What Nonprofits Actually Pay Their Leaders

Boards setting a salary, candidates weighing an offer and journalists writing about a charity all reach for the same figure, and it is not published anywhere useful. So we calculated it from the tax returns of 256,539 organizations.

The first finding is the one nobody expects. Most American nonprofits do not pay their leadership at all.

Where these numbers come from

The IRS publishes a free annual extract of financial data from every Form 990 it processes. It is not a survey and not a sample. Filtering the processing year 2024 file to 501(c)(3) organizations with at least $25,000 of annual expenses leaves 256,539 organizations.

The figure analysed is Form 990 Part IX line 5, which the IRS field documentation labels “Compensation of current officers, directors, etc”. Two things follow from that and both matter enough to state before any number appears.

How much the first row distorts things is measurable rather than a matter of opinion, and it is tested further down.

Most nonprofits pay their leadership nothing

Across every organization in the dataset, the share reporting no officer compensation at all falls steadily with size and never reaches zero.

Among organizations spending under $250,000 a year, 79% pay their officers, directors and key employees nothing. At $1m to $2m it is still 39%. Even above $50m, 14% report nothing on that line.

For the smallest organizations this is the expected picture: a volunteer board, no executive, and a founder who is not on payroll. Higher up it usually means something more specific, most often that the people running the organization are paid as ordinary staff rather than being classed as officers, or that leadership is genuinely voluntary at an institution funded to do one narrow thing.

It matters for reading any compensation figure at all. Every median below describes only the organizations that pay something, which at the small end is a minority of one in five.

What the organizations that do pay actually pay

Read the spread rather than the median. In every band the 75th percentile is roughly two to three times the 25th, which is a wide range for organizations of the same size. Budget size explains a great deal about leadership pay and it does not explain most of it.

The question people actually ask

Search traffic on this subject is dominated by one phrasing repeated a dozen ways: what should a chief executive earn at an organization with a $4 million budget. It is a reasonable question and it has an answer.

A $4m organization sits in the $2m to $5m band. The median reported for officers and key employees combined is $168,486, with a quarter of organizations below $115,960 and a quarter above $270,133.

If that organization has one executive and no other officers on payroll, the median is a fair read of the executive’s package. If it has a chief executive and a finance director both classed as officers, the same figure covers both, and the individual number is lower. Which of those you are looking at cannot be determined from this data, and it is the single most important thing to establish before quoting it.

Leadership pay as a share of the budget

The share of the budget going to leadership falls sharply and consistently, from 26.3% under $250k to 1.2% above $50m.

The top of that range sounds alarming and usually is not. An organization spending $150,000 a year with one part-time executive on $40,000 is spending 26% of its budget on that person, and there is no arrangement in which it does not. Overhead ratios computed at that size measure the arithmetic of small numbers rather than the discipline of the organization.

Total staff cost behaves differently and is worth seeing alongside it, because it moves the other way.

Leadership takes a smaller slice as organizations grow while payroll overall takes a larger one, until around $10m where it levels off. An organization budgeting for growth should expect total staff cost to rise towards half of everything, and should not expect the executive line to rise with it.

Is this one person’s salary?

This is the objection the whole page turns on, so it is worth testing rather than hedging.

The test: restrict to organizations reporting officer compensation and zero other salaries and wages. Those are organizations where the paid officer is almost certainly the only person on payroll, so the aggregate is one person by construction. If that median tracks the median for everyone, the aggregate is a fair proxy. If it does not, it is not.

Up to about $5m the two medians agree within ten percent and mostly within three, so in that range the reported figure is a reasonable read of one person’s compensation. Above $5m they separate by around twenty percent, which is what you would expect once organizations routinely have several officers.

So the top two bands should not be read as a chief executive’s pay. A median of $1,716,373 above $50m is what a group of senior people cost, not what one of them earns. Anyone quoting that as a charity chief executive’s salary is misreading the form, and it happens constantly.

What a board is supposed to do with this

Setting executive pay is one of the few board decisions with a specific legal procedure attached, and following it is worth more than landing on any particular number.

Section 4958 of the tax code taxes excess benefit transactions, and unreasonable compensation is the most common form. The regulations offer a rebuttable presumption of reasonableness if three things happen: the arrangement is approved in advance by an authorized body with no conflict of interest, that body obtains and relies on appropriate data as to comparability beforehand, and it documents the basis for its decision at the time rather than afterwards. Meet all three and the burden shifts to the IRS to prove the amount was unreasonable.

For organizations with annual gross receipts under $1 million there is a lighter standard: the comparability requirement is met with data on compensation paid by three comparable organizations in the same or similar communities for similar services.

Note what that says, because it limits what this page can do for you. The safe harbour asks for organizations in similar communities, and this is national data. It tells you whether a proposed figure is ordinary or unusual for the size, which is genuinely useful context and a good way to sanity check a number before a meeting. It does not by itself satisfy the requirement. Three named peer organizations, whose Form 990s are public, does.

The mechanism for keeping the interested person out of the vote is in the conflict of interest policy, the tax consequences of getting it wrong are in what a nonprofit organization is, and the wider set of things a board is accountable for is in the board of directors guide.

Limits worth stating

Full Form 990 filers only. The 990-EZ and the 990-N do not carry a functional expenses statement, so organizations below those thresholds are absent. The smallest band here is not small nonprofits in general.

It is an aggregate, not a salary. Covered at length above, and the reason the top two bands carry a warning rather than a headline.

No geography, no sector, no job title. The extract carries an EIN and financial fields, not a state, a cause area or a role. A rural food bank and a Manhattan research institute of the same size sit in the same row.

Part IX line 5 is tax year based. The individual figures in Part VII and Schedule J use the calendar year, so a number here will not always match the one a journalist quotes from those sections of the same return.

Processing year, not fiscal year. Returns processed in 2024 cover a spread of fiscal years, mostly 2022 and 2023, so this lags current pay by a year or two.

Reproducing this

The IRS file is a free download and the script that produces every figure above is published rather than described, so the bands, the filters and the definition of a single-payroll organization can all be changed and rerun by anyone who disagrees with them.

Where these numbers came from, and how to read any organization’s own return, is in Form 990. What the same dataset says about liquidity is in nonprofit cash reserves, and the compensation process a board should have written down is in nonprofit policies.

Questions people ask

How much does the average nonprofit executive make?

There is no single figure, because pay tracks budget size closely. Among organizations that pay their officers anything, the median reported for officers and key employees combined runs from $41,445 at organizations spending under $250,000 to $168,486 at $2m to $5m. Medians are used throughout rather than averages, because a handful of very large institutions would drag a mean somewhere useless.

What should a nonprofit CEO earn at a $4 million organization?

A $4m organization falls in the $2m to $5m band, where the median reported for officers and key employees is $168,486, the 25th percentile is $115,960 and the 75th is $270,133. If more than one person is classed as an officer, that total covers all of them and the individual figure is lower.

Do most nonprofits pay their executive director?

No. 79% of organizations spending under $250,000 a year report no compensation at all for officers, directors and key employees, and 39% of organizations spending $1m to $2m report none. Volunteer leadership is the norm at the small end of the sector rather than the exception.

Is this the chief executive's salary?

Not exactly, and above $5m not at all. Form 990 Part IX line 5 is a total for every officer, director, trustee and key employee. Testing it against organizations with only one person on payroll shows the two agree within ten percent up to about $5m, then diverge by roughly twenty percent above that. The $1.7m median in the top band is what a leadership team costs, not one person's pay.

What percentage of a nonprofit budget should go to salaries?

Total staff cost is a median of 35% of expenses at organizations under $250,000, rising to about half between $5m and $50m. Leadership compensation specifically moves the opposite way, from 26.3% of the budget at the smallest organizations to 1.2% above $50m.

How do I find comparable salaries for my own board?

Every nonprofit's Form 990 is public, so the most defensible approach is to pull the returns of three genuinely comparable organizations in your own area and read Part VII, which lists individuals by name and role. For organizations under $1 million in gross receipts that is also what the regulations ask for.

Does using this page satisfy the IRS comparability requirement?

No. The safe harbour for smaller organizations asks for data on three comparable organizations in the same or similar communities, and this is national data with no geography attached. It is a good way to check whether a proposed figure is ordinary for the size before a board meeting. It is not a substitute for named comparables.

Where does this data come from?

The IRS Statistics of Income annual extract of tax-exempt organization financial data, Form 990, processing year 2024. It is a free public download. 256,539 501(c)(3) organizations with at least $25,000 of annual expenses met the criteria used here.