The Three-Comparables Rule: 95% of Paid Officer Positions at Small Nonprofits Have Three Peers in the Same State, 51% in the Same City

Of 53,507 paid officer positions at small nonprofits (49,623 organizations, 49,643 filings, tax years 2021–2024), 95.3% have at least three comparable organizations filing in the same state. Narrow the community to the same city and the share falls to 50.6%. Require the same city and the same NTEE sector and it falls to 18.8%.

The rule

Treasury regulations include a special rule on comparability data for small organizations. Treas. Reg. §53.4958-6(c)(2)(ii) reads:

The rule speaks only to the data. The rebuttable presumption of reasonableness has two further conditions: advance approval by an authorized body whose members have no conflict of interest, and documentation of the basis for the decision made at the time the decision is made (Treas. Reg. §53.4958-6(a)). See the rebuttable presumption of reasonableness for all three.

The regulation does not define a similar community, and this analysis does not decide what counts as one. It measures how many comparable organizations a small organization can find under four readings, from broad to strict.

The finding

The gradient is the finding. Read as the state, the community almost always holds three comparable organizations that file publicly. Read as the city, it holds them for about half of positions. Adding a match on the kind of services, measured here by NTEE sector, leaves fewer than one position in five with three peers.

The state-level share barely moves with size: 95.3% for organizations with revenue under $500,000 (36,605 positions) and 95.4% from $500,000 to under $1 million (16,902 positions). It does move with the role. CauseComp's Executive Director / CEO role has three in-state peers for 99.6% of positions (32,605 positions); every other officer role combined, 88.8% (20,902 positions).

What this means for a small organization's board

Three comparables in the same state are within reach for nearly every paid position in this population. Three in the same city are not, for about half of them, and three in the same city and field are the exception. A board relying on the small-organization rule can record which reading of “same or similar communities” it applied, and why, alongside the organizations it used. For building and recording the peer set, see nonprofit compensation peer groups and what counts as appropriate comparability data. CauseComp's comparables set, drawn from Form 990 filings, is part of the Professional plan; see plans and pricing.

Method and scope

Scope of every figure on this page: the CauseComp Form 990 corpus as refreshed on September 9, 2026, restricted to a frame of 700,545 officer rows from 347,024 filings by 110,004 organizations, tax years 2021–2024, covering the 107 officer roles CauseComp publishes and organizations with a known NTEE sector. Each filing is one Form 990 or 990-EZ return. Every cut reported has at least 20 positions. The page reports counts and shares only, and no compensation figure. Earlier CauseComp research pages use a different frame and count filings differently, so their totals are not comparable to these.

• Small: total revenue under $1 million on the organization's latest filing in the frame (the return's total revenue or, where the return's figure is missing, the revenue on the organization's record in CauseComp's NTEE reference file). The regulation's test is annual gross receipts including contributions, which the corpus does not record. Total revenue is net of some amounts that gross receipts include, such as the cost of goods sold, so it can be lower than gross receipts, and the screen can include some organizations whose gross receipts are $1 million or more. The regulation also permits an average of the three prior taxable years and requires controlled entities to be aggregated (Treas. Reg. §53.4958-6(c)(2)(iii)); this screen uses a single year and no aggregation.
• Position: one organization and one officer role on that latest filing: 53,507 positions at 49,623 organizations, from 49,643 filings. Roles are CauseComp's normalization of filed titles. Where several people share a role at one organization, the role counts once.
• Peer: another organization whose latest filing in the frame reports the same role, with total revenue between half and twice the target's. Peers are counted as distinct organizations and may themselves have revenue of $1 million or more. The peer pool is 179,251 positions at 110,004 organizations.
• Community and services: community is proxied by the state, or the state and city, on the organization's record in CauseComp's NTEE reference file, one of each per organization; services are proxied by the organization's NTEE sector. Neither proxy is a legal definition of a similar community or of similar services.
• Vintages mix: each peer is counted from its own latest filing in tax years 2021–2024.

Who is in the data

Only small organizations that report a compensated officer appear. The corpus extracts Part VII and 990-EZ officer rows only where compensation is at least $10,000 and the return reports revenue of at least $25,000, and Part VII rows also need at least 20 hours a week; Schedule J rows are not subject to those floors. Rows then pass the frame's role, sector and data-quality filters. Of 821,953 full Form 990 filings for tax years 2021–2024 with total revenue under $1 million, 13.6% (112,142) carry an officer row that enters the frame. Of 820,820 Form 990-EZ filings for the same years, 3.9% (31,658) do. The findings describe small organizations with a paid officer in the data, not all small organizations.

1. (a) A count of comparable organizations says the filings exist. It does not say a board would judge each one comparable on every relevant factor.
2. (b) City matching uses the city on the organization's record in CauseComp's NTEE reference file, so spelling variants and neighboring municipalities fall outside a match; the city-level shares may understate what a board reading community as a metropolitan area would find.
3. (c) The shares under “Who is in the data” count filings, not organizations, across tax years 2021–2024.
4. (d) This analysis is research, not legal or tax advice; organizations should consult their advisors on their specific facts.

Frequently asked questions