How Many Staff Nonprofits Have, and What They Cost

Every published figure for nonprofit salaries comes from a survey, which means it comes from the organizations that chose to answer. The tax returns do not have that problem: every filing organization is in them, willing or not.

So we counted staff and payroll across 256,539 charities. 60.7% employ at least one person, and among those the median headcount is 14. The typical organization spends $37,122 per employee per year on everything: wages, benefits and payroll tax combined.

That last number is lower than any published salary survey, and the reason why is the most useful thing on this page.

Why this is not an average salary, and what it is instead

Form 990 Part V line 2a asks for the number of employees reported on the organization’s W-3 for the year. That is a headcount, not a full time equivalent count. A summer camp that employs forty counsellors for eight weeks reports forty employees. A food bank with three full time staff and sixty weekend shift workers reports sixty three.

Divide payroll by that and you get cost per W-2 employee, which is a real and useful number, but it is not what anyone means by a salary. Wherever part time and seasonal work is common, and in this sector it is very common, the figure sits well below what a full time person is paid.

We are publishing it anyway, clearly labelled, because it answers a question no survey can: what does the whole sector spend per person on its payroll. For the actual salary question, the honest measure further down is the count of people paid over $100,000, which no amount of part time dilution can move.

How many people a nonprofit employs

Two things stand out. The first is how small the small end is: the median organization under $250,000 that employs anyone at all employs three people, and 71% employ nobody. The second is how quickly it scales, from a median of 3 to a median of 520.

The ratio that falls out of this is the useful planning number: total budget divided by headcount, which is what a post has historically needed behind it rather than what it was paid. The median runs $59,020 under $250,000, $70,992 to $1m, $80,645 to $5m, $88,210 to $25m and $127,402 above that.

Read it against your band and never against a band boundary, because an organization at the bottom of a wide band and one at the top are nowhere near each other. If your own figure comes out well below your band, either the roles are part time or the budget cannot carry them.

What a nonprofit spends per employee

This is total employment cost: salaries, officer compensation, pension contributions, other benefits and payroll tax, divided by headcount. The sector median is $37,122, of which $31,956 is cash compensation and the rest is benefits and tax.

The gradient is real but it is not only a pay gradient. Larger organizations pay more per person, and they also use proportionally fewer seasonal and part time staff, and both effects push the same direction. Do not read the bottom row as an average salary of $22,606. Read it as: organizations this size spend about $22,606 per person on their payroll, and most of those people are not full time.

Bounding the part time problem

The dilution argument above is easy to assert and worth testing, so we cut the same figure by headcount rather than by budget. An organization reporting exactly one W-2 employee cannot be spreading its payroll across a crowd of seasonal staff.

This gradient runs backwards to the one by budget size, and that is the proof. If low pay were the explanation, cost per head would rise with headcount the way it rises with budget. Instead it falls, from $54,440 at a single employee to $30,994 between 11 and 50, before turning back up over 50 staff where the pay gradient finally outweighs the dilution.

So the $54,440 at the top of that table is the most defensible full time figure in this data: one person, the whole payroll, nothing to average across. It is 47% above the all filers median of $37,122, and that gap is the size of the part time effect rather than anything about wages.

It is still not a salary. A sole employee can be half time, and $54,440 is total employment cost including tax and benefits. Cutting to cash compensation only, the median for a sole employee organization is $49,332 across 11,809 of them. That is the closest figure in this data to a full time nonprofit salary, and it should be read as a floor for a full time post rather than a going rate.

The measure that cannot be diluted

Form 990 asks separately how many individuals received more than $100,000 in reportable compensation. That is a count of people, so part time staff cannot drag it down.

24.4% of charities, 62,485 organizations, report at least one person paid more than $100,000. Below $1m it is rare enough to be remarkable. Above $25m it is universal, and the median such organization has 26 of them.

The jump between $1m and $5m is the one worth noticing: 43.0% have somebody over six figures, and the median count is one. That is the executive director, and it marks the budget level at which the sector starts paying market rate for its senior role. The distribution of that specific salary is on the nonprofit executive compensation page.

Testing the headcount field against the payroll field

A number this convenient deserves a check, so we ran the two relevant fields against each other. An organization reporting payroll expense should report employees, and vice versa.

The second column is negligible. The first is not: between 5% and 13% of organizations report spending money on payroll while reporting zero employees. Some of that is staff paid through an employer of record or a fiscal sponsor, so no W-3 exists in the organization’s own name. Some is simply line 2a left at zero.

The consequence is that 60.7% employing someone is a floor, not an estimate. The true share is higher, probably by most of that 5% to 13%.

It also reconciles two figures that look contradictory. Our expense analysis found that 41.0% of organizations under $250,000 report payroll cost, while only 29.1% report a headcount here. The difference is 11.9 points, and the payroll-without-headcount rate for that band is 12.7%. The two fields disagree by almost exactly the amount the error rate predicts, which is the outcome you want from a consistency check.

Using this for a hiring decision

Three ways to read the tables against your own situation.

Budget per head. Take your total expenses and divide by the number of people you employ or plan to. Compare it with the medians above: $80,645 at $1m to $5m, $88,210 at $5m to $25m. A number far below your band’s means part time roles, and it is worth being explicit about that in a budget rather than discovering it in a job advert.

The first six figure hire. If you are under $1m and considering paying someone six figures, 91% of organizations your size do not. That is not a prohibition, and there are good reasons to be the exception, but it should be a decision the board takes deliberately rather than drifts into.

Whether you can afford the post at all. Cost per employee here includes payroll tax and benefits, which is the number that catches people out. At the sector median, $31,956 of cash compensation carries $5,166 of other employment cost, so budget roughly 16% on top of any salary you advertise. It is higher if you offer health cover.

Whatever the tables say, the constraint is not the benchmark, it is whether the revenue is recurring. 41% of organizations spent more than they received in the year we measured, and a post funded from a one off grant becomes a redundancy in eighteen months.

Method and limits

Source is the IRS Statistics of Income annual extract of tax-exempt organization financial data, Form 990, processing year 2024. It is a free public file covering every processed return, so it is a census rather than a survey. The analysis script is published and every figure here can be reproduced from it.

The population is 501(c)(3) organizations with total functional expenses of $25,000 or more, which leaves 256,539 organizations.

Four limits. The headcount is a W-2 count and not a full time equivalent, as set out above, which is the binding limitation on this page. Organizations filing Form 990-EZ or 990-N are absent, so the smallest charities are underrepresented. The extract carries no state or sector, so none of this can be cut by geography or by field, and a New York hospital and a rural arts council sit in the same row. And processing year is not tax year, so a minority of returns cover earlier periods.

One thing this data cannot tell you at all: what a specific job pays. For that, the individual salaries of the highest paid staff are listed by name on Part VII of any organization’s own Form 990, which is a public document. Comparing three or four organizations that genuinely look like yours will beat any sector median, including ours.