In June, at an advocacy visit, an organizer from a parent advocacy group walked us through the subsidy waiting list.

I went in expecting the size of the list to be the story. It wasn’t. What stopped me was that the list was shorter than the conversation around it implied — and that there were open, funded seats sitting on the other side of it at the same time.

Families waiting. Seats empty. Same city, same month, both true.

I have been arguing for expansion for two years. That was the first moment I wondered whether I had been arguing for the right thing.

The assumption underneath everything I had written

In February I published The Same Question, Eight Years Later, and the argument was about matching: children waiting on one side, capacity sitting unused on the other, providers outside the subsidy system, families who cannot see what exists. I have written some version of that four times now. Every time, the gap was a visibility failure, and the fix was better information and more supply.

That framing has an assumption inside it that I never examined, because it never surfaced long enough to be examined.

I was treating the problem as an access gap — families who need care and cannot get any, because there is not enough of it and they cannot afford what exists.

If that is the diagnosis, expansion is not a policy choice. It is arithmetic. More seats, smaller gap.

The logic is so clean that nobody in the room ever has to say it out loud, which is exactly what makes it hard to notice you have assumed it.

Two gaps wearing the same clothes

There is a second kind of gap, and it produces almost identical evidence.

A fit gap is families who could get a seat but not one that works. Wrong hours for the shift they work. Wrong language. Wrong side of the city on a bus route that doesn’t run. Wrong approach to how a small child should spend a day, measured against how this family wants their child raised. Wrong room for a child who needs more than a typical room can give.

An access gap is a constraint of money and supply. A fit gap is a constraint of match. They look the same from the outside — a family without care — and they respond to completely different instruments.

Because if a band is saturated — open slots sitting against a short eligible list — then adding supply to it does not reach anyone new. It intensifies competition among programs already fighting over a fixed pool.

Expansion is the right instrument for an access gap and the wrong instrument for a fit gap. Nobody in that room could tell which one we had. Including me.

Why we could not tell

Both gaps produce the same numbers, and the numbers we collect were built to see only one of them.

Enrollment is the clearest example. A full program is a full program. But there are at least four different reasons a family is still on your roster this morning, and your enrollment count treats them identically.

Sort them on two axes — whether the program actually fits the family, and whether the family has a real alternative they could act on — and the picture stops being flat:

One of those axes belongs to the program. The other belongs to policy.

Today almost every program in this city lives in the left-hand column, and inside that column Fortunate and Captive are indistinguishable. Both read as enrolled. Both fill a seat. Both show up in the count that gets reported upward as evidence the system is working.

Universal funding is the event that separates them — and it separates them after the fact, once the families have already gone.

San Francisco is already running this experiment

This is the part I would want anyone skeptical to sit with, because it is not a prediction.

In one income band, the money constraint is already gone. Head Start-eligible families can enroll essentially anywhere they like, including license-exempt care. The tier above it reaches nearly everything else. Several different state contract types sit at that same income line, drawing from the same pool of families.

And by my read from the provider side, that band has largely hit its corners. There is no meaningful population left inside it that is going without care and simply waiting to be found.

Which means, inside that band, programs no longer have what they used to have without ever naming it: a captive advantage. Families are there because those programs were where the funding worked. That is not the same as being chosen, and until recently nothing forced anyone to know the difference.

Universal childcare will not introduce competition to San Francisco. It will extend, to every band, a condition one band is already living in.

Money was doing two jobs

I think this is the thing the sector has not priced in.

Money was paying for care. Money was also holding families still. When care is unaffordable and subsidy is scarce, a family’s choice is mostly an assignment — they go where the funding works, and programs receive families they never had to win.

Remove the money constraint and the second job disappears with the first. Nobody planned for that, because nobody was measuring it.

Universal funding does not only add families to the system. It releases the ones already in it.

Funding does not flatten need. It differentiates it.

The nearest thing we have to a preview is special education. The entitlement is real, it is funded, and families still fight — for the right placement, the right expertise, a room that can actually hold their child. Funding resolved the question of cost. It did not manufacture the supply that satisfies the need, and it did not make every provider the right provider.

I expect the same shape here. Once cost stops deciding, families will choose on hours, language, philosophy, distance, trust, and what they want a childhood to feel like. Those preferences were always there. Scarcity just made them unaffordable to act on.

The moment cost stops deciding, everything cost was suppressing starts deciding instead — and that is not a smaller problem than the one we are solving. It is a more specialized one.

What this asks of programs

Here is the uncomfortable implication, and I want to state it carefully, because the cheap version of it is bad advice.

In a captive market, quality can stay tacit. You never had to articulate what you do, because families had nowhere else to go, and the funding arrived either way. In a market where families can act, tacit quality is invisible quality — and invisible quality does not get chosen.

So programs will increasingly be paid not for what they offer, but for what they can communicate that they offer. That is not a cynical observation and it is not an argument for marketing spend. Saying clearly what you actually do is the precondition for being findable by the family you are genuinely right for.

But it does carry a risk that I do not think anyone is holding yet. The providers least equipped to explain themselves in public are often small family child care homes, and providers who do not operate in English.

Left to run on its own, universal funding will quietly redistribute children toward whoever explains themselves best rather than whoever serves them best. That is not a reason to oppose it. It is a reason to build the visibility layer before the money lands, not after.

What I would say differently now

I am not arguing against expansion.

I am arguing against expansion as a reflex — reached for because it was the only instrument we had a unit for. Seats are countable. Fit is not, yet. So the conversation organizes itself around the countable thing, and everyone in the room mistakes that for a judgment.

If I had understood the composition of that waiting list a year earlier, I would not have shown up with a different answer. I would have shown up with a different question. Not how many more seats but which families are actually unserved, what would serve them, and would more seats reach them at all.

The useful part is that most of that is answerable without waiting for data. A director can sit down tonight and ask: which of my families would still be here if money stopped deciding? What do we do well that we have never once had to say out loud? Who is this program actually for?

Planners can ask the same thing one level up: is the constraint we are funding access or fit, and what would we look at to tell them apart?

Nobody has that data today. That is precisely why the framing has to be argued now — because by the time the numbers arrive, the question they answer will already have been chosen.

---

If you work on mixed-delivery planning, subsidy design, or slot allocation anywhere in California, and this maps onto something you are seeing from where you sit, I would like to compare notes. I am not selling anything and I do not have this figured out. I am a provider and a parent who spent two years confident about an answer, and got interested in the question a little later than I would have liked.