Statement of Jill R. Horwitz, Ph.D., Assistant Professor, University of Michigan Law School, Ann Arbor, Michigan; and Faculty Research Fellow, National Bureau of Economic Research, Cambridge, Massachusetts

Testimony Before the Full Committee
of the House Committee on Ways and Means

May 26, 2005

MR. CHAIRMAN AND MEMBERS OF THE COMMITTEE, thank you for the opportunity to speak with you today.  My name is Jill Horwitz.  I am an Assistant Professor of Law at the University of Michigan and a Faculty Research Fellow at the National Bureau of Economic Research, although the opinions I offer today are my own.

Mr. Chairman, in its review of the tax-exempt sector, this Committee has heard many distinguished witnesses discuss the legal requirements governing nonprofit organizations, the advantages that come with nonprofit status, and whether nonprofit organizations provide sufficient public benefits to justify these advantages.  These are particularly important questions for the hospital industry, where for-profit, nonprofit, and government hospitals operate side by side.

In my testimony, I will discuss two questions about the implications of the mix of hospital types:   First, do different types of hospital act differently?  Second, are there significant competitive issues raised by having different hospital types competing in the same market together?

Medical Service Provision

Underlying many of the policy questions about the legal treatment of nonprofit hospitals is one basic issue: do they act the same as for-profit hospitals – and if not, what are the differences and are they big enough to matter?

There are good reasons to expect hospitals of different ownership status to act alike.  They all share common goals of treating sick people; they all employ large numbers of doctors and nurses, using medical technology; they contract with the same employers and insurance companies, and are subject to the same health care regulations.  Superficially, they resemble each other so much that a patient admitted to a hospital is unlikely to be able to tell whether it is a for-profit or a nonprofit.

However, whether you find differences between nonprofit and for-profit hospitals depends on where you look.  Most studies of hospital ownership have examined financial measures, and have found little difference among hospital types.[1]  For example, research has shown that nonprofit and for-profit hospitals are quite similar in their costs,[2] sources of capital,[3] exercise of market power,[4] and adoption of certain types of technology.[5]  Although for-profit hospitals pay higher wages and offer incentives to top managers, nonprofits are increasingly using performance-based pay as well.[6]  Finally, during the early 1990s for-profit hospitals and nonprofits had similar margins, although for-profit margins were higher than those of nonprofits by the late 1990s.[7]  There is some evidence that in the most recent years the average nonprofit hospital had a negative income per admission, while the average for-profit had a positive income per admission.

Such financial measures, however, provide an incomplete picture of a hospital.  Because they are first and foremost providers of care for the sick and injured, to evaluate whether nonprofit hospitals earn their keep we must also know how hospitals differ in the medical care they provide.

In my research on medical services, I have found large, systematic, and long-standing differences among hospital types.  For-profit hospitals are more likely than their nonprofit counterparts to offer the most profitable services, and less likely than either nonprofits or government hospitals to offer services that are unprofitable yet valuable, even essential.

I will offer a few examples.  Psychiatric emergency care is considered an extremely unprofitable service, both because of low reimbursements and because its patients tend to be poor and uninsured.  Comparing hospitals that are similar in terms of size, teaching status, location, and market characteristics, for-profit hospitals were 7 percentage points less likely than nonprofits and 15 percentage points less likely than government hospitals to offer psychiatric emergency services.

Probability of Offering Psychiatric Emergency Services

SOURCE:  Jill Horwitz, “Making Profits and Providing Care:  Comparing Nonprofit, For-Profit, and Government Hospitals,” Health Affairs, v.23, n.3 (2005): 790-801.

NOTES:  Controlling for size, teaching status, location, and market characteristics.

Compare these results to open heart surgery, a service so profitable that is often referred to as the hospital’s “revenue center.”  For-profit hospitals are over 7 percentage points more likely than similar nonprofit hospitals and 13 percentage points more likely than government hospitals to provide open-heart surgery.

Probability of Offering Open Heart Surgery

SOURCE:  Jill Horwitz, “Making Profits and Providing Care:  Comparing Nonprofit, For-Profit, and Government Hospitals,” Health Affairs, v.23, n.3 (2005): 790-801.

NOTES:  Controlling for size, teaching status, location, and market characteristics.

Perhaps what is most striking about for-profit hospitals is how strongly and quickly they respond to changes in financial incentives.  The best illustration of this comes from a set of post-acute care services, such as home health-care and skilled nursing services, whose profitability changed sharply over time.  These services became highly profitable in the early 1990s, then reversed and became less profitable with the 1997 Balanced Budget Act.  All three types of hospitals increased their offerings of home health care when it became profitable, but for-profits did so to a striking degree.  From 1988 to 1996, the probability of a for-profit hospital offering home health services more than tripled – from 17.5 percent to 60.9 percent.  During the same period, nonprofit and government hospitals increased their investment at a much lower rate (nonprofits went from 40.9 to 51.7 percent, government hospitals went from 38.1 to 51.9 percent).  When these services became unprofitable, for-profits were also quick to exit the market, roughly 5 times quicker than nonprofits.  This finding provides evidence that for-profits move quickly and strongly in response to financial incentives.

Probability of Offering Home Health Service

SOURCE:  Jill Horwitz, “Making Profits and Providing Care:  Comparing Nonprofit, For-Profit, and Government Hospitals,” Health Affairs, v.23, n.3 (2005): 790-801.

NOTES:  Controlling for size, teaching status, location, and market characteristics.

In sum, for-profit and nonprofit hospitals act quite differently.  For-profit hospitals are considerably more responsive to financial incentives than nonprofits, not just with respect to their decisions to offer services but also in their willingness to operate at all.  Under financial pressure, for-profit hospitals are more likely to close or restructure than nonprofits.[8]

The most important aspect of these findings is that nonprofits are more willing than for-profits to offer services even though they happen to be unprofitable.  These services include not just psychiatric emergency care, but also child and adolescent psychiatric care, AIDS treatment, alcohol and drug treatment, emergency rooms, trauma services, and obstetric care.

There are a few clear implications of these findings for the question of whether nonprofits provide valuable benefits to society.  First, if the mix of medical services available in a community is strongly determined by the profitability of the services, this is potentially worrisome for all patients -- rich and poor, insured and uninsured.  Patients need what they need, depending on their medical condition not on the price of a service.  Even rich and insured patients sometimes need services that it are unprofitable for hospitals to offer. 

As I noted above, nonprofits are more likely to offer a trauma center than for-profit hospitals with similar characteristics.  One hopes never to be in a serious car crash.  But survivors are more likely close to a trauma center if the accident takes place just outside a nonprofit hospital.

Second, extreme responsiveness to financial incentives can be quite costly to the government.  Medicare spending per patient and increases in spending rates are higher in for-profit hospital markets than others.[9]  This can be explained by investments such as home health.  For example, during that period of ramped up provision of home health care services, home health visits per Medicare beneficiary increased by nearly a factor of seven, and payments for those services ballooned.  Government spending on post-acute care went from 3 percent of Medicare hospital payments to 26 percent.[10]  This increase was not patients getting better care, but hospitals double-dipping – receiving two reimbursements for the same treatment.

Perhaps more troubling is evidence that the relative responsiveness to financial incentives has led to fraudulent billing through a practice known as “up-coding.”  Up-coding occurs when a hospital shifts a patient’s diagnosis to one that receives higher reimbursement from Medicare.  For example, a hospital may label a case of pneumonia as a case of pneumonia with complications, at increased cost to the government of about $2,000 per discharge.  Although all types of hospitals have done this, for-profit hospitals have done this more than nonprofit hospitals.[11]  Moreover, up-coding is contagious.  Nonprofit hospitals are more likely to up-code when they have for-profit hospital neighbors than when they do not.

As a final point on differences in hospital behavior, let me say a word about charity care.  Over the past fifty years, the legal requirements for nonprofit hospitals seeking tax exemption have increasingly shifted from narrow requirements that hospitals relieve poverty to broader demonstrations of charitable benefit.  Yet, public attention to the provision of what is called “charitable care” has remained robust.  Whether nonprofit and for-profit hospitals differ in their provision of charity care is difficult to say – in large part because what is typically measured is overall uncompensated care.  Uncompensated care provided by hospitals represents items that most of us would not consider charitable.  These include bills left unpaid by patients who have the ability to pay or discounts to insurance companies.  Given these measurement difficulties, credible evidence shows that hospital types do not differ much in the provision of uncompensated care.[12]  Even these results are hard to interpret because for-profit hospitals locate in relatively better-insured areas.[13]  My main point in discussing charity care is that although free care for those who are unable to afford it is important, other differences – in services, in quality, in medical innovation – are valuable to all members of society.

Hospital Competition

Do nonprofit hospitals have anti-competitive effects, or represent unfair competition to for-profits?  The arguments about competition boil down to the idea that the nonprofit tax exemption is either unfair or distortionary.  An older generation of research claimed, for example, that the tax exemption gives nonprofits an extra financial boost that makes it difficult for for-profits to compete.  Newer research has dismissed this notion by demonstrating that income tax exemptions do not lower input prices.  Furthermore, as an empirical matter, if there were anti-competitive effects we would not see mixed markets with both for-profit and nonprofit hospitals, but we do.

Some argue that nonprofits are less efficient than for-profits and are able to stay in business because they use their surpluses, including tax savings, to offset higher production costs.  This idea, too, has little foundation.  In determining whether an organization is efficient, it is centrally important to answer the question “efficient at what?”  For-profits are more efficient at earning profits.  In the hospital sector, we care about efficiency in providing health care.  Overall, empirical evidence shows no appreciable differences in efficiency at providing health care between for-profit and nonprofit hospitals.

A final idea is that tax savings leads nonprofits to produce too many goods of too little value.  That is, nonprofits use their financial savings to lower costs and, therefore, patients will buy too much health care.  This argument implies that the health care provided by nonprofit hospitals is too cheap.  The idea that health care is too inexpensive is generally not of great concern, particularly when annual medical inflation rates are back on the rise at 4 percent per year.

The best evidence shows that nonprofit hospitals, rather than using their financial savings to offset inefficient management or lower prices to drive for-profit competitors out of business, provide unprofitable and essential services that are valuable to society.  These come not only in the form of more valuable medical services like trauma care, but also in training physicians and nurses.  It is the vigorous competition among nonprofit hospitals that has produced virtually all the medical innovations on which we rely.  Imagine where we would be without the first small pox vaccination developed at the nonprofit HarvardMedicalSchool or the first brain surgery at Johns Hopkins.  We can thank nonprofits for robotic surgery, pacemakers, artificial skin, kidney transplants, and new technology to save premature infants.  Finally, along with the competition among nonprofit hospitals, having for-profits in the mix provides another dimension of competition, competition between organizational types.

An important lesson of the research I have summarized today is that what you find depends on where you look.  If you look at financial behavior, you will find few differences that justify tax exemption.  If you look at medical treatment, you will find some striking differences of the sort that need to be included in any thorough discussion of nonprofit benefits. 

Thank you for the opportunity to testify today.