What’s in a Number? Ask India

On the first day of September, India's Ministry of Statistics and Programme Implementation reported that the economy had grown 7.8 percent in the April-to-June quarter, beating every forecast on the street, including the Reserve Bank of India's own call of 7 percent. 

Two days later, a former finance secretary went on national television and said the number was 2.6 percent. A day after that, in a separate interview, he moved it again, this time to somewhere "around 5 percent." In between, the country's first chief statistician told a newspaper he wasn't sure the headline figure "passes the smell test," and an executive director at the World Bank went on air and called the entire argument "ill-educated."

Nobody was disputing that India's economy is growing. What nobody could agree on, for about a week, was how fast, compared to what, and using whose numbers should be relied upon.

Whose Number is it Anyway?

The disagreement traces back to a decision Indian statisticians made seven months earlier and largely off the front pages. 

Rebasing gross domestic product calculations to 2022-23 from 2011-12, and, in the same stroke, changing the price index used to strip inflation out of the raw numbers and the mathematical method for doing it. 

Rebasing itself is routine. Every major economy does it every five to ten years, swapping out an aging basket of goods and services for one that reflects how people actually spend money now. 

What is not routine is doing three things at once (new base year, new deflator, new deflation method) and then getting hit, this month, with a quarter of growth so strong that critics like Subhash Garg assumed something in the official machinery had been tuned to produce it.

Mr. Garg gave that suspicion its clearest, most public voice. 

Garg was India's finance secretary, India’s highest-ranking official in its finance department, for five months in 2019 and has kept a habit, since leaving government, of picking holes with official data. On Indian news station NDTV, pressed on whether he was accusing the government of manipulating numbers to make growth look better, he didn't back off.

"I am saying exactly that," he said. His calculations ran like this. 

Under the old GDP series, the government had reported India's nominal GDP for the first quarter of the previous fiscal year, 2025-26, at roughly 86 lakh crore rupees or about $910.14 billion. Under the new series, that same quarter was revised down to about 80 lakh crore. 

Compare this year's nominal GDP of roughly 88 lakh crore against the original 86 lakh crore figure, Garg argued, and growth comes to 2.6 percent, not 7.8 percent. Compare it against the revised, lower base of 80 lakh crore instead, and the growth rate jumps to 10.3 percent nominal, feeding into the headline 7.8 percent real number the government published.

"What has made the government revise the last year's current prices data so drastically?" Garg asked, in the same interview. He went sector by sector. In his calculations, using last year's original government numbers rather than the revised ones, manufacturing actually contracted, and consumption fell. Asked what policy should follow, he didn't vacillate. "Recognize the reality of the Indian economic situation, that there is no real growth," he said, "then take the right reform measures."

What has made the government revise the last year's current prices data so drastically? - Subhash Garg, ex-finance secretary, Government of India

The government's response came fast, and from the top of the statistics bureaucracy.

Saurabh Garg, the secretary in charge of MoSPI, India's official statistician, and no relation to Subhash Chandra Garg, called the manipulation charge "highly unfortunate" and walked through, patiently, why he thought the comparison itself was broken.

"Whenever we compare year to year, we always compare on constant prices, because current prices include price inflation," he said. "The second issue is whenever we do comparison, we need to compare like to like. We can't compare apples to oranges." 

His argument? 

The old series had already become obsolete the moment the new one was published, back in February. Using last year's now-superseded figure as a yardstick for this year's new-series number, he said, wasn't a workaround. It was comparing two different rulers and pretending they measured the same thing. He noted that the lower revised base for 2025-26 was made public in February, months before the Q1 growth print that triggered the row, not slipped in afterward to flatter this year's number, as critics suggested.

For evidence that didn't depend on anyone trusting his deflators, Saurabh Garg pointed to production volumes that don't need adjusting for prices at all. Automobile output, cement, steel, all up somewhere between 12 and 16 percent year on year. "This statement to make that there has been some data fudging done is extremely unfortunate," he said, noting that his ministry had published 64 separate data tables alongside the release.

This statement to make that there has been some data fudging done is extremely unfortunate - Saurabh Garg, Secretary of Ministry of Statistics and Programme Implementation of India

Neelkanth Mishra, India's executive director at the World Bank, was less patient. Asked directly whether the 7.8 percent figure was accurate, he clarified that he'd called the 2.6 percent claim itself "ill-educated," not the man who made it. 

"That was not my allegation," he said, of the personal dig, before repeating the same core objection as the MoSPI secretary. You cannot line up an old series against a new one and call the difference acceptable to professional statisticians, and economists.

Every five years or so, he said, fast-growing economies with big informal sectors need to reset their statistical base, because the previous basket goes stale. India hadn't done it for a decade. "So now we have reset it. When you do that, you find new information, and therefore overall GDP numbers get revised up or down."

Mishra also made an argument that had nothing to do with methodology. Forget the deflators, he suggested. Look at what people are actually buying.

Personal vehicle shipments were up 35 percent year over year in August, even though vehicle exports had grown only 9 percent, meaning the demand was domestic. Two-wheeler sales were up more than 20 percent. Commercial vehicle dispatches were up more than 40 percent. Tax collections were accelerating. Credit growth was outrunning expectations. None of that, he said, depends on anyone's opinion about whether the new PPI-based deflator was built correctly.

Sanjeev Sanyal, a member of the prime minister's Economic Advisory Council to the Prime Minister (EAC-PM), defended the timing of the whole exercise. 

Statisticians typically rebase at the start of a decade, he said, but India's decade opened in 2020 and 2021, the Covid years, which made those years useless as a baseline. 

"We had to wait till 2024 before we got a typical year that we could use as a base year," he said. Once that happened, dying sectors fell out of the basket and faster-growing ones came in, which mechanically lifts the growth rate. "This is precisely what has happened," Sanyal said. "I don't think anybody should complain. We did exactly what the IMF and others were asking us to do."

When Politics Sits Firmly in the Driver's Seat

Two of India's more prominent economists, whose politics don't usually line up, ended up on roughly the same side of the technical argument, if not the same side of how urgent the concerns were. 

Surjit Bhalla, a former IMF executive director, said he'd looked for evidence that the numbers had been politically massaged and found none. 

"Absolutely no politics," he said, describing the statisticians who built the new series as "some of the most conservative economists and statisticians that I've met." Montek Singh Ahluwalia, the former deputy chairman of the Planning Commission, was more circumspect, cautioning against reading too much into a single quarter and saying the size of the downward revision to last year's base deserved a closer look. 

It was a smaller, more careful version of the same discomfort Garg had turned into a headline.

The person who came closest to splitting the difference, and who probably has the most standing to do it, is Pronab Sen, India's first chief statistician and a former secretary at MoSPI himself. 

In a telephone interview, Sen said he didn't put much stock in Garg's specific critique.

The convention, he said, has always been to compare the latest estimate against the latest revised one, not the original. He also thought the 10 percent nominal growth for the quarter was probably close to right.

His concern sat one level deeper, in the price deflators used to convert that nominal growth into a real number. 

MoSPI's new series uses "double deflation" for manufacturing, a technique that requires separate price data for a company's output and for everything that went into making it, a far more data-hungry approach than the old single-deflation method, which needed prices only for finished goods. 

"I have not seen any evidence that they have that data," Sen said. He pointed to one figure that had already drawn attention on its own. Real manufacturing growth of 9.2 percent for the quarter, alongside a manufacturing deflator of negative 1.5 percent, meaning the price of manufactured goods was recorded as falling even as output surged. "It can happen with double deflation," Sen said. "With single deflation, it can't happen." He wasn't calling it wrong. He was saying he had no way to check it, because the government hadn't published the input-price data the calculation depends on.

I have not seen any evidence that they have that data - Pronab Sen, ex-secretary, MoSPI

He raised a second concern about the switch itself, from the Wholesale Price Index to a Producer Price Index that India has never used for this purpose before and that, by MoSPI's own account, is still at an experimental stage. 

Producer prices can't be gathered the way wholesale prices are, by taking quotes in a market. A producer has to hand them over directly, and for decades, Indian companies had refused, citing trade secrecy. 

"I don't know whether they've been able to collect this data," Sen said. "And if they have, I don't know how they've done it, unless Indian companies have become very cooperative." 

His preferred fix wasn't to reject the new index.

It was to have run the old and new series side by side for a period first, so outsiders could see where and why they diverged, rather than retiring one and introducing the other in the same breath. "What you're saying, in effect, is, 'Trust me,'" he said. "But trust is only built when you can compare."

But trust is only built when you can compare - Pronab Sen, ex-secretary, MoSPI

MoSPI has said it will publish its full "sources and methods" documentation within two weeks. 

Sen thought that should have happened when the new base year itself was released, back in February, rather than after a public fight had already broken out over numbers nobody outside the ministry can independently reconstruct. 

In Britain and Australia, he noted, statistical agencies release enough granular supply-and-use data that outside researchers can rebuild most of the GDP calculation themselves. 

A Leap of Faith

India doesn't, and hasn't for years. 

It once compiled a detailed Supply and Use Table that would have made exactly that kind of independent check possible. The ministry stopped producing it. "Not only should they not have stopped," Sen said, "but they should have actually expanded it."

The dispute didn't stay contained to television panels and interview transcripts for long.

Within days it had become a live argument between the Congress party, India’s main opposition party and the ruling Bharatiya Janata Party, each side treating a rebased deflator as a proxy for something much larger. Whether the government's growth story could be trusted at all.

By Sept. 2, Garg himself seemed to have walked down from his initial criticism.

Speaking to the Indian journalist Ashutosh, who goes by one name, he set aside both 7.8 percent and his own 2.6 percent and landed somewhere in the middle. 

"Growth of around 5 percent is definitely expected. Any growth beyond that is being talked about is doubtful." He offered no bridge between that figure and the net-GDP-growth-of-zero claim he had been making days earlier, and at least one critic online asked the obvious question. By what method had he now arrived at 5 percent, having argued gross growth was 2.6 and net growth was nothing at all?

Maybe that's the real lesson sitting underneath the deflator argument. 

A government statistics office switched two variables at once and asked the public to take the new number on faith. A former finance secretary who built his critique on comparing two incompatible series ended up, within seventy-two hours, unable to hold his own number steady either. 

Somewhere between them, the one person with the technical standing to referee the fight wasn't defending either side. He was asking, simply, to see the underlying data. The input prices, the producer index, the supply-use tables. That way the next argument about India's growth rate could be settled with facts and arithmetic instead of conviction.

As of this week, nobody outside MoSPI has seen it yet.

The author is an Executive Director and Head of Research and Analysis at Icarus Asia, a Hong Kong-based risk and advisory firm that specializes in macroeconomic and statistical research. Views are his own and not that of the employer.

Sources

Garg, Subhash Chandra. Television interview on India's Q1 FY27 GDP data. NDTV, September 2026.

"Ex-IMF economist Surjit Bhalla backs 7.8% GDP number: 'No evidence of exaggeration.'" Business Today, September 4, 2026. https://www.businesstoday.in/latest/economy/story/7-8-gdp-row-absolutely-no-politics-in-gdp-numbers-no-evidence-of-exaggeration-says-surjit-bhalla-553394-2026-09-04

"No serious economist will complain': PM's advisor Sanjeev Sanyal defends India's 7.8% GDP growth." Business Today, September 2, 2026. https://www.businesstoday.in/latest/economy/story/no-serious-economist-will-complain-pms-advisor-sanjeev-sanyal-defends-indias-7-8-gdp-growth-552809-2026-09-02

ET Now. "MoSPI Rejects Claims Of GDP Data Manipulation, Explains Constant Price Comparison Method." ET Now, September 2026. https://www.etnownews.com/news/mospi-rejects-claims-of-gdp-data-manipulation-explains-constant-price-comparison-method-article-156032202

IANS. "World Bank's Neelkanth Mishra lashes out at India GDP critics as 'ill-educated and egregiously wrong.'" The Hans India, September 3, 2026. https://www.thehansindia.com/news/cities/new-delhi/world-banks-neelkanth-mishra-lashes-out-at-india-gdp-critics-as-ill-educated-and-egregiously-wrong-1117435

Maitreya, Abhay. "India GDP Growth Hits 7.8% In Q1 FY27 As Economists Question What The Data Misses." IBTimes Singapore, September 1, 2026. https://www.ibtimes.sg/india-gdp-growth-hits-7-8-q1-fy27-economists-question-what-data-misses-93151

Mishra, Asit Ranjan. "Methodology of new GDP series is questionable, says Pronab Sen." Business Standard, September 4, 2026.

NDTV. "'Egregiously Wrong': World Bank's Neelkanth Mishra Slams Critics Of India's GDP Data." NDTV, September 2026. https://www.ndtv.com/business-news/egregiously-wrong-world-banks-neelkanth-mishra-slams-critics-of-indias-gdp-data-12000554/amp/1

Oneindia News. "7.8% GDP Growth or 2.6%? Ex-finance Secretary Subhash Garg Sparks Congress-BJP Row." Oneindia, September 2026. https://www.oneindia.com/business/7-8-gdp-growth-or-2-6-ex-finance-secretary-subhash-garg-sparks-congress-bjp-row-8193665.html

OpIndia Staff. "'India is definitely growing by 5%': Former Finance Secretary Subhash Garg now contradicts his own claim of 2.6% GDP growth." OpIndia, September 3, 2026. https://www.opindia.com/news-updates/india-is-definitely-growing-by-5-former-finance-secretary-subhash-garg-now-contradicts-his-own-claim-of-2-6-gdp-growth/

PTI. "'Ill-educated, egregiously wrong': World Bank ED Neelkanth Mishra on row over GDP data." The Tribune, September 3, 2026. https://www.tribuneindia.com/news/business/ill-educated-egregiously-wrong-world-bank-ed-neelkanth-mishra-on-row-over-gdp-data/