ABA Health eSource
April 2009 Volume 5 Number 8

The IRS Final Hospital Report
By Laura Gabrysch, Fulbright & Jaworski L.L.P., San Antonio, TX 1

AuthorThe Internal Revenue Service (the “ IRS”) in February 2009 issued the Final Report on its Hospital Compliance Project (the “ Report”). 2 This Report has been a long-time coming. The IRS commenced the hospital compliance project in May 2006 when it sent a questionnaire to 544 exempt hospitals asking questions about community benefit and executive compensation.

The IRS had initiated the project because it wanted to gather information about how hospitals were fulfilling the community benefit standard and setting compensation levels. The community benefit standard is the legal standard for determining whether a hospital is exempt from taxation under Section 501(c)(3) of the Internal Revenue Code (the “ Code”). Because the community benefit standard is flexible, does not require a specific amount of charity care, and is based on facts and circumstances, the IRS was concerned that the standard did not distinguish exempt hospitals from for-profit hospitals and resulted in inconsistent reporting. With respect to executive compensation, Code Section 501(c)(3) prohibits exempt organizations from paying more than reasonable compensation and Code Section 4958 establishes procedures for a rebuttable presumption of reasonableness 3 that organizations can use to obtain a presumption that the compensation paid is reasonable. Nevertheless, the IRS was concerned that hospital executive compensation was escalating, in part because of the use of for-profit comparables to support compensation levels.

After reviewing the responses from the questionnaire, the IRS issued an Interim Report in July 2007. The Interim Report presented data from responses to the community benefit questions, but presented no analysis of these responses. The Interim Report also did not address the responses to the questions dealing with executive compensation. Among other things, the IRS in the Interim Report indicated that it wanted to examine the data in the context of varying demographics, such as whether the hospitals served urban or rural communities. 4


As promised in the Interim Report, the Report groups hospitals into demographic categories based on community type and revenue size. The hospitals were also categorized based on health insurance coverage and per capita income in the surrounding communities.

The four community types are:

  1. High population hospitals from the 26 largest urban areas (“ HPHs”), which accounted for 19 percent of respondents;
  2. Other urban and suburban hospitals (“ USHs”), which accounted for 51 percent of respondents;
  3. Critical access hospitals (“ CAHs”), which accounted for 14 percent of respondents; and
  4. Other rural hospitals (“ ORHs”), which accounted for 16 percent of respondents.

Additionally, the Report groups the hospitals according to five categories, based on annual revenues:

  1. Hospitals with revenues of under $25 million, which accounted for 17 percent of respondents;
  2. Hospitals with revenues between $25 million to $100 million, which accounted for 36 percent of respondents;
  3. Hospitals with revenues between $100 million to $250 million, which accounted for 27 percent of respondents;
  4. Hospitals with revenues between $250 million to $500 million, which accounted for 13 percent of respondents; and
  5. Hospitals with revenues over $500 million, which accounted for 7 percent of respondents. 5


There are several limitations on the use of the data reflected in the Report, many of which the IRS acknowledges.

First, it is not certain whether the sample of 500 or so hospitals is representative of the broader tax-exempt hospital sector. 6 Just a few anomalies can skew the results for such a small sample. For example, the Report describes how the presence of approximately 15 research hospitals skewed the community benefit data for the group as a whole. 7 The same can be said for children’s hospitals, which generally accept no Medicare patients but a disproportionate share of Medicaid patients.

Second, the questionnaire did not ask questions in ways that always elicited consistent and complete responses. The questionnaire was drafted prior to the development of the revised Form 990, which was an involved process that resulted in Schedule H. In 2007, the IRS undertook the redesign process for the Form 990 and instructions, which had not been significantly revised since 1979. The IRS was concerned that the Form 990 did not provide for adequate reporting of the community benefit activities of hospitals. After taking into account approximately 700 comments, the IRS released the revised form in December 2007. After taking into account about 120 more comments, the IRS released the revised instructions in August 2008. The changes to the Form 990 and instructions were extensive and included sixteen schedules, including Schedule H, which requires extensive reporting of a hospital’s charity care and community benefit activities, as well as information on bad debt, community building activities, collection practices, joint ventures, and facilities. 8 Unlike Schedule H, which requests information in a manner and format that minimizes inconsistent reporting, the 2006 questionnaire left it to hospitals to determine and measure community benefit and uncompensated care. For example, the questionnaire did not define “uncompensated care,” and hospitals may have reported various items of shortfall and bad debt as uncompensated care, but not charity care. Additionally, the questionnaire asked whether hospitals included various shortfalls measured by charges, but not costs, which may have resulted in underreporting for those hospitals which measure shortfalls by costs.

Third, the reported data generally was not audited or verified. The one exception was the compensation data of the 20 hospitals which were chosen for examination. 9

Fourth, the Report’s demographic groupings based on community type and revenue size, as discussed below, might not be precise. 10 This is the IRS’s first attempt to categorize hospitals in such a manner, and further refinement may be necessary.

Fifth, the Report looks at only a single tax year—2005—which may not be representative of other years or of the economic cycles. 11 Indeed, the recent economic downturn may elicit responses that present a significantly different picture than that presented in the Report.

Sixth, the Report does not address certain community benefits, such as subsidized health services.

Finally, the disclosure requirements made it difficult to provide specifics in some cases. 12

Perhaps the most frequent criticism of the Report is the second one described above—that the questionnaire yielded inaccurate information. In fact, one commentator called the entire Part V of the Report “completely worthless,” opining that “[w]hile the distributions and breakdowns all look very impressive, the simple fact is that all the charts and numbers are meaningless because the numbers that are reported by each hospital vary in their content and methodology.” 13 Perhaps this criticism is overly harsh. After all, the IRS can only work with the information that is available. Nevertheless, an IRS official has acknowledged this issue, and has publicly commented that persons reading the Report should focus on the on the patterns and general findings, rather than specific numbers. 14 In that spirit, the discussion below on the Report findings will mercifully omit many of the specifics and concentrate on the big picture.


What Was Reported as Community Benefit?

The IRS commented that there was a wide range of reporting with respect to community benefit across community types and revenue sizes. Even so, by far uncompensated care was the largest portion of community benefit. The percentage mix for community benefit expenditures was 56 percent uncompensated care, 23 percent education, 15 percent research, and 6 percent community programs. 15

However, the impact of a group of 15 hospitals, which reported 93 percent of the research expenditures, illustrates the degree by which a few hospitals can skew the results. If these research hospitals were excluded, the percentage mix would be 71 percent for uncompensated care, 21 percent for education, 1 percent for research, and 7 percent for community programs. 16

Community Benefit Findings

The average community benefit expenditures were 9 percent of total revenues, while the median community benefit expenditures were 6 percent of total revenues. 17 Not surprisingly, HPHs had higher average and median community benefit expenditures than CAHs and ORHs, with the average expenditures ranging from 6.3 percent for CAHs to 12.7 percent for HPHs. 18

The IRS observed that a small number of hospitals provided the bulk of the community benefit, with 9 percent of the hospitals reporting 60 percent of the community benefit expenditures and 19 percent of hospitals reporting 78 percent of the community benefit expenditures. 19 Additionally, 21 percent of hospitals reported community benefit expenditures of less than 2 percent of revenues and 47 percent reported community benefit expenditures of less than 5 percent of revenues. 20 CAHs and the smallest hospitals were more likely to have expenditures that fell below the 5-percent and 2-percent levels. Among CAHs, 39 percent reported community benefit expenditures of under 2 percent of total revenues and 61 percent reported community benefit expenditures of under 5 percent. ORHs and hospitals with revenues under $25 million reported numbers similar to CAHs. On the other hand, only 11 percent of HPHs, 17 percent of USHs, and 5 percent of hospitals with revenues of over $250 million reported community benefit expenditures of less than 2 percent, and 32 percent of HPHs, 46 percent of USHs, and 27 percent of hospitals with revenues of over $250 million reported community benefit expenditures of less than 5 percent. 21

Excess Revenues

The IRS also measured excess revenues to determine their impact on the ability to provide community benefits. While 79 percent of all hospitals reported excess revenues, 21 percent of the hospitals reported no excess revenues or a deficit. 22 Of the hospitals reporting excess revenues, 39 percent reported excess revenues as a percentage of total revenues as 5 percent or less and 40 percent reported excess revenues as percentage of total revenues of over 5 percent. 23 Of the hospitals reporting no excess revenues or a deficit, most of these were hospitals with revenues of less than $25 million (35 percent), and the percentage of hospitals reporting deficits or no excess revenues decreased as size increased (10 percent for hospitals with revenues of over $250,000). 24 Additionally, 34 percent of CAHs reported no excess revenues or deficits, compared to 13 percent for ORHs. 25

On average, hospitals reported a percentage of excess revenues over total revenues of 4.6 percent. 26 This percentage was 3.3 percent for the smallest hospitals and increased to 5.5 percent for hospitals with revenues of over $500 million. 27 Among community types, CAHs reported the smallest percentage of excess revenues (3.5 percent) while ORHs reported the largest percentage (6 percent). 28 HPHs and USHs reported excess revenues of 4.5 percent and 4.6 percent, respectively, to total revenues. 29 The Report also concluded that 77 percent of the aggregate excess revenue was generated by only 16 percent of the hospitals. 30

Other Demographics

All community types reported private insurance as the largest portion of patient mix, followed by Medicare, Medicaid, uninsured, and public programs, in that order. The average percentages were 43 percent private insurance, 31 percent Medicare, 15 percent Medicaid, 3 percent public insurance, and 8 percent uninsured. 31 All community types reported uninsured patients as 7 percent to 8 percent of total patients. 32 CAHs had the highest percentage of Medicare patients, while HPHs had the highest percentages of Medicaid and public insurance patients. 33

The Report found no correlation between community benefit expenditures levels and per capita income of the community served by the hospitals. 34 However, the Report did find that community benefit expenditures increased as uninsured rates in the community increased. 35

What was reported as Uncompensated Care?

As with community benefit, there was a wide variation in how uncompensated care was reported. The numbers reported in this section are likely inaccurate. The 2006 questionnaire asked whether hospitals treated as uncompensated care the amounts “charged for services” less amounts received for private insurance, public insurance, self-pay, Medicaid, and Medicare. The questionnaire did not ask whether hospitals included shortfalls measured by costs as uncompensated care. Thus, it is possible that some hospitals may have answered “no” to these questions, resulting in these shortfalls not being captured in the Report.

Between 18 percent and 20 percent of hospitals reported including private insurance shortfalls, Medicare shortfalls, Medicaid shortfalls, and public program shortfalls in uncompensated care—low percentages, especially for Medicaid shortfalls, which may result from the wording of the questionnaire. 36 The number of hospitals reporting bad debt as uncompensated care was 51 percent and the number of hospitals reporting self-pay shortfalls as uncompensated care was 44 percent. 37 Generally, all hospitals reported higher percentages of self-pay shortfalls and bad debt as uncompensated care as compared to other shortfalls. 38 Rural hospitals (CAHs and ORHs) reported less bad debt as uncompensated care, but include more private insurance, Medicare, Medicaid, and public insurance shortfalls. 39 By revenue size, hospitals with revenues of over $250 million reported bad debt as their largest item of uncompensated care, whereas the smaller hospitals generally reported higher levels of private insurance, Medicare, Medicaid, and public insurance shortfalls. 40

To avoid identification, hospitals with revenues of over $500 million were grouped with hospitals with revenues from $250 million to $500 million. Even though no specifics were reported, the Report does indicate that hospitals with revenues ranging from $250 million to $500 million reported the highest percentage of Medicaid shortfalls, whereas hospitals with revenues of over $500 million reported the smallest percentage of Medicaid shortfalls. Further, hospitals with revenues of over $500 million reported the smallest percentage of Medicare shortfalls, public insurance shortfalls, and private pay shortfalls than any other category. 41

The Report also compared uncompensated care expenditures for hospitals that included the various items of shortfall or bad debt as uncompensated care as opposed to those that did not. On average, there was very little difference in revenues reported as uncompensated care with respect to private insurance shortfalls, public insurance shortfalls, and self-pay shortfalls. However, the average percentage of revenue spent on uncompensated care was significantly higher for hospitals that reported including Medicare shortfalls (11.6 percent v. 5.6 percent), Medicaid shortfalls (12.6 percent v. 5.6 percent), and bad debt (9.7 percent v. 4.8 percent). 42

Uncompensated Care Findings

The median percentage of patients receiving uncompensated care was 3.4 percent and the average was 9.8 percent. 43 By community type, the medians ranged from 1.7 percent for ORHs to 6 percent for HPHs and the averages ranged from 6.7 percent for CAHs to 11.2 percent for HPHs. 44 In all cases, the median percentages were less than the percentage of uninsured patients in the community, which ranged from 11 percent for HPHs to 8 percent for ORHs. 45 Uncompensated care as a percentage of community benefit expenditures was greatest for CAHs (77 percent), ORHs (76 percent), and hospitals with revenues of less than $25 million (93 percent), as compared to HPHs (42 percent), USHs (69 percent), and hospitals with revenues over $500 million (35 percent). 46

Between 94 percent and 96 percent of each community type reported uncompensated care expenditures. 47 The average uncompensated care expenditure was 7.2 percent of total revenues, while the median was 3.9 percent of total revenues. 48 CAHs reported the lowest average (5.6 percent) and median (2.1 percent) of uncompensated care expenditures while HPHs reported the highest average (7.9 percent) and median (4.8 percent) percentages. 49 Of the hospitals reporting uncompensated care expenditures of 3 percent or less of total revenues, 33 percent consisted of HPHs and 59 percent consisted of CAHs. 50 In general, 58 percent of all community types reported uncompensated care expenditures of 5 percent or less of total revenues. 51


The executive compensation portion of the Report details information from the responses to the 9 relevant questions on the questionnaire, plus results from examinations of 20 hospitals which were selected based on the questionnaire responses.

Findings Based on Questionnaire.

The average and median total compensation amounts for the top management official were $490,437 and $377,256. 52 By community type, CAHs and ORHs paid more in compensation than HPHs and USHs, with averages and medians for the top management official ranging from $780,600 and $565,500, respectively, for HPHs and $177,600 and $169,000, respectively, for CAHs. 53 By revenue size, the average and median compensation increased as revenue size increased, with averages and medians ranging from $1,092,400 and $786,100, respectively, for hospitals with revenues of over $500 million and $171,000 and $139,500, respectively, for hospitals with revenues under $25 million. 54

The Report found that 73 percent of hospitals had a written compensation policy, although only 54 percent of hospitals with revenues of under $25 million had a written compensation policy. 55 Additionally, nearly all hospitals approved compensation in advance by non-conflicted individuals and reported using comparables and independent personnel to establish compensation, and these results did not vary by demographic types. 56 Nearly all demographic types used education and experience, specific responsibilities, geographic areas, similar services, and number of beds as factors for setting compensation. 57 The questionnaire did not ask about use of for-profit comparables. 58

With respect to business relationships, 65 percent of hospitals reported at least one relationship with an officer, director, trustee, or key employee, with the most common business relationship being the furnishing of goods, services, and facilities to the hospital. 59 HPHs, USHs, and larger hospitals reported more business relationships than CAHs, ORHs, and smaller hospitals. 60

Findings Based on Examinations.

Of the 20 hospitals selected for examination, 6 were HPHs, 4 were CAHs or ORHs, and 10 were USHs. 61 By revenue size, 8 had revenues of under $250 million, 9 had revenues between $250 million and $500 million, and 3 had revenues of over $500 million. 62 The average and median total compensation paid to the top management official was $1.4 million and $1.3 million 63 .

Nearly all of the examined hospitals had a written conflict of interest policy and a written compensation policy. 64 Compensation was approved in advance in all cases and by non-conflicted individuals in most cases. 65 Among the tools used to measure compensation were published surveys (70 percent), related outside experts (65 percent), and unrelated outside experts (45 percent). 66

With respect to business relationships between the hospital and an officer, director, trustee, or key employee, 40 percent of the examined hospitals reported business relationship (as compared to 65 percent of respondents in the general survey). 67 None of the business relationships resulted in excess benefit penalties. 68 Nearly all (17 out of 20) hospitals complied with rebuttable presumption of reasonableness procedures. 69 No excess benefit penalties were assessed on any of the hospitals that followed the rebuttable presumption, although the IRS was still considering whether excess benefit penalties would be assessed on any of the remaining three hospitals. 70


As noted, the Report has been criticized for reporting inaccurate and incomplete information. Even so, several observations can be made from the Report.

First, many of the traditional concerns of the IRS have not been alleviated by the Report findings. The IRS continues to be concerned that a small number of hospitals may be providing most the community benefit and uncompensated care. 71 The IRS is also concerned that tax-exempt hospitals may be too similar to for-profit hospitals, which is evident in its plans to study for-profit hospitals. 72

Second, the IRS acknowledged its difficulty in enforcing the community benefit and reasonable compensation standards. 73 This admission may result in continued scrutiny of these standards. While Schedule H of the Form 990 should help with community benefit reporting, it is telling that the IRS does not feel it has the ability or information necessary to enforce the exemption standard for hospitals.

Third, the Report made clear that exempt hospitals differ from each other in many ways, such as by revenue size, financial ability, communities served, and community benefits offered. The Report introduced new ways of grouping hospitals based on these various demographics, and these groupings and measurements may serve as an analytical framework for future discussion and studies of tax-exempt hospitals. 74

Fourth, the IRS acknowledged that a single objective standard for measuring community benefit and uncompensated care may have unintended disparate impacts on rural hospitals and smaller hospitals. 75 It is probable that much of the community benefit and charity care expenditures are concentrated in a small number of hospitals because these hospitals have the bulk of the financial resources. ORHs, CAHs, and smaller hospitals reported lower percentages of community benefit and uncompensated care, but also have less excess revenues to devote to community benefit. Many of these hospitals would be unable to meet a standard such as the one proposed by Senator Grassley in 2007, which would have required hospitals to provide charity care 76 in an amount equal to at least 5 percent of patient revenues or expenses in order to retain tax exemption under Code Section 501(c)(3).

Fifth, with respect to executive compensation, the Report indicates that there was widespread compliance with the tax standards for reasonable compensation, as well as widespread use of good governance practices for developing, monitoring, and approving executive compensation. 77 If nothing else, this part of the Report shows that the rebuttable presumption really works, even when the IRS suspects the compensation levels are high. However, unlike with the community benefit portion of the Report, where the IRS painstakingly used neutral language, the executive compensation portion of the Report opined that the compensation levels could be considered high by some, despite legal compliance. It is doubtful that this wording was accidental.


Schedule H reporting will provide much better information than the questionnaire. However, this information will not be available for several more years. In the meantime, the IRS will continue its scrutiny of the community benefit standards and executive compensation.

With respect to community benefit, the IRS plans to: (1) compare tax-exempt hospitals to for-profit hospitals; (2) gather more information on research as a community benefit, including funding and public availability of such research; (3) examine the accuracy of costing methodologies for measuring community benefit; and (4) continue to study whether and to what extent Medicare, bad debt, and community building should be included as community benefit. 78 Additionally, the IRS will evaluate how the community benefit standard and tax-exempt hospitals fit in as part of the ongoing healthcare debate. 79 With respect to executive compensation, the IRS plans to: (1) examine the quality of comparables used to establish executive compensation, especially for-profit comparables; and (2) examine the use of the initial contract exception to see impact on compensation amounts. 80

Meanwhile, Senator Grassley—the perennial tax-exempt hospital critic—continues to urge an overhaul of the community benefit standard, either by Treasury or Congress. 81 However, there seems to be a softening on the 2007 Grassley proposal that would have imposed a single quantitative test. In fact, one congressional aide recently commented that “[a] flat percentage community benefit standard probably isn’t the right answer.” 82

Finally, hospitals can continue to follow the rebuttable presumption of reasonableness and be extra vigilant about the way executive compensation is viewed in today’s economic climate. Additionally, hospitals must continue to do their best to provide, identify, measure, and report community benefits, while at the same time dealing with declining or even negative margins.

1 Laura Gabrysch is a senior associate at Fulbright & Jaworski L.L.P. where her practice encompasses a wide variety of matters related to Federal and local taxation of corporations, partnerships, limited liability companies, and tax-exempt organizations. Laura currently serves as chair of the ABA Health Law Section Tax and Accounting Interest Group.
2 IRS Exempt Organizations (TE/GE) Hospital Compliance Project Final Report (Feb. 2009), available at http://www.irs.gov/charities/charitable/article/0,,id=203109,00.html (herein after, “Hospital Report”).
3 To establish this rebuttable presumption of reasonableness, the following must occur: (1) the compensation arrangement must be approved in advance by the organization's governing body (or a committee of the governing body) composed entirely of individuals who do not have a conflict of interest with respect to the arrangement; (2) the governing body or committee must obtain and rely upon appropriate data as to comparability of compensation before making its determination; and (3) the governing body or committee must adequately document the basis for its determination concurrently with making that determination. Without the rebuttable presumption, an exempt organization must prove the reasonableness of the compensation in any audit by the IRS.
4 Hospital Report, at 1.
5 Id. at 3.
6 Id.
7 Id. at 94-97.
8 IRS Background Paper—Summary of Form 990 Redesign Process (Aug. 2008), available at http://www.irs.gov/charities/article/0,,id=201398,00.html
9 Hospital Report, at 2.
10 Id.
11 Id.
12 Id.
13 John D. Columbo, Some Musings on the IRS Hospital Report, Nonprofit Law Prof Blog, available at http://lawprofessors.typepad.com/nonprofit/2009/02/some-musings-on.html.
14 Statement of Ronald Schultz, Senior Technical Advisor to the Commissioner, TE/GE, Internal Revenue Service, IRS Nonprofit Hospital Report Discussed at D.C. Bar Program, EO Tax J. 2009-36 (March 13, 2009).


Hospital Report, at 4.
16 Id at 96.
17 Id. at 40.
18 Id. at 41.
19 Id. at 57.
20 Id. at 63.
21 Id. at 64-65.
22 Id. at 23.
23 Id.
24 Id. at 27.
25 Id. at 32.
26 Id.
27 Id. at 27.
28 Id at 32.
29 Id.
30 Id. at 35.
31 Id. at 22.
32 Id. at 39.
33 Id.
34 Id. at 94.
35 Id.
36 Id. at 98.
37 Id.
38 Id. at 99.
39 Id.
40 Id. at 102.
41 Id.
42 Id. at 105.
43 Id. at 46.
44 Id.
45 Id. at 40.
46 Id. at 58 & 67.
47 Id. at 45.
48 Id. at 47.
49 Id.
50 Id. at 50.
51 Id.
52 Id. at 124.
53 Id. at 126.
54 Id. at 127.
55 Id. at 129 135.
56 Id.
57 Id. at 136.
58 Id. at 138.
59 Id.
60 Id. at 139.
61 Id. at 141.
62 Id.
63 Id. at 143.
64 Id.
65 Id.
66 Id. at 144.
67 Id. at 145.
68 Id.
69 Id.
70 Id. at 146.
71 Id. at 169.
72 Id. at 171; see Statement by Lois Lerner, Director of the IRS Exempt Organizations Division, on the IRS Report on Nonprofit Hospitals, at Press Briefing (Feb. 12, 2009), available at http://www.irs.gov/charities/charitable/article/0,,id=203109,00.html (hereinafter, “Lerner Statement”).
73 IRS Exempt Organizations Hospital Study Executive Summary of Final Report (Feb. 2009), available at http://www.irs.gov/charities/charitable/article/0,,id=203109,00.html.
74 Hospital Report at 169.
75 Id.
76 “Charity care” in Senator Grassley’s proposal was much more restrictive than uncompensated care that was reported as a result of the questionnaire, and generally would have excluded bad debt and Medicare shortfalls.
77 Id. at 169-170.
78 Id. at 169; Lerner Statement.
79 Lerner Statement.
80 Hospital Report, at 169-170.
81 Press Release of Sen. Grassley, IRS Non-profit Hospitals Study is Helpful, Treasury Should Look at Restoring Charity Care Standards (Feb. 12, 2009).
82 Statement of Theresa Pattara, Tax Counsel , Republican Staff, Senate Finance Committee, IRS Nonprofit Hospital Report Discussed at D.C. Bar Program, EO Tax J. 2009-36 (March 13, 2009).

The ABA Health eSource is distributed automatically to members of the ABA Health Law Section . Please feel free to forward it! Non-members may also sign up to receive the ABA Health eSource.