Nobody to borrow from

Thailand feeds a schoolchild on about ฿24 a day.

The Cabinet approved ฿2,955 million for lunches — and this article previously said what that money was for, and was wrong.

It is not the national school lunch budget. It is a specific allocation for pupils in Mathayom 1 to 3 — lower secondary, roughly ages 12 to 15 — at 7,344 “opportunity-expansion” schools, the ones that carry pupils beyond primary. It covers 575,983 students, and it ran from 1 December 2024 to 30 September 2025.

That works out at about ฿5,130 per pupil for the period, or roughly twenty-five baht a day across a school year of about two hundred days.

The rate runs on a sliding scale, ฿22 to ฿36 a head per day, with the lowest rate applying to schools of 120 pupils or more and smaller schools receiving more per child — a sensible piece of design acknowledging that a kitchen cooking for sixty cannot buy like a kitchen cooking for six hundred.

Now set that against the number this site established four years ago.

Article 13 What a Bangkok street food cart actually earns put a plate of cooked street food at ฿50.

The budget for a child’s lunch is under half the price of the cheapest cooked meal in Thailand.

The concession, which is real

I’m not going to run that comparison without dismantling it first, because it isn’t like-for-like and pretending otherwise would be dishonest.

A street vendor’s ฿50 contains things a school kitchen doesn’t pay for. Her own income. Her pitch fee. Her gas, her transport, her losses on what doesn’t sell. She’s buying for one cart; a school buys at wholesale volume from the same distribution layer article 31 Where the ฿50 plate actually comes from described, and it cooks in bulk.

So ฿24 buys more food in a school kitchen than ฿24 buys on a street. That is genuinely true and anyone quoting this article should quote this paragraph with it.

It is still ฿24, for a cooked meal with protein in it, for a growing child, in a year when article 97 Cheaper than mending recorded factory closures outpacing openings and article 100 Borrowing to eat put household debt at ฿16.41 trillion. Thai coverage has been blunter than English coverage: one assessment of the increased allocation was headlined, roughly, a bigger budget, and the quality still lags.

What happened in 2018

In May 2018, photographs circulated on Thai social media of kindergarten pupils being served ขนมจีน — fermented rice noodles — with fish sauce. That was the meal.

The school was in Surat Thani. Investigators found falsified purchase documents: paperwork recording the acquisition of regular, nutritious supplies that had not been bought.

The former director was convicted on 77 counts of corruption, each carrying five years. The Supreme Court confirmed a term of 20 years and six months.

That was not an isolated case. Thai courts have handed down a 385-year term in one school lunch prosecution, 50 years in another, 25 years in a third.

Thailand punishes this ferociously, and keeps having to. Which tells you something important: enforcement is not the binding constraint. A country handing out three-figure sentences and still generating cases every few years does not have a deterrence problem.

The thing article 92 didn’t see

Article 92 The first toll found a universal uniform subsidy of about ฿400 against a uniform costing ฿1,500 to ฿3,000, and concluded that universal is not enough — it has to be adequate too. A benefit set below the cost of the thing it’s for produces the same outcome as no benefit, by a different route.

This is that finding again, in a different ministry. But there’s a second mechanism here that the uniform piece doesn’t have, and I think it’s the more important one.

When a budget is set below the cost of the thing, the gap gets closed by someone.

For uniforms, the closer is the parent. 27% of Thai parents reported borrowing (from relatives, pawnshops, informal lenders) to keep a child in school. Painful, and visible, and countable.

For lunch there is nobody to borrow from. A seven-year-old cannot take out a loan against the difference between ฿24 and what a meal costs. The gap doesn’t get closed. It gets eaten — as a smaller portion, or less protein, or noodles.

And that produces the consequence nobody states:

An underfunded programme makes theft from it invisible.

If ฿24 legitimately buys a thin lunch, a thin lunch is not evidence of anything. It looks exactly like compliance. A parent seeing a poor meal cannot distinguish a school doing its honest best on an impossible budget from a school stealing, and neither can an auditor, without going to the paperwork — which is precisely what was falsified.

Fund it properly and a bad lunch becomes evidence. That is the argument for adequacy that has nothing to do with generosity: a budget with slack in it is a budget you can audit by looking at the plate.

Whether it works — and it does

The programme deserves its due, because a piece that only reports the failures misdescribes it.

School feeding in Thailand is universal, free, and it reaches everyone in kindergarten and primary. Research finds that increased real spending on school lunches significantly reduces the likelihood of children being underweight, stunted or wasted. The money does what money is supposed to do.

Stunting among Thai under-fives is around 13.4% — down over time, and low by regional standards.

But the national figure conceals the thing this site keeps finding. In Ranong the stunting rate is around 26% — double the national average. In the southernmost provinces it runs near 20%. And UNICEF has reported that one in ten children in Thailand live in severe child food poverty.

A universal programme, and the outcomes are still twice as bad in the places with the fewest alternatives. Article 89’s pattern, arriving through a system that was designed to prevent exactly that.

The figure I can’t reconcile

Reported programme scale does not agree with itself and I’d rather say so than pick.

One set of figures describes 7,344 schools: 6,855 under the Office of the Basic Education Commission, 376 under local administration, 109 under the Bangkok Metropolitan Administration, and four run by the Border Patrol Police. Another describes the programme reaching around 30,000 schools.

That conflict is now resolved, and the answer is that they were never the same programme. The 7,344 is the count of opportunity-expansion schools in the lower-secondary allocation described above. The ~30,000 belongs to the long-standing universal programme for kindergarten and primary, which is a separate and much larger budget.

They were being read as two estimates of one thing. They are one figure each from two different things, and this article was among the places making that mistake.

⚠️ What this article still cannot give you is the universal programme’s own budget and per-head rate. Those are the numbers a reader actually wants, and the figures reached here describe the smaller, newer extension instead. That the two are so easily confused — and that the confusion survived into a published article about exactly this subject — is itself the point about how legible a national feeding programme is.

What would actually change it

Raise the rate to the cost of a meal, and index it. Not as generosity, as the precondition for the programme being auditable at all.

Publish the per-school allocation and the menu. Article 80’s finding, applied here: a published standard and a published price is what makes an informal system safe, and it is exactly what would make this one checkable by the people best placed to check it, who are parents.

Weight it towards the provinces with the worst outcomes. Ranong at 26% against a national 13.4% is not a national problem needing a national rate.

And stop treating prosecution as the answer. A 385-year sentence is a statement about how seriously Thailand takes this. It is not a system for making sure a child eats, and the cases keep coming.

Common misconceptions

“Thailand doesn’t feed its schoolchildren.” It runs a free, universal school lunch programme for kindergarten and primary, and separately allocated ฿2,955 million in 2025 to extend lunches to lower-secondary pupils at 7,344 opportunity-expansion schools.

“฿24 is comparable to a ฿50 street plate.” Not directly — schools buy at wholesale and cook in bulk, and the vendor’s price includes her income and her pitch. The gap is smaller than it looks and it is still a gap.

“The corruption is the problem.” The corruption is a problem. The budget being below the cost of a meal is what makes the corruption hard to see.

“Sentences that long must be deterring it.” Terms of 20, 25, 50 and 385 years have been handed down, and cases continue. Deterrence is not the binding constraint.

“Stunting is low, so nutrition is fine.” Nationally around 13.4%. In Ranong around 26%, and in the far south near 20%.