A May 2025 Pennsylvania Supreme Court decision drew national attention to a question that nonprofit hospital boards have long treated as settled: what does it actually take to defend executive compensation at a tax-exempt hospital? A central part of the court’s answer was comparability.
The case began in 2023, when a Pennsylvania appellate court denied property tax exemptions for four Tower Health hospitals. The court found that the system’s executive compensation structure – with 40% of bonuses tied to financial performance – reflected a profit motive inconsistent with charitable status. The system’s five highest-paid executives each earned more than $1 million in 2019, with the president exceeding $2 million. Management fees charged to Pottstown Hospital reached as much as $23 million in a single year – figures the Supreme Court ultimately found insufficient on their own to disqualify Pottstown, since a subsidiary is ordinarily evaluated separately from its parent system.
Pottstown Hospital subsequently won its exemption back before the Pennsylvania Supreme Court. To assess whether the hospital’s pay was reasonable, the court imported the factors federal law uses for nonprofit compensation – beginning with what comparable organizations pay for comparable roles. It concluded that executive pay was “within fair market value as compared to similar executives at similar healthcare institutions.”
The lesson for hospital boards is practical: when compensation is challenged, the defense turns on comparability. A board that can demonstrate its pay against a credible peer group is in a far stronger position than one that cannot.
A More Demanding Regulatory Environment
Scrutiny of nonprofit hospitals’ tax-exempt status intensified in 2026, and executive compensation is one of the threads running through it.
In April, Treasury announced that the IRS will revise Form 990 to require more detailed disclosure on governance and fiscal management, with an initial focus on government grants, contracts, and fiscal sponsorships. No proposed rule has been issued yet. Schedule J already requires reporting compensation above $150,000, and Form 990 remains a public document, available to any stakeholder at no cost.
The House Ways and Means Committee has also turned its attention to nonprofit hospitals. At an April hearing on hospital prices and consolidation, Chairman Jason Smith told the CEOs of CommonSpirit Health, ECU Health, HCA Healthcare, and NewYork-Presbyterian that the prices they charge patients are “borderline extortion.” In July, the committee advanced the Tax-Exempt Hospital Transparency Act, which would tighten charity care obligations for tax-exempt hospitals. Supporters of the bill cited executive compensation among the sector’s problems. The hearing’s focus was prices, not pay, but the direction is clear: the tax-exempt status of large hospital systems is under active congressional scrutiny.
The Lown Hospitals Index now grades nonprofit hospitals publicly on pay equity and community benefit spending. Its review of the Tower Health matter noted that at least 40 nonprofit hospital CEOs in Pennsylvania alone earned $1 million or more in 2019, with fourteen hospitals receiving a grade of “D” on pay equity in the 2022 Index. These scores are searchable by policymakers, journalists, community members, and anyone evaluating a hospital’s claim to tax-exempt status.
The legal standard is the one that has applied since 1996: the rebuttable presumption under Section 4958 rests on independent board approval, appropriate comparability data, and contemporaneous documentation. The difference now is how many people are looking.
Where the Compliance Gap Concentrates
Large health systems typically maintain compensation governance practices adequate to current scrutiny. They engage specialized consultants, participate in proprietary benchmarking surveys, and produce structured analyses for board review on an annual cycle.
The gap is more pronounced among independent community hospitals and critical access hospitals with revenues below $500 million. These organizations face identical Section 4958 compliance requirements, with substantially fewer resources to meet them.
The typical process at this level involves a compensation committee reviewing executive pay once or twice annually, drawing on publicly available Form 990 filings and perhaps a third-party report. HR leaders at some organizations supplement this by calling counterparts at nearby health systems to collect informal salary figures – an approach that typically produces only a few data points and may not resolve the comparability problem, since proximate hospitals often differ considerably in revenue, service mix, and organizational structure. Board minutes note that comparability data was reviewed, and the compensation package is approved.
Several structural weaknesses undermine this approach.
Peer group construction is problematic. A $200 million community hospital in a rural market shares little in common with a $3 billion academic medical center, yet both appear in a broad search for nonprofit hospital executive compensation. Controlling for revenue, geography, organizational type, and mission complexity is essential to meeting the comparability standard. Unfiltered data does not accomplish this.
The major proprietary compensation surveys present a third challenge. By Sullivan Cotter’s ownpublished participant data, its Health Care Management and Executive Compensation Survey – the sector’s standard reference for nearly 35 years – draws on 361 parent organizations. Of these, 161 (45%) have revenues above $3 billion. Only 42, or 12% of the sample, report revenues under $500 million.
The survey’s participants reflect Sullivan Cotter’s client base, which is primarily large health systems. Smaller hospitals rarely engage a major compensation consultant, so their data seldom makes it into the survey. A $200 million hospital that buys it is benchmarking against organizations that look nothing like it, and it may find fewer than ten truly comparable data points.
Geographic variation compounds each of these issues. A salary of $400,000 in Birmingham, Alabama is not directly comparable to the same figure in Seattle, Washington. Labor markets for healthcare executives vary considerably across regions, and a compensation committee that applies national medians without geographic adjustment is not meeting the standard for appropriate comparability data. Most surveys publish regional breakdowns but do not normalize for wage-level differences across markets.
The Three Requirements of the Rebuttable Presumption
IRS Section 4958 establishes three specific elements that, when satisfied, create a legal presumption of reasonableness that the IRS bears the burden of overcoming.
Independent approval requires that the compensation arrangement be approved by an authorized body composed entirely of individuals without a conflict of interest. Compensation committee members cannot include the executives whose pay is under review, nor individuals with financial relationships that could compromise their independence.
Comparability data requires that the authorized body obtain and rely on appropriate data before making its determination. The IRS expects data drawn from organizations of similar size, scope, geography, and complexity. A structured analysis controlling for these variables satisfies this requirement. A collection of general 990 lookups from organizations of varying size does not.
Contemporaneous documentation requires that the basis for the determination be recorded at the time of the decision. Documentation should specify what comparability data was used, how the committee obtained it, the methodology applied, the peer organizations included, the committee’s conclusions, and the date of the determination. Board minutes limited to “compensation was discussed and approved” fall short of this standard.
When all three elements are present, the burden of proof shifts to the IRS. When any element is missing or inadequate, that burden remains with the organization – and even compensation that is substantively reasonable becomes difficult to defend.
An Operational Consideration for Boards
The rebuttable presumption is a safe harbor, not a mandate, but it is the strongest position a board can be in if its compensation is ever challenged. Pottstown shows that courts outside the IRS context are reaching for the same test: the hospital won because the record showed its pay was in line with comparable institutions.
For boards and compensation committees at mid-size nonprofit hospitals, the relevant question is whether current documentation practices would satisfy this standard under direct scrutiny.
Organizations that are not confident in their answer may benefit from examining three dimensions of their current process: the recency of the underlying data, the rigor of peer group construction, and the specificity of the committee’s contemporaneous documentation.
CompIntel was built to address each of these dimensions for mid-size organizations. It draws from the complete universe of nonprofit hospital IRS filings to construct peer groups with hundreds of data points in the sub-$1 billion revenue range, matched by revenue, geography, and organizational type. Geographic wage indexing is applied to every comparison, so that benchmarks reflect the labor markets in which a hospital actually competes. The resulting documentation is structured to meet the specificity the rebuttable presumption requires.
David Lieu is the founder of CompIntel and the former COO at Sullivan Cotter.