Two nutrition numbers

Last series was cheques. This one is children.

Same country. Same LinkedIn. Different file.

Every month the government publishes what Anganwadis measured.

Ministries quote it. State dashboards quote it. CSR decks that work in “health and nutrition” quote it without always knowing which thermometer they picked up.

There is another number. The household survey. NFHS-6, released this year: about 29% of children under five are stunted. About 19% are wasted.

On Poshan Tracker in August 2026, the district average in our file is:

• Stunting 25%. Close enough that people blur them.
• Underweight 10%. The survey says 32%.
• Wasting 3%. The survey says 19%.

Stunting is a family argument. Wasting is two countries.

That gap is the series. Not an average that would not budge. Named districts. Months. A map.

What you are looking at

NFHS is a household sample. Children whether or not they walk into an Anganwadi. One round, 2023–24.

Poshan Tracker is the programme’s own census: children on the roll, measured this month, typed into a phone. Forty-one months of it in this file. Thirty-six states and UTs. Seven hundred and thirty-nine districts with a complete history.

One is prevalence. One is what the machine wrote down.

If you average them you get a polite 27% and a useless sentence. If you keep them apart you can see where the dashboard is doing a different job from the survey — and where it is quietly agreeing.

It agrees on the shape. The heavy belt is still central and eastern India. Next week we name 132 districts the dashboard itself still puts in very high risk. West Singhbhum. Nandurbar. Thirty-eight of them in Madhya Pradesh.

It disagrees on acute malnutrition. Researchers flagged that in 2024. NFHS-6 did not close it. Part 5 asks whether Tracker wasting even follows a season. Spoiler the file already whispers: it mostly just falls.

Why this, after CSR

You already know I will sit with a public dump until it produces one stubborn fact.

The last stubborn fact was ₹35 lakh. See CSR Ticket Size.

This one is a place-list, and a 3% that should not be used as 19%.

If you fund schools, anganwadis, or “nutrition” lines — Part 6 of the last run was you — this is the dashboard those rupees sit under. I will not lecture you about row count. I will tell you which districts the machine still paints red, which states moved the wrong way, and which number you should stop mixing into a slide.

This week’s only job

Hold two numbers. Do not pick a side in the first post.

Survey: how stunted, wasted, underweight are India’s children.

Dashboard: what 14 lakh centres recorded in August.

The gap: a finding, not a rounding error.

Next — Part 2: 132 districts . The map.

Data: Part 1 of 5. Poshan Tracker monthly dumps, April 2023–August 2026. Latest month on the public site: August. District averages are unweighted. Tracker percentages are whole numbers as published. NFHS-6 is the 2023–24 household survey. This is programme data, not a second NFHS. Also on LinkedIn.