure would put health organizations into a net-loss position in the aggregate. Further, exempt organizations have no reason, so long as they are indeed exempt, to undertake even the slightest efforts at tax minimization. For example, although the annual tax return filed by charities allows them to take deductions for depreciation of their buildings and equipment, a charity would have no reason to seek the maximum deductions possible in circumstances that would permit a range of options as to depreciation methods to business taxpayers. Similarly, because there is no tax amount against which credits could be claimed, there is not even a line item for such credits on the charitable organizations' return.⁶⁴ Thus, one imagines that if hospitals were suddenly to lose their exempt status, they would be able to arrange their affairs in such a way that they would have an even greater deficit, in the aggregate.

Of course, aggregate negative numbers would not be inconsistent with the possibility that some hospitals, in some years, would achieve significant profits. But another feature of the income tax as applied to businesses may then come into play: the opportunity to net profits and losses across years, using net operating-loss deductions accumulated in loss years to offset income in up to two preceding and twenty following years.⁶⁵ Thus, even among hospitals sometimes showing a profit, only those that consistently do so over time would be exposed to any ultimate tax liabilities.

There is no question that such hospitals are rare, and there is further no doubt that those lucky few are grateful for their exemption from income tax. The aggregate numbers, however, together with one's conjectures about what the aggregate numbers would look like if nonprofit hospitals had an incentive to undertake even a modest amount of tax planning, suggest that not much revenue is lost due to the exemption. Since the subsidy inheres in the lost [*pg 132] revenue, we can safely conclude that federal tax subsidies directly due to tax-exempt status are not likely to be a significant force affecting the market for health services. Notably, the federal government makes no estimate of the "tax expenditure" -- roughly, the revenue foregone because of favorable tax features that deviate from an ideal income tax -- associated with exempt status for health care institutions.⁶⁶ This is probably because exemption could itself be considered among the baseline features of the tax system, so there is no deviation from the norm involved.⁶⁷

C. Charitable Contributions Deductions

In contrast, the deduction available for charitable gifts to qualifying nonprofit health care institutions constitutes a clear, and measurable, subsidy. This tax feature is technically independent of the institution's status under I.R.C. § 501(c)(3) because the charitable contributions deduction has its own set of rules, and its own criteria for deductibility, under I.R.C. § 170(c)(2)(B). However, the language of the latter section is virtually identical to section 501(c)(3), for all purposes relevant here.⁶⁸

The ordinary workings of the charitable contribution deduction are straightforward: if a donor gives a dollar to a charitable organization, such as a qualifying nonprofit hospital, that amount can be deducted from gross income, yielding a savings equal to that one dollar times the taxpayer's marginal tax rate. At the moment, marginal tax rates run up to thirty-five percent, so a contributed dollar may result in a tax savings to the donor of up to thirty-five cents. This could be characterized as a matching-grant program, under which the government pays up to thirty-five cents for every sixty-five cents contributed by the private donor.⁶⁹ Corporate donors face much the same deduction arithmetic, since medium and large corporations are taxed at marginal rates of either thirty-four percent or thirty-five percent.⁷⁰ Of course, some donors are in lower tax brackets and receive correspondingly lesser benefits.⁷¹ And, because the charitable contributions deduction is an itemized [*pg 133] deduction, those taxpayers who do not itemize their deductions receive no tax benefits from their contributions at all.⁷²

In some cases, the benefits of contributions can be even greater because donors can ordinarily deduct the fair market value of appreciated property that they contribute to charities. Suppose, for example, that an entrepreneur has $1000 of founder's stock with a negligible basis that he is considering selling or contributing. If he sells, he will face a capital gains tax of, typically, fifteen percent, leaving $850 after that tax. If he contributes the stock, he will not be taxed on the capital gain, and he will be able to shelter $1000 of unrelated income with a deduction for the contribution of $1000 of property, generating as much as $350 of tax savings. Thus, the net cost of making the gift under these assumptions may be as little as $500 for a $1000 gift, which resembles a 100% matching grant program.⁷³

Although this treatment is quite favorable, its impact on the health care industry is limited by the fact that health care institutions do not seem to be among the favorite targets of donors. Because taxpayers are not required to disclose the names of their charitable donees, no official information on the amount of contributions can be compiled. A recent and widely respected estimate, however, is that charitable gifts to the health subsector were just under $22 billion in 2004.⁷⁴ This is roughly consistent with the Office of Management and Budget's estimate of the amount of federal revenue foregone due to charitable contributions to health organizations, which is that such gifts were associated with a revenue loss of a bit over $3 billion in fiscal year 2004, and with about $3.7 billion in fiscal year 2006.⁷⁵

The impact of the contribution deduction on the economics of health care is thus not large in the aggregate, and it is even further diminished by the fact that [*pg 134] much of what is given is specifically designated for medical research or international programs, not for the direct provision of medical care to Americans.⁷⁶

D. Tax-exempt Financing

The final member of the trio of favorable supply-side tax features is the opportunity provided to qualifying nonprofit health care institutions to borrow money for capital projects by issuing bonds whose interest payments are exempt from federal income tax. I.R.C. § 103(a) permits the exclusion of interest paid with respect to state and local government bonds, and § 103(b), in effect, extends that favorable treatment to any of several types of "qualified bonds," among which are "501(c)(3) bonds," which are described in I.R.C. § 145. Unsurprisingly, "501(c)(3) bonds" are those that finance property owned by organizations described in I.R.C. section 501(c)(3),⁷⁷ so qualification to issue such bonds is essentially automatic upon grant of the basic tax exemption. Although there is a state-by-state volume cap on the issuance of all "private activity bonds,"⁷⁸ (of which section 501(c)(3) bonds are a subcategory), the 501(c)(3) bonds are exempt from that cap.⁷⁹ Until 1997, there were limits on bond financing imposed on each institution.⁸⁰ But even under those limits, hospitals were treated favorably, being exempt from the institutional limits as long as ninety-five percent of any particular bond issue was devoted to hospital construction.⁸¹

Exemption of the interest on 501(c)(3) bonds from income taxation of course allows issuers of such bonds to market them at rates of return that are somewhat lower than issuers of otherwise comparable bonds would need to pay, thus lowering the capital costs of the institutions endowed with this privilege. The Office and Management and Budget estimates that the federal subsidy, in terms of foregone revenue due to the interest exemption feature, will amount to about $2.2 billion in fiscal year 2006.⁸²

[*pg 135]

IV. CONCLUSION

Thus, the three supply-side subsidies provided through the federal tax rules do not, even cumulatively, bring any very impressive forces to bear on the market for health care. Exemption from taxation surely saves some institutions some tax they would otherwise pay, but most institutions would likely be able, by careful tax planning or, more commonly, by having genuinely unprofitable operations, to avoid all or most of the tax. Indirect subsidies provided to health care institutions by the deductibility of contributions to them, and the exclusion of interest paid by them, are not trivial, but add up to less than seven billion dollars per year -- hardly of much significance in an industry whose contribution to the gross domestic product now exceeds two trillion dollars per year.⁸³

One way of putting the supply-side subsidies in perspective is presented by the Office of Management and Budget's list of the top tax expenditures, which ranks the various departures from an ideal income tax in terms of their budgetary effects over the next five fiscal years.⁸⁴ At the very top of the list, hundreds of billions of dollars ahead of the next-most-expensive item, is the exclusion of employer contributions for medical insurance and medical care, amounting to over $760 billion.⁸⁵ This, of course, is the centerpiece of the demand-side subsidy discussed elsewhere in this volume.⁸⁶ In contrast, the charitable-contribution deduction ranks twenty-ninth at a total cost of about $21 billion over the same period;⁸⁷ and the exclusion of interest on hospital construction bonds ranks thirty-ninth, at a total cost of less than $12 billion over this period.⁸⁸

It seems likely, however, that despite the relatively modest dollar value of the federal subsidies on the supply side, the presence of substantial numbers of nonprofit institutions affects the market for medical services in a variety of ways. For example, it seems possible that in an oligopolistic market, the presence of a significant number of providers whose pricing structures may include explicit efforts to cross-subsidize one group of users through above-market charges imposed on other users might create something of a pricing "umbrella" that could be mimicked by the profit-seeking entities in the same market. The mimicking, however, would likely not involve similar cross-subsidies; that is, the profit-seeking entities might employ the "subsidizing" price schedule where the market would bear that structure, without using the surplus created thereby to subsidize anyone but their own shareholders. Similarly, it is possible that inefficiencies that some suspect afflict nonprofit [*pg 136] entities in particular could provide a similar pricing umbrella for the profit-seeking competitors in the same region.

It is also possible that nonprofit hospitals have become repositories of stores of assets that are in excess of any reasonable capital needs of the health care industry. This could be the result of enjoying monopoly profits, and, in effect, having nothing else to do with them (because of the ban on distributions of profits) than to invest them internally.⁸⁹ This is, however, very difficult to demonstrate with available accounting data. As noted earlier,⁹⁰ the annual reports filed by nonprofit hospitals with the IRS show little if any aggregate surplus.

To be sure, "shadow surpluses" could be hidden in either or both of two ways: the potential availability of surpluses could simply lessen pressures on expenses, so that expenses effectively rise to meet the revenues; alternatively or in addition, some of the "expense" could be in the form of depreciation allowed with respect to capital expansions that may be unnecessary, and would in any event have been foregone but for the potential presence of monopoly profits that make expansions feasible. The process could be this: nonprofit hospitals can charge more for their services than those services would cost if efficiently provided, and so they do. The excess is either absorbed in inefficiency or used to pay the debt service on bonds -- themselves issued under favorable terms -- used to finance unnecessary expansion of the hospital's capital plant.⁹¹

So nonprofit hospitals could be awash with potential surpluses that are never reflected in their financial reports. This may be an instance of a failure to prove a negative: the presence of surpluses might prove the presence of monopoly profits, but the absence of surpluses cannot prove the absence of monopoly profits.

But even if there are "shadow" monopoly profits in the nonprofit hospital subsector, it is unclear what the effects of those monopoly profits may be. Some may be quite benign. For example, it is likely that some of the shadow surplus is consumed by the research budgets of university hospitals.⁹² If and to the degree that this is what is going on, it is hardly problematic. Basic research is rich in positive externalities: it often leads to profitable applications, but it is usually not profitable in itself. Recognition of this provides one of the justifications for tax subsidies of basic research.⁹³

[*pg 137]

But in that most nonprofit hospitals do not pursue significant research agendas, this explanation is at best reassuring only as to a small part of whatever shadow surpluses may exist. It must be admitted that more probable candidates to explain the bulk of shadow surpluses are either simply waste -- due to the combination of inefficiency and lack of budget pressures -- or bloat -- an accumulation of excessive amounts of capital within the nonprofit hospital subsector. What if the structure of the nonprofit-hospital subsector is rife with either or both of these?

Waste is difficult to defend. But it may be worth noting that if the health care industry is, in effect, rigged, so that monopoly profits are available, and can be leveraged by the addition to the mix of third-party payer moral hazards,⁹⁴ then the presence of nonprofit participants in the market may simply reduce the monopoly profits available to the for-profit participants. Waste is still waste, and it might be preferable to have for-profits enjoy even higher profits if the alternative is dead-weight loss due to waste. But from the point of view of consumers in this market, it may not make much difference: monopoly prices will prevail, whether the benefits of that are enjoyed by shareholders of for-profit providers, or wasted by nonprofit ones.

Bloat is an equally unattractive characterization, but it too must be analyzed within the context of the health care industry overall. If bloat occurs because nonprofits have nowhere else to go with their surpluses, one must ask what influence that might have on the for-profit participants in the same market? They suffer no comparable barrier to the distribution of accumulated capital within the provider entities; they are free to pay dividends if they wish. In most industries, there is an optimal level of capital investment -- higher in competitive industries, lower in monopolized ones. If the health-services industry behaves according to this principle, then the overinvestment by nonprofits may well be offset by lesser capital investments by for-profit firms in the industry, so that the overall level of capital investment remains optimal.⁹⁵ One might expect to see nonprofits in this scenario to be more likely to have expensive, but relatively infrequently used, items of equipment, while for-profits confine their capital expenditures to items that are reasonably sure to experience high demand.

Of course, a fundamental premise of this symposium is that the health care industry is not like other monopolized industries, particularly in the sense that exploitation of the monopoly opportunities does not take the usual form of reduced output at higher prices. The third-party-payer situation arguably makes possible both the charging of monopoly prices and the expansion of output. If that is the case, then there may be no natural optimum of capital [*pg 138] invested in the industry. However, if that is true, one must ask again whether the presence of nonprofit providers in the industry makes any difference. If the incentives toward bloat are built into the context -- in particular here, the tax-subsidized third-party payer problem -- will they not affect for-profit firms in much the way they affect nonprofit firms? And if so, does the mix of the two types matter? And does it ultimately make any difference that nonprofit firms have an additional constraint that virtually obligates them to pursue bloat as a more or less conscious strategy?

These questions certainly deserve exploration, although they may not be amenable to conclusive resolution. But framing them as I have done in this concluding section brings me back to some of the points emphasized in the preceding ones. In particular, it would seem that the primary forces that lead to the bizarre features of the American health care market are not ones that have much to do with the exempt status of some of the providers within that market. Rather, they have to do with the dominance of third-party payers financed by pre-tax dollars. Solving the mysteries of this market may involve, as it often does in detective fiction, following the money. And the real money in this industry is in the tax-subsidized third-party payers.

FOOTNOTES

*	Professor of Law at Duke University.
1.	Clark C. Havighurst and Barak D. Richman, Distributive Injustice(s) in American Health Care, 69 LAW AND CONTEMP. PROBS. 7, 20-22 (Autumn 2006).
2.	Id.
3.	Id. at 16.
4.	Id. at 20.
5.	I.R.C. §§ 105, 106 (2005).
6.	I.R.C. § 213 (2005). This provision is accurately described as allowing only extraordinary expenses as deductions because it imposes a nondeductible floor on deductions equal to 7.5% of the taxpayer's adjusted gross income. Id. As a consequence, fewer than nine million taxpayers were able to claim a medical expense deduction in 2003-a number that is just 19.7% of all taxpayers who itemize their deductions and less than 6.7% of all taxpayers in the aggregate. Richard L. Schmalbeck, unpublished calculation (on file with author) (based on Michael Parisi and Scott Hollenbeck, Individual Income Tax Returns, 2003, 25 STAT. OF INCOME BULL., No. 2, at 9, 25, 42 (Fall 2005)).
7.	Lawrence Zelenak, Of Head Taxes, Income Taxes, and Distributive Justice in American Health Care, 69 LAW & CONTEMP. PROBS. 103 (Autumn 2006).
8.	Havighurst & Richman, supra note 1, at 22-23.
9.	I.R.C. § 501(c)(3) (2005).
10.	I.R.C. § 170(a) (2005).
11.	I.R.C. §§ 103(a), 141(e)(1)(G), 145(a) (2005). See infra note 57 for further detail on exempt bond financing.
12.	See text and infra notes 58, 59.
13.	I.R.C. § 501(c)(3). This language permits exemption from tax for the organization. Similar language allowing donors to deduct their contributions is found at I.R.C. § 170(c)(2)(B), except that the latter provision does not allow deductions for public-safety testing organizations.
14.	Treas. Reg. § 1.501(c)(3)-1(d)(2) (1990).
15.	See Daloia v. Franciscan Health Sys., 679 N.E.2d 1084, 1086 (Ohio 1997); David Villar Patton, The Queen, the Attorney General, and the Modern Charitable Fiduciary: A Historical Perspective on Charitable Enforcement Reform, 11 U. FLA. J.L. & PUB. POL'Y 131, 162 (2000) (referring to the hospital population as the "sick-poor").
16.	38 Stat. 168 (1913).
17.	Rev. Rul. 56-185, 1956-1 C.B. 202, 203-04.
18.	Id. The requirements of little interest here related to having organizational documents specifying charitable purpose, banning private inurement, and so on. The key point for this discussion is the specific attention devoted to alleviation of the effects of poverty.
19.	Id.
20.	Prop. Treas. Reg. § 1.501(c)(3)-1(b), 21 Fed. Reg. 460, 464 (Jan. 21, 1956).
21.	Id.
22.	The ruling in particular contains references to earlier, less general IRS announcements, such as a 1941 General Counsel's Memorandum that suggests that the positions in the 1956 ruling, later reflected in the proposed regulations, were consistent with a continuous ruling position of the government. I.R.S. Gen. Couns. Mem. 22,554 (June 21, 1941).
23.	Rev. Rul. 69-545, 1969-2 C.B. 117.
24.	And, again as is often the case, the distance between the polar hypothetical hospitals in this ruling is so wide that only limited guidance is provided on the precise location of the line separating qualified hospitals from those that are not.
25.	Rev. Rul. 69-545, 1969-2 C.B. at 117. The hypothetical hospital found not to qualify for exempt status was controlled by the doctors who founded it, was not open to more than a few other doctors, and had a "relatively inactive" emergency room. Id.
26.	Id.
27.	Id. at 118.
28.	(Simon II), 426 U.S. 26 (1976).
29.	Id. at 26.
30.	E. Ky. Welfare Rights Org. v. Shultz, 370 F.Supp. 325 (D.D.C. 1973).
31.	E. Ky. Welfare Rights Org. v. Simon (Simon I), 506 F.2d 1278, 1291 (D.C. Cir. 1974).
32.	Simon II, 426 U.S. at 26-27.
33.	Treas. Reg. § 1.501(c)(3)-1(d)(2) (1990).
34.	Neither phrase appears in the regulations themselves, but the ruling reads, in relevant part: The promotion of health, like the relief of poverty and the advancement of education and religion, is one of the purposes in the general law of charity that is deemed beneficial to the community as a whole even though the class of beneficiaries eligible to receive a direct benefit from its activities does not include all members of the community, such as indigent members of the community, provided that the class is not so small that its relief is not of benefit to the community. Rev. Rul. 69-545, 1969-2 C.B. 117, 118. There is no elaboration on the minimum size of the group that must be benefited in order to qualify as a "community benefit."
35.	See discussion infra note 38of the subsequent rulings that seem to confer special status on hospitals.
36.	Treas. Reg. § 1.501(c)(3)-1(d)(2).
37.	Rev. Rul. 83-157, 1983-2 C.B. 94, 94-95.
38.	To be sure, even after 1969 qualification for exemption was not available to any sort of nonprofit hospital. In addition to the normal exempt-status requirements banning private inurement and the like, Revenue Rulings 69-545 and 83-157 both make it clear that factors relating to the openness of the hospital to all physicians in the community, broad community representation on the board of directors, and the use of any surpluses for research or capital improvements were matters to be considered in evaluating the case for exemption. See Rev. Rul. 69-545, 1969-2 C.B. 117, and Rev. Rul. 83-157, 1983-2 C.B. 94.
39.	72 T.C. 687 (1979), aff'd, 625 F.2d 804 (8th Cir. 1980).
40.	See Fed'n Pharmacy Servs., Inc., 72 T.C. at 689-90 (stating that the elderly and handicapped were entitled to become "members" of the organization, which entitled them to automatic five percent discounts, while nonmembers were allowed to purchase drugs, but only at prices that were intended to replicate the full retail price for such drugs charged by commercial enterprises).
41.	Id. at 692.
42.	Id. at 691-92.
43.	The IRS did announce, in 2003, its intention to review its position on the possible exempt status of HMOs, and to provide further guidance. IRS Notice 2003-31, 2003-1 C.B. 948. To date, however, no further guidance has been issued.
44.	See IHC Health Plans, Inc. v. Comm'r, 325 F.3d 1188, 1201-03 (10th Cir. 2003) (denying HMOs tax exemption because they did not operate for purpose of promoting health for benefit of community, and because the HMOs were found not to be an integral part of affiliated tax-exempt hospital division); see also Geisinger Health Plan v. Comm'r, 985 F.2d 1210, 1219 (3d Cir. 1993) (holding that HMOs that provided no significant benefits to anyone other than their paying subscribers did not qualify for tax-exempt status).
45.	71 T.C. 158 (1978), acq., 1981-2 C.B. 1, action on dec., 1981-127 (June 10, 1981).
46.	325 F.3d 1188, 1197 (10th Cir. 2003).
47.	Id.
48.	Id.
49.	IHC Group, Inc. v. Comm'r, 82 T.C.M. (CCH) 606, 615 (2001).
50.	IHC Health Plans, Inc., 325 F.3d at 1204.
51.	See Geisinger Health Plan v. Comm'r, 985 F.2d 1210 (3d Cir. 1993) (concluding that the HMO, despite having a subsidized dues plan for low-income members, was not entitled to exemption). A later case involving the same plan, Geisinger Health Plan v. Comm'r (Geisinger II), decided upon remand that the HMO could not qualify for exemption on the basis of its relationship with other entities that were exempt. 30 F.3d 494 (3d Cir. 1994). See also I.R.S. Technical Advice Memorandum 98-37129, in which the IRS considered the status of an HMO and found it not qualified for exemption on similar grounds.
52.	Medicaid Act, 42 U.S.C. §§ 1396-1396v (2000).
53.	Health Insurance for the Aged Act (Medicare Act), Pub. L. No. 89-97, 79 Stat. 290 (codified as amended in scattered sections of 42 U.S.C.).
54.	Havighurst & Richman, supra note 1, at 20-21.
55.	I.R.C. § 501(c)(3) (2005).
56.	I.R.C. § 170 (2005).
57.	I.R.C. § 103(a) (2005) exempts interest paid on certain bonds from inclusion in the gross income of the recipient, as long as the bonds are "qualified bonds." I.R.C. § 141(e)(1)(G) (2005) makes it clear that "qualified 501(c)(3) bonds" are within this exemption. A qualified 501(c)(3) bond is defined in I.R.C. § 145(a) (2005) to include any bond issued by an organization exempted under I.R.C. § 501(c)(3), subject to a few conditions that are not salient here.
58.	An encyclopedic summary of state laws on these questions is beyond the scope of this article, but the following examples from the laws of North Carolina will serve as illustrations: North Carolina allows nonprofit charitable organizations an exemption from the state corporate income tax. N.C. GEN. STAT. ANN. § 105-130.11(a) (West 1998). And, by incorporating the federal definition of income into the North Carolina personal income tax, it effectively allows deductions from state income taxes for charitable contributions to the extent those contributions are deductible for federal income tax purposes. N.C. GEN. STAT. ANN. § 105-134.5(a) (2005).
59.	Again using North Carolina as an illustration, N.C. GEN. STAT. ANN. § 105-164.14(b) (1995) allows exempt organizations to obtain refunds of sales taxes paid. N.C. GEN. STAT. ANN. § 105-275(7) (2003) exempts real and personal property of such organizations from the reach of taxes otherwise applicable to holdings of those properties.
60.	The favorable postal rates for exempt organizations are described at 39 U.S.C.A. §§ 3626. See also I.R.C. § 3306(c)(8) (2005), which exempts charitable organizations from federal employment taxes.
61.	Very infrequently, a hospital or other organization might be organized as a charitable trust, in which case I.R.C. § 1(e) (2005) would be the operative tax provision from which the organization was exempt.
62.	Internal Revenue Serv., Charities and Other Tax-Exempt Organizations, 2002, 25 STAT. INCOME BULL. 263, 264 (Fall 2005), available at http://www.irs.gov/pub/irs-soi/02eochar.pdf. Organizations other than churches must, pursuant to I.R.C. § 6033(a), file an annual return (IRS Form 990), on which this report is based, unless their income is generally less than $25,000. See DEP'T OF THE TREASURY I.R.S., FORM 990: RETURN OF ORGANIZATION EXEMPT FROM INCOME TAX (2005), available at http://www.irs.gov.pub/irs-pdf/f990.pdf. It is thus safe to assume that these data include all nonprofit hospitals.
63.	Internal Revenue Serv., supra note 62, at 264. I.R.C. § 102 allows a broad exclusion for amounts received as gifts. It is possible that some grants, however, might not qualify for this exclusion, so some part of the $41 billion in this category could imaginably be included in income for tax purposes if the organization receiving it were not exempt.
64.	Among the credits for which hospitals might be able to qualify would be the research credit under 26 U.S.C.A. § 41(a), the employer-provided-childcare credit under 26 U.S.C.A. § 45F(a), the work-opportunity credit under I.R.C. § 51(a), and a number of others.
65.	I.R.C. § 172 (2005).
66.	See OFFICE OF MGMT. AND BUDGET, ANALYTICAL PERSPECTIVES: BUDGET OF THE UNITED STATES GOVERNMENT, FISCAL YEAR 2006 , at 316 (2006), available at http://a255.g.akamaitech.net/7/255/2422/07feb20051415/www.gpoaccess.gov/usbudget/fy06/pdf/spec.pdf.
67.	For a full development of this view, see Boris I. Bittker and George K. Rahdert, The Exemption of Nonprofit Organizations from Federal Income Taxation, 85 YALE L.J. 299 (1976). Reduced to its essentials, this article demonstrates that the federal-income-tax principles and rules were simply not designed to tax any excess of receipts over disbursements in any annual period that might be achieved by an exempt organization.
68.	See discussion supra note 13.
69.	Sixty-five cents is the appropriate denominator, rather than one dollar, because the net burden to the donor of making a gross gift of one dollar is that one dollar less the thirty-five cents of tax savings.
70.	I.R.C. § 11 (2005).
71.	Currently, tax rates range from as low as ten percent up to the thirty-five percent maximum rate.
72.	In 2003, 33.7% of all individual taxpayers itemized their returns. Author calculations based on tables 1 and 3, Michael Parisi and Scott Hollenbeck, supra note 6. "However, the decision to itemize is closely linked to income, with 85.2% of taxpayers with adjusted gross incomes exceeding $75,000 being itemizers." Id. Hence, while a minority of taxpayers itemize, it is reasonable to infer that a majority of contributed dollars are deducted by itemizers who obtain some tax benefits from making their contributions.
73.	To make the arithmetic a bit more explicit, assume taxpayer has a $1000 gain and $1000 of unrelated income. The total tax will be as much as $150 on the gain and $350 on the unrelated income, leaving $1500 after tax. If she gives away the $1000 asset instead of selling it, both taxes are avoided, leaving the $1000 of unrelated income intact. This outcome, $1000 of after-tax income, is only $500 worse than the result that would have obtained had no gift been made at all.
74.	CTR. ON PHILANTHROPY AT INDIANA UNIV., GIVING USA 2005: THE ANNUAL REPORT ON PHILANTHROPY FOR THE YEAR 2004, at 123 (2005).
75.	OFFICE OF MGMT. AND BUDGET, supra note 66, at 318. The precise estimates were $3090 million in FY 2004 and $3670 million in FY 2006. Id. Each federal fiscal year begins in October of the preceding numbered year and runs through September of the numbered year. For example, FY 2006 runs from October 1, 2005 through September 30, 2006. The revenue loss is "roughly consistent" with a much larger gross-giving estimate, because the tax loss is no more than the marginal rate of tax times the amount of the gift. See RICHARD SCHMALBECK AND LAWRENCE ZELENAK, FEDERAL INCOME TAXATION, at 356 (2004). Also, some gifts are made by individuals who do not itemize their deductions, by individuals whose contributions are nondeductible because they exceed the limitations of I.R.C. § 170(b) on individual deductions (which generally limit deductions to no more than fifty percent of the donor's adjusted gross income), or by other tax-exempt institutions. Id.
76.	For example, by far the largest gifts to the health subsector in recent years have been those made by the Bill and Melinda Gates Foundation, which have focused on research and treatment efforts targeted at AIDS, malaria, and tuberculosis epidemics, largely in third-world regions. See CTR. ON PHILANTHROPY AT INDIANA UNIV. supra note 74, at 126-27.
77.	See I.R.C. § 145(a)(1) ("[A]ll property which is to be provided by the net proceeds of the issue is to be owned by a 501(c)(3) organization or a governmental unit . . . .").
78.	I.R.C. § 146(a).
79.	I.R.C. § 146(g)(2).
80.	The limit was a fairly generous $150 million of bonded indebtedness outstanding at any one time per institution. See I.R.C. § 145(b) (2000). This provision is still in the Code, but is subject to a built-in expiration of effectiveness by the terms of I.R.C. § 145(b)(5), which was added in 1997. Taxpayer Relief Act of 1997, Pub. L. No. 105-34, 111 Stat. 788 (1997).
81.	I.R.C. § 145(b)(1).
82.	OFFICE OF MGMT. AND BUDGET, supra note 66, at 318 (stating the precise estimate at $2160 million).
83.	Havighurst & Richman, supra note 1, at 12.
84.	OFFICE OF MGMT. AND BUDGET, supra note 66, at 324-26.
85.	Id. at 324.
86.	Havighurst & Richman, supra note 1, at 36-39; Zelenak, supra note 7, at 109-20.
87.	OFFICE OF MGMT. AND BUDGET, supra note 66, at 324.
88.	Id.
89.	Havighurst & Richman, supra note 1, at 23.
90.	See text and notes at supra note 61.
91.	Of course, this possibility would be sporadically constrained by certificate-of-need requirements imposed by various state authorities. See Havighurst & Richman, supra note 1, at 61.
92.	Havighurst and Richman report estimates that $20-25 billion is spent by hospitals on medical education and research. See Havighurst & Richman, supra note 1, at 21 n.36.
93.	Note that scientific research is one of the purposes that may be pursued by organizations exempt from tax under I.R.C. § 501(c)(3). One may object at this point that the subsidy for basic research is coming from the pockets of the insured employees whose premiums pay for medical services set at monopoly prices. See Havighurst & Richman, supra note 1, at 20-21. But those premiums are paid with pre-tax dollars, rather than being paid with post-tax dollars, as most consumption is. It is difficult to be certain about the magnitudes involved on either side of this trade-off, but it is not unreasonable to imagine that the savings from the pre-tax feature offset the lost consumer surplus due to monopoly pricing.
94.	Havighurst & Richman, supra note 1, at 20-25.
95.	Again, certificate-of-need requirements, being based on all facilities in a geographic area, whether for-profit or not, would operate as overall constraints. Thus, overinvestment by nonprofits would tend to constrain investment by for-profit firms.