GDP - Gross Denominator Politics

A decent democracy does not retire dissent when a public official leaves office. The government car may have to be returned, but not the licence to question government, policies, data or motives.

It should, however, leave the footnotes intact.

The latest GDP controversy began with an ex-top finance ministry bureaucrat arguing that India’s nominal GDP had grown by only about 2.6% in the April–June quarter, rather than the reported 10.3%. By the time the television clip had completed one lap of social media, 2.6% had crossed over from nominal GDP to “real GDP growth”, on the strength of repeated forwarding.

The calculation producing 2.6% is wonderfully simple. Take the latest nominal GDP figure of ₹88.27 trillion. Compare it with last year’s figure of ₹86.05 trillion. Press the division key. Announce that the economy has barely grown.

The arithmetic is correct. The comparison is not.

The calculator is blameless; it will divide any two numbers you enter. It is not trained to conduct background checks.

The problem is that ₹88.27 trillion belongs to the new GDP series, with a base year of 2022–23. The ₹86.05 trillion figure belongs to the old series, based on 2011–12. Both numbers are official. They are simply not comparable.

Under the new series, nominal GDP in April–June 2025 was ₹80.00 trillion. Nominal GDP in April–June 2026 was ₹88.27 trillion. This produced a reported nominal GDP growth of 10.3%. Real GDP, after adjusting for prices, grew by an officially estimated 7.8%.

The ₹80 trillion figure was also not conjured up on August 31. It had already appeared in the new series in February as ₹80.32 trillion, when April-June 2025 was estimated.

It would be an unusual statistical conspiracy: the supposedly incriminating revision had been publicly available for roughly six months.

The previous new-series estimate, released in June, was ₹80.44 trillion. Its subsequent revision to ₹80.00 trillion was about 0.6%. Even if the June estimate holds, nominal GDP growth works out to roughly 9.7%.

SBI Research entered the debate with its thesaurus set firmly to combat mode. Its note described the mixed-series calculation in terms including “unsolicited”, “fallacious” and “intellectual sclerosis”.

Within days, one side had accused the government of manufacturing GDP, and the other sent the denominator for neurological examination.

National accounts had become national name-calling.

Behind the rhetoric, however, SBI Research made a useful point. GDP estimates are routinely revised as more complete information becomes available. Revisions are not mechanically a one-way process, and the direction has varied under previous GDP series as well.

The latest April–June estimate is provisional and will pass through several rounds of revision before the estimates for the period are finalised. By then, the GDP number will have accumulated a much fuller statistical record. Many of today’s social-media opinions may also have been revised by then, though probably without a press release.

SBI Research has also argued that, using an alternative treatment of GDP deflator would be about 7.4% rather than the official 7.8%. That provides room for a perfectly legitimate 7.4-versus-7.8 debate. It does not provide a route to 2.6%.

None of this makes the official 7.8% estimate holy scripture.

The government often treats a favourable GDP release like a child’s school report card. It is photographed, enlarged, laminated and forwarded to the entire extended family. Every additional decimal point is introduced to the neighbours.

Serious questions remain about the latest numbers. The unusually low GDP deflator deserves examination. So do the composition of growth, employment generation, household incomes, investment and the distance between aggregate statistics and lived experience. Statistical institutions must explain their methods clearly, particularly when a new base year, revised methodology and new data sources materially alter the estimated level and growth profile of GDP.

Nor should scepticism about official statistics be confused with permission to assemble an unofficial denominator.

Governments should be cross-examined. Statistics should be cross-examined. What should not be crossed are the statistical series.

Criticism may be ideological. Arithmetic must remain non-aligned.

Otherwise, the fastest-growing part of the Indian economy will not be manufacturing, agriculture or services. It will be Gross Denominator Politics, an unregistered sector with no base year, no seasonal adjustment and, judging by recent activity, 100% capacity utilisation.