Qualified Contributions vs. Operating Gifts: The Two-Gift Structure That Funds an SGO

Every SGO business plan reaches the same sentence and stalls: who pays for the staff?

The arithmetic that produces the stall is simple. At least 90% of what sits in a Section 25F segregated account has to go out as scholarships, which leaves at most 10% to run the organization on. Raise $200,000 in a state and the entire operating budget for that state is $20,000 — software, audit, insurance, payment processing, and whatever fraction of a person administers the program. Most organizations do the multiplication, conclude the program only works at scale, and put the plan down.

The conclusion is wrong, and it is wrong because the 10% is not the only money available. It is only the money that comes out of the accounts.

There are two entirely separate gift instruments in this program. Nearly every conversation collapses them into one, and the collapse is what makes an SGO look unfundable.

The Two Instruments, Side by Side

Read the last row twice, because it is the whole point. Under the safe harbor Treasury previewed in June 2026, income for the 90% test is measured by what the segregated account holds. Money that never enters the account is not in the denominator. Spending it does not move the ratio, does not consume the 10%, and does not reduce by one dollar what reaches students.

An SGO funded entirely on qualified contributions has a $20,000 budget on $200,000 raised. The same SGO, with $60,000 in operating support from sources that never touch the accounts, has an $80,000 budget — and still sends 90% of every qualified contribution to students. Nothing was taken from anyone. A second instrument was used.

Why Overhead Belongs Outside the Accounts

Three reasons, in ascending order of how much they matter.

The allowance is small and already spoken for. Payment processing alone can claim a quarter of it on card rails — the fee arithmetic is its own post — before audit, insurance, or software. There is not room in 10% for a salary at most realistic scales.

The cap binds per state account, so thin states cannot carry themselves. Because the 90/10 test is a withdrawal cap rather than an expense rule, a state holding $80,000 can release $8,000 and not a dollar more, however much the organization spends serving it. A big state cannot subsidize a small one through the accounts. Operating money has no such geography.

Year one is structurally broken, and only outside money fixes it. Formation, IRS recognition, state listing, systems, and the first campaign are all spent before the first qualified contribution arrives. A partial first year cannot absorb them inside 10% of a number that is still near zero. Notice 2025-70 asked whether the regulations should provide start-up relief or multi-year smoothing; that question is still on the open list. Model as though the answer is no, and the launch has to be funded from operating money — because it does.

There is also a fundraising consequence worth naming. Any campaign that does not return better than 10:1 in the same state cannot be paid for from the accounts at all. Put donor acquisition on the operating side and that constraint disappears — you are spending money that was never in the denominator.

Making the Ask: One Donor, Two Gifts

The instinctive objection is that asking the same person twice is a harder ask. In practice it is an easier one, because the two gifts feel completely different to the person writing them.

The qualified contribution is close to free. A donor who owes federal tax gives $1,700, claims a $1,700 credit, and is out nothing — the gift redirects tax they were going to pay anyway. That is not a generosity conversation; it is a redirection conversation, and it converts at rates ordinary fundraising does not.

The operating gift is the real gift. It costs the donor money. But it is asked after they have already seen their scholarship dollars cost them nothing, which is the most favorable moment a development office is ever going to get.

The sequence that works is the obvious one: lead with the credit, because it is the remarkable thing and it is free. Then make the smaller, honest ask — the credit sends money to students, and something has to keep the lights on so it can keep happening. A donor who has just been shown a $1,700 gift that costs them nothing is unusually receptive to a $250 gift that costs them $250.

Two things to keep straight in the ask. The credit is non-refundable, so a donor with no federal tax liability gets nothing from the qualified contribution this year, though it carries forward five years. And the operating gift is deductible only if the donor itemizes, which most do not. Neither is a reason to skip either ask — they are reasons not to promise an outcome you cannot deliver.

The Part That Breaks: Receipting and Books

This is where organizations get into trouble, and it is entirely avoidable if the separation is built before the first gift rather than reconstructed afterward.

Two receipts, two formats. A qualified contribution gets a Section 25F acknowledgment carrying the IRS-method unique donor number the credit is matched against. An operating gift gets an ordinary Section 170 contemporaneous written acknowledgment. They are not interchangeable, and a Section 170 letter will not support a credit claim. If your system can only produce one kind of receipt, it cannot run a two-gift program.

Two ledgers, never commingled. The statute requires the SGO to prevent commingling of qualified contributions with other amounts by maintaining separate accounts used exclusively for them. Operating gifts must land somewhere else from the moment they arrive — not be swept later.

The intake has to capture intent at the moment of the gift. Whether a gift is a qualified contribution or operating support is a decision the donor makes when giving, not one an administrator makes at month end. A giving form that cannot ask the question produces gifts nobody can correctly receipt.

Do not let a rejected gift silently become the wrong instrument. Someone will attempt an appreciated-stock gift or a donor-advised fund grant to the scholarship account. Neither can be a qualified contribution — the credit runs to the individual taxpayer, and the contribution must be cash. The right handling is to route it to operating support and say so plainly. The wrong handling is to accept it into the segregated account, which puts a non-qualifying amount in the denominator of the 90% test.

That last point has an upside most organizations miss. The DAF balances and appreciated securities that are useless for the credit are perfectly good operating money — and they are frequently the largest gifts a donor is capable of making.

Where Operating Money Actually Comes From

Individual operating gifts are one source and rarely the largest. The others, roughly in order of how reliable they are:

• A parent or affiliated organization. If the SGO was formed as a dedicated affiliate — which the safe harbor strongly rewards — the parent's dues, reserves, and program revenue can support it without ever entering the accounts. For an association or a diocese, this is usually the answer.
• Foundation grants for operations. Ordinary grantmaking. A funder who will not underwrite scholarships directly will often underwrite the infrastructure that moves them.
• Corporate sponsorship. Subject to the usual unrelated-business-income care around substantial return benefits, but conventional.
• Partner-school fees, where the model includes them. Worth naming precisely: a partner fee is paid from the operating allowance, not from the 90%. It is a use of the 10%, not a second source outside it.

Four Cautions

"Largely scholarship-granting" is undefined, and the safe harbor depends on it. The safe harbor is available to organizations whose activities are largely scholarship-granting. Treasury has not said what "largely" means. A two-gift structure does not endanger that on its own — raising operating money to run a scholarship program is scholarship-granting activity — but an organization that grows a large non-scholarship program alongside it is betting on a definition that does not exist yet.

No quid pro quo, in either direction. An operating gift cannot buy anything, and it especially cannot buy influence over awards. A donor who gives operating support does not thereby get a say in who receives a scholarship, and nothing about the two-gift structure loosens the earmarking prohibition.

Operating donors can still become disqualified persons. Substantial-contributor status is measured against the organization, and it is not obvious that operating gifts are excluded from that calculation. A major operating donor may therefore acquire disqualified-person status that reaches their own family's eligibility for awards. Ask counsel before soliciting a very large operating gift from a family with children in your schools.

Everything here rests on a preview, not a regulation. The safe harbor that makes the denominator argument work was previewed on June 9, 2026. Proposed regulations are due by the end of September 2026. The structure described here is defensible under the previewed rules and would survive most plausible refinements, but nothing in this program is settled yet.

The Takeaway

An SGO that funds itself only from the 10% is a small organization by construction, and its first year is close to impossible. An SGO that runs two gift instruments — qualified contributions that become scholarships, operating support that pays for the machine that delivers them — has an ordinary nonprofit budget attached to an extraordinary fundraising offer.

The structural work is not difficult, but it has to be done before the first gift: two receipt formats, two ledgers, an intake that captures intent, and a development plan that asks for both. If you are still deciding whether to form the entity at all, the formation guide covers the sequence; if you are deciding what has to be running on day one, the software has to do both or you will be reconstructing records under audit.