In the 1990s, India began feeding children in government schools.

Decades later, researchers found that the girls who ate those meals went on to have children who were taller for their age than the children of women who never got them.

This came out of a 2021 study in Nature Communications, which tracked Indian mothers and children across national surveys from 1993 to 2016. Children of women who had the full run of mid-day meals in primary school were measurably taller for their age, by 0.40 standard deviations, than children of women who never got them.

In a country where child stunting has been one of the hardest problems to solve, that is a serious number. The same study estimated that the scheme may have accounted for 13 to 32 percent of India’s improvement in child stunting between 2006 and 2016.

Most of that effect would have shown up decades after the meal was eaten, which is exactly the kind of return a budget has no way of recording.

Now compare that to the bill.

PM POSHAN, the scheme most of us still call the mid-day meal, feeds nearly 11.2 crore children in government and government-aided schools. For primary school children, the meal is meant to provide 450 calories and 12 grams of protein.

These are the numbers the state can hold on to: how many children, how much grain, what the cooking costs, what the Centre pays, what the state pays, what the year's allocation comes to.

A file can tell you what the plate costs today.

It has no way of recording what happens when a girl stays in school two years longer and her child is born healthier a decade later.

A year later, another group led by H.P.S. Sachdev challenged the study in the same journal.

They were not saying school meals are useless. Their argument was narrower: the intergenerational effect may not be as clear as the first paper made it sound. 

Because girls who received meals may also have come from households or districts that were already doing better in other ways. Those advantages could have affected the children regardless.

The first team replied and defended its method.

The disagreement still sits there, which is frustrating but also revealing.

Which is why the cost ends up being easier to argue about than the benefit. 

One comes with receipts. The other comes with a list of things researchers still cannot rule out.

Frédéric Bastiat wrote about this problem in 1850, in an essay called That Which Is Seen, and That Which Is Not Seen.

The mid-day meal has the reverse of the usual problem. Usually the hidden thing is the cost. Here it may be the return.

What is visible is the plate, the budget line, the cook's honorarium, the fuel bill. What is not visible is the girl who did not drop out, the marriage that happened four years later than it might have, the pregnancy that went better, the child born less stunted…

That difference decides things in a budget meeting, because a number you can point to always sounds more responsible than a return you have to take on trust, even when there is real evidence behind it.

India launched the national school feeding programme in 1995, but for years it did not become the cooked meal most of us now imagine. Many states were still giving dry rations, and coverage was weak. The Nature paper notes that in 1999, only 6 percent of girls aged 6 to 10 received mid-day meals in school.

Then the Supreme Court stepped in.

In November 2001, in the right-to-food case, the Court directed states and Union Territories to provide cooked mid-day meals in government and government-aided primary schools. The official PM POSHAN page still traces the cooked meal to that order.

That is worth sitting with for a second.

One of the country’s strongest long-term investments had to be pushed into schools by a court order.

Maybe that is what hidden returns need.

You see the same pattern outside school meals too.

Prevention loses because nobody sees the thing that did not happen.

A new hospital wing has a photo. A case of malnutrition avoided does not.

A new school building has a plaque. A girl who did not drop out because lunch made school worth attending does not.

This is not only a government problem. I have made the same mistake in business.

I have cut onboarding because it looked like overhead, then watched customers churn later for reasons that began in the first week. I have treated team rituals like wasted time, then wondered why people stopped speaking honestly. At the time of the decision, the cost was right in front of me. The return was somewhere else.

That is usually how we make the mistake.

The thing that saves us later looks expensive today.

The thing that hurts us later often looks like savings today.

And then there is the cook.

Before the first child reaches school, she has usually been there for hours. In many places, the work means handling firewood, vessels, grains, vegetables, water and smoke before the cooking even starts. Some cooks have said that collecting firewood is part of the job. Some also help grow vegetables.

Then she cooks for the children, serves the meal, and washes the vessels after everyone has eaten.

For this, the Centre pays ₹1,000 a month. After a pension deduction, ₹810 may reach her hand. The Centre has not raised its share of the honorarium since December 2009.

Officially, she is a cook-cum-helper. Often, she is treated like a volunteer.

That word is convenient. It lets the country depend on her work without pricing it like work.

And maybe she is the clearest part of the whole story.

We can count what she costs. We have no way of counting what she helps make possible.

Because the chain is hard to prove, her work remains easy to underpay.

A video did the rounds while I was writing this edition.

Officials confirmed both, and the suspensions came quickly, but only after the videos travelled. One of those incidents had happened months earlier and nothing moved until the clip surfaced.

That is a different problem from the one I am writing about, and it deserves a proper edition of its own.

Next week.

Maybe Chakrabarti is right, and the mid-day meal helped change a generation.

Maybe Sachdev is right, and the long-term effect is not that clean.

Either way, treating the return as zero because it is hard to calculate would be a mistake.

Some things do not give you a neat number when you need one.

The school lunch may be one of those things. So may the woman who cooks it.

The plate costs ₹6.78.

That may be the least important true thing about it.

Hit reply and tell me: what is something you cut, or nearly cut, because you could not see what it was quietly returning?

I read every email.

Until next week,
Ritesh

P.S. The court order was the emergency fix. The lasting fix came in 2013, when the meal became part of the National Food Security Act. The surest way to protect a return nobody can see clearly is to stop making it ask for permission every year.