Half the Charity Sector Runs on Government Money. Most Charities See None of It.

The Australian charity sector’s largest funder is the government, and this is not a secret. The regulator prints it. In the 2024 reporting period, of $239 billion that moved through registered charities, $117.9 billion came from government — a shade under half, published by the Australian Charities and Not-for-profits Commission in its own annual report on the sector.

Recomputing it from the underlying returns gives 49.4%: $115.5 billion of $233.6 billion, across 43,148 charities. The difference from the regulator’s figure is 0.1 percentage points, which is what a matching population definition looks like.

So the headline number is settled and it is not the story. The question the aggregate cannot answer is which charities that money reaches — and the answer is that it reaches almost none of them.

The average charity gets nothing#

Of the 43,148 charities that reported positive revenue, 26,905 — 62.4% — recorded no government revenue whatsoever. The median charity’s government share is zero. Half the sector’s money comes from government; more than half of the sector’s organisations never touch it.

Among the 16,243 that do receive some, the distribution has two peaks and a hollow middle. The largest group, 4,719 charities, gets a little: under a fifth of revenue. The second largest, 4,329 charities, is at the other extreme, above 80%. Only 2,127 sit in the 40–60% band that the sector average would suggest is typical.

The dollars run the opposite way to the counts. The 4,719 charities in the least-dependent band hold 1.7% of all government revenue between them. The two most-dependent bands, 7,168 charities, hold 68.9%.

Half the sector’s money comes from government. Sixty-two per cent of its organisations never see any of it.

It is not spread thin#

The concentration was the first thing tested, with the line drawn in advance: the top 100 recipients would take at least half of all government revenue reaching charities.

They take 47.7%. The prediction fails by 2.3 percentage points, which is worth saying plainly because the threshold was set before the number was computed and the number came in under it.

What the curve does show: 100 organisations, 0.62% of the charities that receive any government money, take just under half of it. A thousand take 80.8%. Two and a half thousand take 92.7%.

The single largest recipient is the Victorian Catholic Education Authority, which reported $3.53 billion of government revenue against $3.63 billion of total revenue — 97.3%. The rest of the top ten is Catholic school authorities, universities, a national early-learning provider and archdiocesan trusts.

Size predicts it almost perfectly#

The clearest structure in the data is the plainest. Government funding rises with charity size, and it rises without exception: small charities draw 14.9% of their combined revenue from government, medium charities 33.9%, large charities 50.7%.

The medians say it more sharply than the aggregates. The median small charity gets 0% from government. The median medium charity gets 7.7%. The median large charity gets 50.1% — so the sector-wide average, that tidy half, is very nearly a description of one specific animal: the typical large charity. Four in five large charities receive some government money; fewer than one in four small charities do.

The most government-funded purpose is religion#

The next test predicted a gradient by charitable purpose: charities registered for advancing social or public welfare would be at least 20 percentage points more government-funded than those registered for advancing religion.

The result reverses. Religion-purpose charities draw 56.0% of their combined revenue from government, the highest share of any purpose tested, and 9.1 points above welfare at 46.9%. Environmental charities are lowest at 18.3%.

The explanation is in the overlap, and it took a decomposition to find. Of 7,959 charities registered for advancing religion, only 16.8% receive any government money — a far lower hit rate than welfare’s 44.8%. But 1,015 of them are also registered for advancing education, and those 1,015 receive $12.67 billion, which is 88.7% of all government revenue going to religion-purpose charities, at a 67.9% government share. The 6,944 religion-only charities receive $1.61 billion between them, at 23.6%.

Read plainly: the religion figure is a schools figure. Australia funds religious school systems through the charity structure, those systems register “advancing religion” as a purpose alongside education, and the purpose-level average inherits their scale. The top five religion-purpose recipients are all Catholic school authorities.

Nothing moved#

The fourth failed prediction is the most boring and possibly the most useful. In a matched panel of 40,323 charities reporting positive revenue in both 2023 and 2024, the government share was predicted to rise by at least a percentage point.

It rose 0.34 points, from 49.5% to 49.8%.

The flat line hides traffic. Within that same panel, 1,153 charities crossed from minority to majority government funding and 1,245 crossed the other way. The sector-level stability is two thousand four hundred organisations changing their funding character and almost exactly cancelling out.

A contractor class, and a column that cannot be trusted#

The last test asked whether a distinct group exists: charities that are, in revenue terms, government service providers. Two conditions were set in advance — at least 5,000 charities drawing more than half their revenue from government, and that group holding at least twice its share of the sector’s staff as its share of charity numbers.

There are 8,288 such charities, 19.2% of those analysed. They hold 53.8% of the sector’s revenue and 81.5% of every government dollar that reaches a charity.

The staffing half of that test could not be answered from the data as published. Adding up the full-time-equivalent staff column across the sector gives 3,477,245 FTE. The ACNC’s own report says charities employ about 1.6 million people, roughly 11% of the Australian workforce. A single record accounts for more than half the discrepancy: Murdoch University’s return reports 1,868,177 full-time-equivalent staff on the same form that reports 1,259 full-time, 575 part-time and 2,765 casual employees.

That is the same failure mode as the $54 billion government-revenue figure sitting in the same extract — a number no arithmetic check would catch, because it breaks no identity the form tests.

Screened, the contractor group holds 61.8% of the sector’s paid staff against 19.2% of its organisations — a ratio of 3.21, above the pre-registered 2.0, so the hypothesis holds. It holds on a population defined by a rule invented after looking at the column, and that is a real qualification, not a formality.

What could explain this instead#

“Government funding” here is not one thing, and calling it dependence overstates it. The Annual Information Statement asks for a single combined figure. A public hospital paid per activity, a school funded on enrolments, a disability provider billing the NDIS and a charity handed a discretionary grant all report in the same column. A 97% government share at a school authority describes a funding formula for a public service; it is not the same condition as a small advocacy charity whose grant is up for renewal. Nothing in this data separates them.

The concentration may be an artefact of how charities are structured, not of how money is allocated. Several of the largest recipients are _ACNC Group entities — consolidated returns covering many underlying organisations. A diocese that reports as one group looks like extreme concentration; the same diocese reporting parish by parish would not. The register’s grouping conventions shape the top of the distribution, and no attempt is made here to unwind them.

The U-shape may be a definitional shadow of the size gradient. Large charities are disproportionately government-funded and disproportionately likely to be contracted service providers, so a distribution with a peak near 100% could simply be the large-charity population showing through, not a distinct funding regime. The size and dependence findings are not independent tests of separate claims.

The 62% who receive nothing are mostly very small, and may be measured badly. Small charities face the lightest reporting obligations, and a zero in a revenue component can mean “none” or “not itemised”. Zeros were read here as reported. If some fraction of them are unreported rather than absent, the share of charities receiving nothing is overstated — the direction of that error is knowable, its size is not.

There is a version of the Australian charity sector that matches the public idea of one: 26,905 organisations funded by donations, fees and volunteers, receiving nothing from government at all. There is another version that is 8,288 organisations holding four-fifths of the public money that flows to charities and three-fifths of the paid staff. Both are counted in the same total, described by the same average, and regulated under the same form.