New Delhi: Former bureaucrat Ashok Khemka, who came to be known as an anti-corruption whistle-blower in the run-up to the 2014 Lok Sabha polls, has joined the controversy over massive revisions in previous years' GDP figures. Writing in the Indian Express on Friday (September 4), Khemka says the government must be transparent about how it has been revising the GDP estimates.

"It is wrong to calculate current growth by combining an old-series denominator with a new-series numerator. But it is equally unsatisfactory to dismiss questions about a massive revision simply by invoking “change of base year,” he writes.

Close watchers of the GDP figures have pointed out in recent days that if the GDP of the first quarter of the previous financial year had not been revised downwards by Rs 6 lakh crore, the April to June 2026 quarter economic growth rate would have been much lower – perhaps as low as 2%, or even 0%.

But Khemka writes that since last year's revision of the GDP number was due to the base year changing from 2011 to 2022, this adjustment is normal practice.

Base year changes are meant to bring statistics up to date with economic realities, such as new additions to the manufacturing mix or a shift towards services or away from agriculture and so on.

He writes that once the base year is revised to reflect such shifts, the old GDP should not be used to assess how the economy is faring today – including how fast it may be growing.

Khemka refers to this as "a fatal methodological flaw" in the former finance secretary Subhash Garg's estimate that the economy grew by no more than 4–5% in the first quarter of 2026-27, if not lower. The Ministry of Statistics and Programme Implementation (MoSPI) has said the economy grew 7.8% in the first quarter.

Beyond this, Khemka illustrates the same problems with the government's GDP statistics as Garg and several others have, such as former chief statistician and MoSPI secretary, economist Pronab Sen. Khemka, too, says the government's numbers can make sense, but only if the sources are explained.

When 2011-12 was the base year, nominal GDP for the first quarter of 2025-26 was estimated at Rs 86.05 lakh crore. The second quarter in the same year was estimated at Rs 85.25 lakh crore. Under the new 2022-23-base-year series, the first quarter data was revised to roughly Rs 80 lakh crore and the second quarter (July-August 2025) to approximately Rs 80 lakh crore.

Nominal GDP is a measure of total economic activity at a time, whereas the real GDP is what the economy looks like after removing the effects of price change.

Also read: Why India's 7.8% GDP Growth Wonder Is Much Ado About Nothing

"The first-half estimate has fallen by around Rs 11 lakh crore, or about 6.5 per cent," Khemka writes, adding, "This is not a rounding error. The most striking feature is that this enormous revision is concentrated in the first two quarters."

As per Khemka, there is no complete old-series annual estimate against which the new series can be reconciled quarter by quarter. "MoSPI should open that window," he writes.

On its part, MoSPI has explained that the 2022-23 exercise involved much more than simply changing the reference year for future GDP estimates. It says the new series incorporates new data sources, revised coverage, updated administrative data, a new IIP, new PPI, a revised Banking Services Price Index and methodological changes.

This is where Khemka, Garg and others such as Congress MP and former commerce minister Jairam Ramesh agree – apart from the declaration that all these new factors have been taken into account, there is no other information on what the "updated" data used by MoSPI are.

"There may well be legitimate statistical reasons for the revision. But they need to be quantified," Khemka writes.

"MoSPI should publish a simple reconciliation bridge showing, in rupees, how the old estimate became the new estimate, he writes, saying this should be done for the first two quarters of the last financial year.

The public should be told how much of the revision is attributable to revised source data, changes in sectoral coverage, methodological changes, revised estimates of taxes and subsidies, changes in price indices and deflators and other statistical adjustments, he writes.

"That would settle the argument effectively," he writes.

"It is wrong to calculate current growth by combining an old-series denominator with a new-series numerator. But it is equally unsatisfactory to dismiss questions about a massive revision simply by invoking 'change of base year'," he writes.

The Wire is now on WhatsApp. Follow our channel for sharp analysis and opinions on the latest developments.