When a national statistics office changes the year it measures growth against, it changes every comparison that runs across the change. This is normal, necessary and periodically disruptive. India’s national accounts were rebased to 2022-23, and in the days after the June-quarter release on 31 August 2026, the disruption arrived.
The mechanics are worth setting out because the dispute is not really about the growth rate. A base year is the period whose price and production structure defines the weights used to add up an economy. Rebasing updates those weights to reflect what the economy now produces — which matters enormously in a country where the share of e-commerce, digital services and new manufacturing has changed quickly since the old base. Every rebase therefore involves revisions to past estimates as well, because the historical series is recomputed on the new weights. India’s previous base was 2011-12.
The argument concerns what those revisions did to the level. The current-price GDP for the June quarter of 2025 had earlier been estimated at roughly ₹86 lakh crore. Under the new base it is about ₹80 lakh crore. Real growth of 7.8 per cent for June 2026 is measured against that lower figure. Compare against the older, higher level, and nominal growth looks far weaker than the headline implies. Both comparisons use numbers from the same release.
SBI Research made the counter-argument: that comparing a new-base estimate with an old-base estimate is not a like-for-like comparison, and that doing so produces a number that is simply wrong. On its reading the correct comparison gives nominal growth near 9.7 per cent and real growth of about 7.4 per cent. It also pointed out that revisions are routine — 25 upward and 12 downward between 2022-23 and 2024-25 — and that the June 2026 figure will not be final until February 2029.
The deeper issue is that both sides are arguing from a statistical artefact that no revision can remove: there is no single true growth rate, only a rate measured against a chosen base. A rebase that captures a changed economy more accurately will also, by construction, make the years on either side of it harder to compare. That is not a defect of the exercise; it is what the exercise does.
What follows for a reader is not a verdict but a method. When a growth figure is quoted across a base-year change, the questions to ask are which base each number uses, whether the comparison is like-for-like, and whether the estimate is provisional — because for the most recent quarters it always is, and will be revised for years. The 7.8 per cent is real. So is the reason it can be disputed. Anyone reporting one without the other is describing a number rather than an economy.
