IRS Comparability Data Requirements
Under IRC Section 4958, the IRS can impose excise taxes on nonprofit executives who receive unreasonable compensation. The rebuttable presumption of reasonableness — commonly called safe harbor — protects organizations that follow specific comparability data procedures. Here is what you need to know.
Chat with any nonprofit's data
Coming Soon
Instant, data-backed answers on compensation, financials & more.
Join WaitlistSmart Data Platform
Key Takeaways
The IRS rebuttable presumption of reasonableness (safe harbor) shifts the burden of proof to the IRS rather than the organization when compensation is challenged.
Safe harbor requires three elements: an independent board or committee, appropriate comparability data, and concurrent documentation of the decision.
Acceptable comparability data includes Form 990 filings, independent compensation surveys, and written expert opinions from qualified consultants.
Organizations must match comparables by budget size, geographic region, sector, and job complexity — not just title alone.
Failing to establish safe harbor does not automatically mean compensation is unreasonable, but it leaves the organization with the burden of proof.
What Is Comparability Data?
The foundation of reasonable compensation under IRS rules.
Comparability data is compensation information from similarly situated organizations used to determine whether a nonprofit executive's pay is reasonable. The concept comes from IRC Section 4958 and the Treasury Regulations that implement it (26 CFR 53.4958-1 through 53.4958-8), specifically the <a href="/nonprofits/resources/excess-benefit-transactions-intermediate-sanctions">intermediate sanctions</a> rules enacted in 1996 as part of the Taxpayer Bill of Rights 2. [1] [2]
The IRS does not define a specific dollar threshold for "reasonable" compensation. Instead, reasonableness is defined as the amount that would ordinarily be paid for like services by like enterprises under like circumstances (26 CFR 53.4958-4). [3] This is where comparability data becomes essential — it is the evidence that supports your compensation decisions.
Excess Benefit Transaction
A transaction in which a disqualified person (typically a key executive or insider) receives compensation or other economic benefit from a tax-exempt organization that exceeds the value of what the organization received in return. Defined in IRC Section 4958(c) and 26 CFR 53.4958-4. [4]
Disqualified Person
Any person who was in a position to exercise substantial influence over the affairs of the organization at any time during a five-year lookback period. Defined in IRC Section 4958(f) and 26 CFR 53.4958-3. This typically includes officers, directors, key employees, and their family members. [5]
The Rebuttable Presumption of Reasonableness
How safe harbor shifts the burden of proof.
The rebuttable presumption of reasonableness is the IRS's safe harbor mechanism for nonprofit compensation. When properly established, it creates a legal presumption that the compensation arrangement is reasonable. The IRS can still challenge the arrangement, but the burden shifts — the IRS must prove the compensation is unreasonable rather than the organization having to prove it is reasonable.
This is a significant legal advantage. Without safe harbor, an organization facing an IRS audit on compensation must affirmatively demonstrate reasonableness. With safe harbor, the IRS must overcome the presumption with clear and convincing evidence — a much higher bar.
Why Safe Harbor Matters
Even if the IRS never audits your organization, establishing safe harbor is important. It demonstrates governance best practices to donors, state regulators, and watchdog organizations. It also protects individual board members from potential personal liability for approving excess benefit transactions.
The rebuttable presumption applies to compensation arrangements and property transfers between a 501(c)(3) or 501(c)(4) organization and its disqualified persons. It is established through a three-part test outlined in 26 CFR 53.4958-6. [6]
The Three-Prong Test
All three requirements must be met to establish safe harbor.
Requirements for the Rebuttable Presumption (26 CFR 53.4958-6)
Independent Body Review (53.4958-6(c)(1))
The compensation arrangement must be approved by an authorized body composed entirely of individuals with no conflict of interest regarding the transaction. This is typically a board compensation committee or the full board with conflicted members recused. [6]
Appropriate Comparability Data (53.4958-6(c)(2))
Before making the decision, the authorized body must obtain and rely upon appropriate comparability data. This means actual compensation data from organizations of similar size, scope, and geography — not gut feelings or informal benchmarks. [6]
Concurrent Documentation (53.4958-6(c)(3))
The authorized body must adequately document the basis for its decision at the time the determination is made. Minutes must record the terms of the arrangement, the comparability data relied upon, the members who voted, and the rationale for the decision. [6]
All Three Are Required
Missing any single prong eliminates the presumption entirely. The most common failure is the third prong — organizations obtain good data and have an independent committee but fail to document their deliberations contemporaneously. Documenting after the fact does not satisfy this requirement.
Acceptable Data Sources
What the IRS considers appropriate comparability data.
26 CFR 53.4958-6(c)(2) identifies several types of data that qualify as appropriate comparability data. The key is that the data must reflect compensation paid by similarly situated organizations for functionally comparable positions. [6]
IRS-Accepted Comparability Data Sources
IRS Form 990 filings from comparable tax-exempt organizations, including <a href="/nonprofits/resources/form-990-schedule-j-explained">Schedule J</a> compensation data for officers, directors, and key employees.
Independent compensation surveys conducted by recognized firms that cover the relevant sector, geography, and organization size range.
Written opinions from independent compensation consultants who are qualified to render expert opinions on nonprofit executive compensation.
Actual compensation data compiled from job postings, published salary databases, or compensation studies by industry associations.
Data from taxable organizations doing comparable work, if tax-exempt comparables are insufficient — for instance, a nonprofit hospital comparing to for-profit hospital systems.
Small Organization Exception
For organizations with average annual gross receipts of less than $1 million (averaged over the three prior tax years), 26 CFR 53.4958-6(c)(2)(ii) provides a simplified standard. Comparability data can include data from three or more comparable organizations in the same or similar communities for similar services. The regulation permits data gathered even through informal telephone surveys, as long as a written summary is prepared. [6] [7]
Form 990 data is particularly valuable because it is publicly available, covers the vast majority of tax-exempt organizations, and includes detailed breakdowns of base compensation, bonus and incentive pay, deferred compensation, nontaxable benefits, and other reportable compensation (reported in Part VII and <a href="/nonprofits/resources/form-990-schedule-j-explained">Schedule J</a>). Form 990 Part VI, Section B, Line 15 also asks whether the organization used a comparability study when setting compensation — a public disclosure that creates transparency pressure. [9] [10] RoundPaper's platform aggregates this data across 3.6M+ filings to make comparability analysis faster and more accurate.
Matching Comparables
Selecting the right peer organizations for comparison.
Using comparability data is not simply looking at average CEO salaries across all nonprofits. The IRS expects organizations to compare against peers that are genuinely comparable on multiple dimensions. Getting the peer group wrong can undermine the entire analysis.
Factors for Selecting Comparable Organizations
Budget size and total revenue — A $2M community nonprofit should not benchmark against a $200M university hospital.
Geographic location — Compensation varies significantly by region. A nonprofit in San Francisco faces different labor markets than one in rural Mississippi.
Sector and mission type — Health care, education, social services, arts, and religious organizations have distinct compensation markets.
Organizational complexity — Number of employees, number of programs, multi-state operations, and regulatory complexity all affect what constitutes reasonable pay.
Job responsibilities — Compare actual duties and scope of authority, not just title. An ED at a 5-person nonprofit has a fundamentally different role than an ED at a 500-person organization.
How Many Comparables Do You Need?
The regulations do not specify an exact number for organizations above $1M, but best practice is to use at least five to ten comparable organizations. Too few comparables can be dismissed as cherry-picked. Too many unfiltered comparables dilute the relevance. For organizations under $1M in average gross receipts, 26 CFR 53.4958-6(c)(2)(ii) requires at least three comparables in the same or similar communities. [6] [7] [12]
Documentation Requirements
What concurrent documentation must include.
The third prong of the safe harbor test — concurrent documentation — is where many organizations fall short. 26 CFR 53.4958-6(c)(3) is specific about what must be documented and when. Documentation must be prepared before the later of the next meeting of the authorized body or 60 days after the final action. [6]
Required Documentation Elements
The terms of the compensation arrangement (salary, bonus structure, benefits, deferred compensation, severance).
The date of the board or committee approval.
The members of the authorized body who were present during debate and those who voted.
The comparability data obtained and relied upon, and how it was obtained.
Any actions taken by a member of the authorized body who had a conflict of interest (e.g., recusal).
Documentation that the authorized body determined compensation was reasonable based on the comparability data.
If compensation exceeded the range in the comparability data, the basis for the determination that it was still reasonable.
Timing Is Critical
Per 26 CFR 53.4958-6(c)(3)(ii), the documentation must be concurrent — meaning recorded before the later of the next meeting of the authorized body or 60 days after the final action. Retroactive documentation, such as drafting minutes months later or assembling comparability data after the fact, does not satisfy the concurrent documentation requirement and will not establish the rebuttable presumption. [6]
Common Mistakes
Pitfalls that can undermine your safe harbor protection.
Using stale data
Relying on compensation surveys or Form 990 data that is more than two to three years old. Compensation markets change, and outdated data may not reflect current conditions.
Mismatched comparables
Comparing against organizations that are significantly larger, smaller, or in a different sector. A $500K budget arts nonprofit should not benchmark against a $50M hospital system.
Conflicted decision-makers
Allowing the executive whose compensation is being set to participate in the discussion or vote. Even if the executive is a board member, they must recuse themselves.
Incomplete documentation
Recording only the final salary number without documenting the comparability data reviewed, the discussion, or the rationale. This fails the concurrent documentation prong.
Ignoring total compensation
Looking only at base salary while ignoring bonuses, deferred compensation, housing allowances, vehicle benefits, and other perquisites. The IRS evaluates the total value of all compensation and benefits per 26 CFR 53.4958-4(a). [3] [4]
Rubber-stamping without review
Having the board approve compensation without actually reviewing the data or engaging in meaningful deliberation. The authorized body must genuinely evaluate the information.
Consequences of Non-Compliance
What happens when compensation is deemed an excess benefit transaction.
If the IRS determines that an <a href="/nonprofits/resources/excess-benefit-transactions-intermediate-sanctions">excess benefit transaction</a> has occurred, IRC Section 4958(a) and (b) impose a two-tier excise tax structure. The penalties apply to the individual who received the excess benefit — not the organization — although the organization can also face consequences. [1] [8]
Excise Tax Penalties
First-tier tax (IRC 4958(a)(1)): 25% of the excess benefit amount, imposed on the disqualified person who received the excess compensation. [1]
Second-tier tax (IRC 4958(b)): 200% of the excess benefit amount if the transaction is not corrected within the IRS-specified taxable period. Correction requires returning the excess amount to the organization. [1]
Organization manager tax (IRC 4958(a)(2)): 10% of the excess benefit (up to $20,000 per transaction) on any organization manager who knowingly approved the transaction. [1] [8]
Revocation of Tax-Exempt Status
In extreme cases, the IRS can revoke an organization's 501(c)(3) status entirely. While intermediate sanctions were designed as an alternative to revocation (see IRS EO CPE Text, Topic H), the IRS retains the power to revoke exemption for repeated or egregious excess benefit transactions. This outcome is rare but possible. [2] [11]
Beyond IRS penalties, excess compensation can trigger state attorney general investigations, negative media coverage, and loss of donor confidence. Many state regulators actively review Form 990 data and may independently challenge compensation that appears unreasonable. [10] [12]
Best Practices Checklist
A practical framework for establishing and maintaining safe harbor.
Annual Compensation Review Process
Establish a standing compensation committee of independent board members with no financial relationship to the executives being evaluated.
Gather current comparability data annually — ideally from multiple sources including Form 990 filings, <a href="/nonprofits/resources/nonprofit-executive-compensation-benchmarking">compensation surveys</a>, and expert analysis.
Define your peer group before reviewing data: match by budget size (within 50-200% of your revenue), geographic region, sector, and organizational complexity.
Review total compensation including base salary, bonuses, deferred compensation, benefits, housing, vehicle allowances, and all other economic benefits.
Document the committee's deliberations in real time: record the data reviewed, the discussion, the vote, and the rationale in board minutes.
If compensation falls outside the comparable range, document the specific reasons justifying the deviation (e.g., retention risk, unique qualifications, market conditions).
Have conflicted parties recuse themselves from all deliberation and voting — not just the final vote.
Retain all supporting materials (surveys, Form 990 printouts, consultant reports) with the meeting minutes for at least seven years. [10]
Review and update your compensation policy annually as markets change and organizational needs evolve. [12]
Consider engaging an independent compensation consultant for complex arrangements or when compensation exceeds $200,000. [12]
How RoundPaper Helps
RoundPaper's data platform aggregates compensation data from 3.6M+ IRS Form 990 filings, making it fast and easy to pull comparability data filtered by budget size, geography, sector, and role. Instead of manually searching individual 990s, you can generate a peer comparison in seconds. Join the waitlist for early access.
Sources & Citations
Primary sources used to research and verify this resource.
This resource is for informational purposes only and does not constitute legal or tax advice. Consult a qualified attorney or tax professional for advice specific to your organization.
Ask anything about
any nonprofit
Get instant, data-backed answers about nonprofit compensation, financials, and trends. Join the waitlist for early access. Free tier included at launch.
Trusted by nonprofit professionals
3.6M+
IRS Filings
1.7M+
Organizations
28M+
Comp Records
Keep Reading