On 27 February 2026 the Indian government published a notice called "New Series of Gross Domestic Product (GDP) Estimates with Base Year 2022-23."
Nobody shares that. It is a line of pure jargon with two dates attached.
The rebuild nobody covered
Across nine weeks, three measuring systems were replaced or opened for replacement.
| Date | Release |
|---|
| 13 January | Discussion paper on a chain-based industrial production index |
| 28 January | Discussion paper on seasonal adjustment of that index |
| 12 February | First price index published on a new base year of 2024 |
| 27 February | New national accounts series, base year 2022-23 |
| 16 March | Work begins on revising the industrial production base year |
The base year that was replaced dated from 2011-12. Fifteen years old.
First lesson: the dullest headline is often the biggest change
Anything genuinely structural tends to carry a boring title. These are written by the office that did the work, in the language that office uses.
The exciting releases are the ones written to be read — launches, inaugurations, summits. They are designed to travel. That is what makes them a poor guide to what matters.
So invert your instinct. In a government feed, a headline that looks like homework is worth more of your time than one that looks like news. Methodology, revision, series, framework, base year — those words mark the releases that change the meaning of every number that comes after.
Two spikes, two different causes
Two ministries had enormous months in 2026, and comparing them teaches something.
The electronics ministry published 114 releases in February, against a monthly average of 36 — more than three times normal. Reading the titles, they cluster around one thing: a large international summit on artificial intelligence held in New Delhi, with a declaration attached and many countries signing on.
The chemicals and fertilisers ministry published 89 in March, against an average of 27. Those titles are different in kind: fertiliser supply, hoarding enforcement, licences cancelled, production figures, a pharmaceutical funding announcement.
Same size of spike. Completely different causes. February's was an event with a date. March's was a financial year closing, with enforcement running alongside it.
Second lesson: ask whether a spike is an event or a deadline
An event spike has a venue and a guest list. It rises fast, peaks in days, and leaves nothing behind in the following month.
A deadline spike has no venue. It builds through a month, involves money being sanctioned or rules being enforced, and it recurs at the same point every year.
Second lesson: when you see an office publishing far more than usual, find out which of the two it is before you read anything into it. One is a photograph. The other is a process.
What was and was not there
Covid appears in zero of the year's 4,494 releases.
The word that climbed instead was global, in 182 releases across nine months — already above the 169 it managed in all of 2025.
And in September the coal ministry had its busiest month of the year, 34 releases against an average of about 12.
Third lesson: check that two numbers were measured the same way
This is the lesson 2026 exists to teach, and it comes straight out of that dull February notice.
When the base year of an index changes, the numbers before and after it are not built the same way. A base revision is not a fresh coat of paint on the same method. The notice itself says so, which is why it carries the words "new series" rather than "revised figures".
That is usually an improvement. It is always a break.
Third lesson: before you compare any two numbers, check they were measured the same way. In India, for anything spanning 2026, the answer is often no. That does not make either figure wrong. It makes the comparison a thing you have to handle carefully rather than a thing you can do casually.
---
Do these still work? Test them on last week
Take the five working days to 25 September 2026 — 81 releases, 22 of the 26 ministries publishing, four silent.
The dullest headline. The week's most consequential release was titled "Provisional Estimates of Index of Core Industries for the Month of August 2026" — and it ran on, with the final index for July attached. Homework words, both halves. It is the first official reading of industrial activity each month.
Event or deadline. The coal ministry's seven releases that week were mostly cleanliness campaigns and retrospectives — neither an event with a venue nor a deadline with money attached. Worth checking, so you do not read volume as activity.
Measured the same way. That same week one release reported steel demand across April to August, while another reported the core industries index for August alone. Five months against one. Both true, both counting different spans, and a reader who lines them up carelessly will draw a conclusion neither supports.
One idea to keep
2026 is the year India changed the instrument it measures itself with, and the notice announcing it was written to be ignored.
That is the pattern worth carrying. The things that reshape how you will understand everything afterwards almost never arrive as news. They arrive as administration.
Read the boring ones.