Agency implications of debt in not-for-profit hospitals: a conceptual framework and overview. (2001 Best Student Paper Award Recipient *).

This article argues that mainstream finance theories about the agency role of debt in for-profit corporations may not apply fully in a not-for-profit hospital context. We outline a set of propositions about the impact of not-for-profit status on capital structure decisions and related agency relationships. By examining not-for-profit hospitals in the context of finance theories on debt, we raise concerns that debt financing may exacerbate or create agency problems, rather than resolve them.

An important premise of this research is that, since the community-at-large "owns" not-for-profit institutions, the community holds a position analogous to that of corporate shareholders. Community ownership is increasingly being codified into law. For example, when not-for-profit hospitals convert to for-profit status, a foundation must be established to recognize the community as the residual owner and repay the community for years of tax exemptions and other support. McLean (1989) supports this analytical framework, and maintains that not-for-profit agency conflicts are actually thornier and more difficult to resolve because "management... is an agent without an effective principal" (p. 68). The community-at-large is diffuse and ill defined as compared with a corporate shareholder.

In addition to the community-at-large, government agencies and donors may also be considered shareholders. The roles of government agencies and major donors may be likened to those of mutual funds and venture capitalists, respectively. Government agencies may be the formal entities that permit tax exemption or provide funding, but they invest to serve the wishes of voters--as mutual fund managers invest to serve the wishes of investors. Donors often provide much of the initial financing for projects, at times in exchange for special relationships (e.g. naming buildings)--just as venture capitalists may receive special classes of stock. Part of the rationale for the not-for-profit form of organization is that "[d]onations can substitute for the resources provided by residual claimants" (Fama & Jensen 1985, p. 115).

Hospital boards are responsible for ensuring that government agencies' conditions of tax-exemption are followed and for ensuring that governments' and donors' contributions are used in accordance with their wishes, since outsiders often have difficulty monitoring the output of the hospital over time (Pauly, 1987). Agency relationships (and potential agency conflicts) between managers and governments and between managers and donors parallel those between managers and the community-at-large, and the mediating role played by hospital boards is similar. Throughout this paper we will generally only refer to communities as principals, but governments and donors may also fit in this classification.

In addressing the manager-community agency relationship, this research aims to build upon a framework of corporate finance in order to explain the role of debt in not-for-profit hospitals and lay the groundwork for future studies of debt's agency effects in not-for-profit hospitals.

PRIOR RESEARCH

Only in the past fifteen years have financial economists begun to apply the tenets of corporate finance to not-for-profit hospitals. Several recent articles have addressed not-for-profit hospitals' capital structures and motivations for issuing debt. For example, Wedig, Sloan, Hassan & Morrisey (1988), Trigeorgis & Brindamour (1993), and Ligon (1997) all found that cost-based reimbursement contributed to the hospital sector's relatively high financial leverage, compared with other industries. Until its phase-out in the 1990s, cost-based reimbursement meant that a portion of hospitals' interest expense could be passed through to payers, especially Medicare. By contrast, the costs of obtaining not-for-profit "equity" (primarily philanthropic donations and retained earnings) have never been directly reimbursed.

Another motivation for hospitals to issue debt is the indirect tax benefit from tax-exempt bonds (Wedig, Hassan, & Morrisey, 1996). Bondholders, who do not owe taxes on the interest from these bonds, receive direct tax benefits, which are partially passed on to the hospitals. Consequently, not-for-profit hospitals can obtain debt financing on favorable terms (lower interest rates) compared with for-profit institutions.

However, hospitals have been somewhat constrained by a project-financing rule specific to municipal finance that requires fixed investment outlays to match or exceed tax-exempt debt flows. (Of course, there are also other IRS-based constraints on arbitrage behavior.) Tax-exempt debt can be used only for a qualifying capital investment project, resulting in a possible bias toward capital-intensive investment. Wedig et al. (1996) hypothesize that hospitals with a shortage of high-yielding, tax-exempt projects may be induced to seek out lower-yielding projects in order to justify increasing their debt financing to target levels, which balance the tax advantages of debt against the potential costs of financial distress. This interaction of financing and investing decisions is not believed to occur in a typical for-profit corporation; indeed, it is contrary to the separation theorem in corporate finance.

To our knowledge, none of the published articles on hospital debt has explicitly considered agency conflicts between managers and community shareholders. However, the articles generally support the hypothesis that debt can cause rather than correct agency problems. Clement, Smith & Wheeler (1988) acknowledge potential agency conflicts between bondholders and community shareholders. Bond covenants, restrictions in insurance or contracting provisions associated with debt, and the need to service debt through periodic interest expenses and principal repayment all limit hospitals' flexibility in providing charitable services desired by the community. Debt-related requirements might even divert hospitals from their charitable missions. Charitable services are typically unprofitable projects and unattractive from a bondholder's perspective.

PROPOSITIONS ABOUT DEBT IN NOT-FOR-PROFIT HOSPITALS

Since not-for-profit hospitals vary along a number of dimensions (not the least of which may be their sponsorship community, religious, educational or other), modeling not-for-profit hospital behavior is multifaceted and complex. Nonetheless, we raise a set of propositions about the agency implications of debt in a not-for-profit hospital. These propositions may not apply to all not-for-profit hospitals, but to a sufficient number of community hospitals to justify further examination. Conditions under which the propositions are likely to hold are developed in two theory sections of the paper which follow presentation of the propositions.

Proposition 1: Debt may crowd out charity care. Sources of financing other than debt (such as government grants or philanthropic contributions) may encourage or require the provision of uncompensated care, or community benefits more generally (which we group together with "charity care"). Debt may crowd out some of these benefits. Given the differences in communities along a host of socio-demographic characteristics, there may be variation in hospitals' commitment to charity care. Debt may force the hospital's managers to adopt a more bottom-line orientation and to focus on profitability, rather than on whatever target they might otherwise have for charity care.

The crowd-out of charity care may occur by two distinct mechanisms: (1) Hospitals and managers that are highly committed to charity care may be constrained in funding it because of restrictive bond covenants and debt-service obligations; (2) Hospitals and managers that are less committed to charity care may use the need to pay down debt as a bargaining chip with payers (public and private) and the community over rates (i.e., charges to paying patients) and the volume of charity care provided. Baldwin (1983) observed that debt-related obligations help management to oppose demands from unions for higher wages. This may be considered a "pseudo crowd-out" effect.

Proposition 2: Debt may crowd out other sources of financing. Apart from the managerial effects of debt, donors may decrease their philanthropic contributions (or their unrestricted donations) to a not-for-profit hospital when it increases its financial leverage, as they may fear that new donations will be used to repay debt rather than to fund charitable services. This hypothesized effect is a direct, albeit new, application of the "debt overhang" theory discussed below.

Proposition 3: Debt may change hospitals' investment priorities such that they emphasize capital investment over less tangible forms of investment. The project-financing rule may be responsible for this phenomenon. However, an observation of "excessive" capital investment (relative to capacity utilization or other factors) is also consistent with theories of empire building and the hypothesis of Newhouse (1970) that not-for-profit hospitals seek to maximize a mixture of quantity and quality.

In the following section, we outline how principal-agency theory applies to the debt financing decisions of corporations. We then discuss the particular application of these theories to the not-for-profit hospital. This discussion concludes with support for the three propositions.

PRINCIPAL-AGENCY THEORY AND DEBT

Standard finance theories hold that increased debt helps to reduce principal-agency problems by aligning managers' incentives with shareholder interests (although it may exacerbate conflicts between managers and existing bondholders, and between managers and other stakeholders). Providing that debt does not rise to the level where financial distress becomes an imminent risk, it can discipline managers to work harder on behalf of shareholders. Managers face incentives to maximize profits, improve efficiency and invest appropriately. There are at least three mutually reinforcing reasons for debt's disciplining effect, as outlined below. All three reasons presume that financial markets are more efficient (faster acting) than product markets or labor markets at correcting sub-optimal managerial behavior.

1. Managers will own a bigger stake in the company: Since debt is often issued in exchange for stock (e.g., through share repurchases in a leveraged recapitalization), some equity previously held by outside investors will be replaced by debt. As a result, managers will become larger shareholders and may be less likely to consume perquisites, shirk their responsibilities, or engage in empire building (Jensen & Meckling, 1976; Barnea, Haugen & Senbet, 1981; Chew, 1993).

2. Managers will seek to avoid financial distress for their companies and the potential loss of their own jobs: Managers have strong incentives to stave off financial distress, which can lead to the takeover of their companies and the loss of their jobs. Because higher debt raises the risk of financial distress occurring and can accelerate the pace of a corporate default, managers will work harder to prevent such distress (Grossman & Hart, 1982; Jensen, 1986; Jensen, 1989; Harris & Raviv, 1990).

3. Debt service will require steady cash flows: The discipline of meeting debt-service obligations focuses attention on cash flows and may necessitate tighter management control of both finances and operations. Debt can be a substitutetpossibly a superior one--for dividends in motivating the pay out of cash flows (Jensen, 1986). If companies that take on higher debt distribute the debt proceeds and other "free cash flows" to shareholders through special dividends or share repurchases, they face even more pressure to generate steady cash flows (Jensen, 1986; Jensen, 1989; Kauer & Silvers, 1991; Chew, 1993). Managers may issue debt in order to signal to shareholders that they expect higher cash flows from new business ventures or capital investments (Ross, 1977). Thus, debt effectively commits managers to generating such cash flows (Grossman & Hart, 1982; Harris & Raviv, 1990).

These three factors, though not completely distinct or separable, form a useful conceptual framework for understanding debt's disciplining role. Note that the finance theories do not just concern debt's effect on managerial incentives; they also concern the rationale for issuing debt in the first place. Yet the disciplining impact is a common thread, regardless of when, how, and why the debt is issued.

Limited empirical support of the disciplining effects of debt can be found in analyses by Jensen (1986 & 1989), Kaplan (1991) and Baker (1992). The agency role of debt in corporations remains an area of active research. Financial economists debate the persuasiveness of the existing evidence and reach different conclusions when reviewing case studies of companies that increased their financial leverage.

Do the above theories and evidence apply in a not-for-profit context or, instead, might higher debt misalign not-for-profit managers' incentives? We address this question by considering agency problems in not-for-profit hospitals and whether the disciplining roles of debt apply. We then return to agency theory in not-for-profit hospitals and the three aforementioned propositions.

AGENCY PROBLEMS IN NOT-FOR-PROFIT HOSPITALS

Principal-agency problems may well be more intractable in a not-for-profit context than in a for-profit one. Consider that the community-at-large is a "shareholder" in all the not-for-profit hospitals serving a neighborhood. By contrast, management and trustees focus on the interests of their individual hospitals. In order to attract doctors, patients, and payers, hospitals have incentives to invest in capacity, technology, and equipment that may be duplicative or excessive from the community's perspective (Pauly & Redisch, 1973; McLean, 1989; Kane & Magnus, 1995). Payers in particular often want to contract with a hospital that can provide the full range of needed services ("one-stop shopping"). This potential misalignment of institutional and shareholder interests simply does not exist in the for-profit world, in which management, trustees, and shareholders alike have a stake in an individual corporation.

Hospitals also lack oversight by public trading markets and may have fewer self-interested or community-interested shareholders. McLean (1989) notes that, without a viable equity market, it becomes much harder to stage a takeover of a not-for-profit hospital whose management is driving down its value. The market for corporate control is much less developed in the not-for-profit market than in the for-profit one. The managerial labor market does serve a protective role, but career moves of executives occur at a slower rate and with higher transaction costs than the exchange of shares of stock.

Another agency problem is the difficulty of giving management an equity position in the hospital. For-profit corporations often compensate managers with stock or stock options--a disciplining device unless managerial entrenchment becomes a factor. Obviously, such forms of compensation are not feasible in a not-for-profit institution. It is illegal for a manager, or any individual, to be a residual claimant on the hospital's profits. Perhaps the closest analogy to equity in the not-for-profit setting is managers' personal identification with the mission of their institutions. Managers' "psychic" equity could increase if they are given more latitude to pursue projects of special interest (Rose-Ackerman, 1987). However, with higher debt in a hospital's capital structure, managers' perceived equity may decrease since they may have less control over their hospital's service output.

Still another challenge is assessing the value of a not-for-profit hospital, either in terms of its social contributions or its profitability. In a for-profit corporation, the stock price serves as an external measure of performance that aggregates the latest public and private information about the corporation. Because there is no stock price or trading volume for a not-for-profit hospital, social learning about the hospital is likely to occur at a slower pace. This is a potential problem since information is a key tool for reducing agency conflicts. Consider the 1998 bankruptcy of the not-for-profit Allegheny Health, Education and Research Foundation (AHERF) in Pennsylvania, which largely took observers and the local community by surprise. Poor public information and oversight have been implicated in the system's collapse (Bums, Cacciamani, Clement & Aquino, 2000).

DO THE THREE DISCIPLINING ROLES OF DEBT APPLY?.

Having identified distinct agency problems in the not-for-profit hospital setting, we now ask if debt can help to resolve these problems.

1. Following debt issuance, do managers own a bigger stake in the hospital? No, they cannot be equity holders at all. In fact, if debt covenants are restrictive and creditors' oversight is tight, managers may feel that their stake in the hospital has diminished.

2. Will managers work harder when financial leverage increases? Maybe, maybe not. With higher debt in a hospital's capital structure, managers may make a greater effort to avoid financial distress and potential job loss.

3. Does debt service require steady cash flows? Yes. However, the need to repay debt may create an unusual kind of "slack," peculiar to the not-for-profit context, because it may increase managers' latitude to scale back community services that they do not deem glamorous or prestigious. Similarly, it may improve their negotiating position with regulators and insurance companies over rates or charges (Ligon, 1997).

Reexamining these three points raises questions about the applicability of the standard finance theories to not-for-profit hospitals. Below, we sketch out an alternative theory that suggests that debt may exacerbate the agency problems unique to not-for-profit hospitals.

AGENCY THEORY IN NOT-FOR-PROFIT HOSPITALS

The concept of free cash flow, a special application of debt overhang and investment under the project-financing role form the basis of interpreting agency theory in not-for-profit hospitals. The role of free cash flow relates to proposition 1, that debt may crowd out charity care. The role of debt overhang and the project-financing rule relate to proposition 2, that debt may crowd out other sources of financing. Investment under the project-financing rule also relates to proposition 3, that debt may change hospitals' investment priorities.

The role of free cash flow in a not-for-profit hospital is likely very different from that in a for-profit company. In a for-profit setting, shareholders want free cash flow kept to a minimum in order to prevent management from wasting cash on perks, low-yielding investments, and empire building. A typical strategy is to dispense free cash flow in the form of regular or special dividends or through share repurchases. Forcing management to appeal to external capital sources subjects them to market discipline, as they must justify new projects to disinterested parties. Only when a company is constrained in revealing information about new investment projects--for example, when competitive positioning requires that the information be kept proprietary--is free cash flow deemed advantageous.

In the not-for-profit hospital, the community wants managers to have sufficient cash to fund charitable services--enough cash to "do good," if not to "do well." A substantial amount of unrestricted, unencumbered cash may be necessary because of not-for-profit hospitals' limited ability to raise capital from donors (Rose-Ackerman, 1987), other equity sources (e.g., retained earnings or government agencies), and debt markets. A key problem facing not-for-profit hospitals is that the same cash used to fund charity care is potentially wasted on perks and empire building. Ensuring that not-for-profit managers have the "fight" amount of cash at their disposal is a difficult balancing act.

In the for-profit context, debt plays the unambiguous role of reducing free cash flow, as cash must be generated and used to service debt. In the not-for-profit context, however, debt's role is less clear (Kauer & Silvers, 1991). Debt in a not-for-profit hospital decreases the cash flow available for charity care and management perks, but increases that available for commercial purposes and capital investment (because of the project-financing rule). Through these cash-flow effects, debt may misalign managers' incentives.

To shed further light on the agency implications of debt in the not-for-profit context, we turn to the question of whether "debt overhang" is a problem for not-for-profit institutions and a constraint on not-for-profit managers. The standard debt overhang theory (Myers, 1977) holds that increased debt can prevent profitable projects from being funded. Shareholders, who hold only a residual claim on the cash flows generated by the firm, will not fund investments that primarily benefit bondholders, either through direct payments to bondholders or a reduced risk for outstanding debt. Debt can motivate under-investment by weakening shareholder incentives to commit fresh equity capital.

Does debt overhang apply in a not-for-profit context? It may be mitigated since the project-financing rule permits not-for-profit hospitals to impute indirect tax benefits to the marginal project. These benefits lower the effective cost of the project and hence the requisite return. In fact, the cost may decline to such a degree that a not-for-profit organization might undertake a negative net present value project (Kauer & Silvers, 1991). This would occur only when the indirect benefits of the project exceed the interest and principal-repayment costs, as well as such indirect costs of debt as potential financial distress.

Regardless of the hospital's target debt level and its current debt position, debt overhang may apply in another, unique way in the not-for-profit context. Debt financing could crowd out philanthropic contributions and hence charitable output that the contributions would fund. Donors may be concerned that, with the need to pay down debt, a hospital will use new philanthropic contributions to service its debt. Donors supply "equity" only when they expect to reap a fair share of the rewards, such as witnessing an expansion of community benefits. They may not see such benefits, however, if the hospital is saddled with debt obligations and if the funds flowing into the hospital are fungible (such that contributions can effectively be reassigned to servicing debt).

Both contractual obligations connected with debt and the need to maintain ongoing access to debt markets are likely to force hospitals to pay off debt before funding charity. While state and federal regulation sets minimum standards for the provision of charity care, hospitals enjoy wide discretion in determining their charitable output. Donors may require that their funds be used for specific purposes, but hospitals may reduce spending from other sources of funds, or reduce spending on other forms of charity care that do not interest a particular donor. Further, recall that tax-exempt debt cannot be issued directly for the purpose of funding charity care, and that bondholders usually do not look favorably upon an increased provision of charity care. Hence, debt overhang could prevent or discourage not-for-profit hospitals from investing in new or expanded forms of community benefits.

Debt financing may actively contribute to not-for-profit hospital over-investment. The project-financing rule requires that the dollar value of capital investments must be greater than or equal to the tax-exempt debt proceeds. When this rule binds, debt capacity is high relative to investment opportunities, hospitals aim to increase the debt level and they may seek out lower-yielding investments than they would otherwise (Wedig et al. 1996). As noted earlier, financing and investing decisions interact in this scenario.

From a shareholder perspective, the deleterious consequences of hospital overinvestment include: negative returns from projects, at least at an aggregate industry level (if not for a particular project in a particular hospital); induced demand for unnecessary medical services to justify available projects; and a tradeoff or substitution effect, such that increased capital investment could decrease the funds available for charity care. The potential contribution of debt financing to hospital over-capacity is a serious policy concern.

CONCLUSIONS

Mainstream principal-agency theories in finance hold that increased debt in for-profit corporations helps to resolve agency conflicts between managers and shareholders. However, increased debt may not discipline managers as effectively in not-for-profit hospitals. It may even misalign not-for-profit hospital managers' incentives with respect to the wishes of shareholders, who include the community-at-large, donors and governments.

In the case of not-for-profit hospitals, there are at least three relations that are of interest: (1) debt and the provision of uncompensated care, or community benefits, (2) debt and the receipt of unrestricted donations, (3) debt and project returns. Mainstream finance theories would suggest positive relations in all three cases. The propositions presented in this paper suggest the possibility of observing negative relations.

If empirical research finds negative relations, the ramifications for both management and policy are potentially far-reaching. There may, for example, be a need for increased governmental oversight of tax-exempt debt markets and not-for-profit hospital performance. Most policy debates regarding for-profit vs. not-for-profit hospital performance have centered on product markets, which encompass both commercial goods and "social goods," such as charitable services. However, financial-market differences between for-profit and not-for-profit hospitals could be more pronounced than product-market differences, and could even influence behavior in product markets, as with the debt-charity relation of proposition 1.

The key implication of this analysis is that it is much harder in a not-for-profit hospital to set debt levels in order to minimize agency conflicts between managers and shareholders. The presence of debt itself may also raise agency problems that do not exist, and could not even be imagined, in a for-profit corporation. Empirical work is needed in order to test the propositions and theoretical insights offered in this paper.

REFERENCES

Baker, G.P. (1992, July). Beatrice: a study in the creation and destruction of value. Journal of Finance, XLVII(3), 1081-1119.

Baldwin, C.Y. (1983) Productivity and labor unions: an application of the theory of self-enforcing contracts. Journal of Business, 56(2), 155-185.

Barnea, A., Haugen, R.A., & Senbet, L.W. (1981, Summer). Market imperfections, agency problems, and capital structure: a review. Financial Management, X(3), 7-22.

Burns, L.R., Cacciamani, J., Clement, J., & Aquino, W. (2000, January/February). The fall of the house of AHERF: the Allegheny bankruptcy. Health Affairs, 19(1), 7-41.

Chew, D.H., Jr. (1993). Introduction: financial innovation in the 1980s. In D.H. Chew, Jr. (Ed.) The New Corporate Finance: Where Theory Meets Practice. New York, NY: McGraw-Hill, ix-xix.

Clement, J.P., Smith, D.G., & Wheeler, J.R.C. (1988, Fall). Relations between financing and output in the not-for-profit hospital. Medical Care Review, 45(2), 255-289.

Fama, E.F., & Jensen, M.C. (1985). Organizational forms and investment decisions. Journal of Financial Economics, 14, 101-119.

Grossman, S.J., & Hart, O.D. (1982). Corporate financial structure and managerial incentives. In J.J. McCall (Ed.) The Economics of Information and Uncertainty, Chicago, IL: University of Chicago Press, 107-140.

Harris, M., & Raviv, A. (1990, June). Capital structure and the informational role of debt. Journal of Finance, XLV(2), 321-349.

Jensen, M.C. (1986, May). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review, 76(2), 323-329.

Jensen, M.C. (1989, September/October). Eclipse of the public corporation. Harvard Business Review, 67(5), 61-74.

Jensen, M.C., & Meckling, W.H. (1976, October). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305-360.

Kane, N.M., & Magnus, S.A. (1995). U.S. corporate executives play new roles in managing hospital costs. Business & the Contemporary World, VII(I), 77-82.

Kaplan, S.N. (1991). The staying power of leveraged buyouts. Journal of Financial Economics, Vol. 29, 287-313.

Kauer, R.T., & Silvers, J.B. (1991, Fail). Hospital free cash flow. Health Care Management Review, 16(4),67-78.

Ligon, J.A. (1997, Spring). The capital structure of hospitals and reimbursement policy. Quarterly Review of Economics and Finance, 37(1),59-77.

McLean, R.A. (1989, Winter). Agency costs and complex contracts in health care organizations. Health Care Management Review, 14(1), 65- 71.

Myers, S.C. (1977, November). Determinants of corporate borrowing. Journal of Financial Economics, 5(2), 147-175.

Newhouse, J.P. (1970, March). Toward a theory of nonprofit institutions: an economic model of a hospital. American Economic Review, 60(1), 64-74.

Pauly, M.V. (1987, May). Nonprofit firms in medical markets. American Economic Association Papers and Proceedings, 77(2), 257-262.

Pauly, M.V., & Redisch, M. (1973, March). The not-for-profit hospital as a physician's cooperative. American Economic Review, 63(1), 87-99.

Rose-Ackerman, S. (1987). Ideals versus dollars: donors, charity managers, and government grants. Journal of Political Economy, 95(4), 810-823.

Ross, S.A. (1977, Spring). The determination of financial structure: the incentive-signaling approach. Bell Journal of Economics, 8(1), 23-40.

Trigeorgis, L., & Brindamour, P. (1993, August). Distortions in capital asset acquisition and financing under cost-based reimbursement. The Financial Review, 28(3), 417-429.

Wedig, G.J., Hassan, M., & Morrisey, M.A. (1996, September). Tax-exempt debt and the capital structure of nonprofit organizations: an application to hospitals. Journal of Finance, LI(4), 1247-1283.

Wedig, G.J., Sloan, F.A., Hassan, M., & Morrisey, M.A. (1988, March). Capital structure, ownership, and capital payment policy: the case of hospitals. Journal of Finance, XLIII(1), 21-40.

* Stephen A. Magnus
University of Kansas (USA)

Dean G. Smith
University of Michigan Ann Arbor (USA)

John R.C. Wheeler
University of Michigan Ann Arbor (USA)

The authors gratefully acknowledge research support from the Blue Cross Blue Shield of Michigan Foundation (Detroit, MI), the Nonprofit Sector Research Fund of the Aspen Institute (Washington, DC), and Merritt Research Services, LLC (Cedar Rapids, IA). Finally, the authors thank two anonymous reviewers and the editors of the journal for their helpful insights that improved the paper.

Address for correspondence: Stephen A. Magnus, University of Kansas School of Medicine, Department of Health Policy and Management, 3901 Rainbow Blvd., 4040A Varnes Center, Kansas City, KS 66160 USA, [e-mail].