India’s GDP Overhaul Shrank Last Year and Lit a 7.8% Fight

India’s GDP growth printed 7.8% for April-June after the statistics ministry rewrote last year’s rupee total lower under a new 2022-23 series. Real output is put at ₹81.36 lakh crore, up from ₹75.46 lakh crore a year earlier, in the ministry’s quarterly GDP estimates for April-June.

Former finance secretary Subhash Chandra Garg says that cut turned a modest rise into a headline rate. His 2.6% alternative compares two different GDP series, which the ministry calls the error. The political problem is that India rebuilt these accounts to answer an IMF grade of C, then published a print whose base had just been shrunk.

Mixing Two Series Turns 10.3% Into 2.6%

On August 31, the National Statistics Office said nominal GDP, the current-price total, rose 10.3% to ₹88.27 lakh crore from ₹80.00 lakh crore. Real GDP, which strips out price changes, rose 7.8%. Real gross value added rose 8.2% to ₹73.82 lakh crore, and nominal GVA rose 11.5%.

Garg, finance secretary in 2018-19, went on television and compared this year’s ₹88.27 lakh crore with the ₹86.05 lakh crore figure first published for April-June 2025 under the old 2011-12 series. That mix yields about 2.6% in current prices. He has said real growth would then be close to zero, and that a change in method does not explain “what went out of the production.”

The arithmetic of his 2.6% is not in dispute. ₹88.27 lakh crore against ₹86.05 lakh crore is a 2.6% rise. The dispute is whether those two totals measure the same economy. They do not. The ₹88.27 lakh crore print sits in the 2022-23 series. The ₹86.05 lakh crore print does not.

Chief Economic Adviser V. Anantha Nageswaran called that approach “cherry-picking.” He said some quarter numbers get bumped up and some get bumped down when a series is rebuilt, and that users should look at consistency inside one series. Commerce Minister Piyush Goyal put the government’s line more bluntly, calling the 7.8% print a reality.

https://x.com/PiyushGoyal/status/2095378981889675396

A widely shared clip of Garg refusing to retreat is what moved a statistics footnote into a political argument. The 2.6% figure is a mixed-series artefact. The live question underneath it is why last year’s rupee GDP is now so much smaller, and why the implied GDP deflator for this quarter is only 2.5% when wholesale inflation was above 9%.

The Overhaul Meant to Lift India’s IMF Grade

On November 26, 2025, the IMF’s Article IV staff report kept a C rating for national accounts, the second-lowest mark on a four-step scale, while the median grade across India’s data set stayed at B. Staff wrote that the numbers arrive on time and with “broadly adequate granularity,” but that “some methodological weaknesses somewhat hamper surveillance.”

The Fund named an outdated 2011-12 base year, wholesale price indices used as deflators because a producer price index was missing, “excessive use of single deflation,” gaps between the production and expenditure sides, thin coverage of the informal sector, and no seasonally adjusted quarterly accounts. A C grade, in the Fund’s own wording, means the data “have some shortcomings that somewhat hamper surveillance.”

On February 27, 2026, the ministry launched a new GDP series with 2022-23 as the base, the first full reset since 2015. Statistics Secretary Saurabh Garg, who is not related to Subhash Chandra Garg, had said the package would largely answer the Fund: the old base would go, single deflation would be replaced by double deflation or single extrapolation, and supply-use tables would shrink the gap between the two ways of adding up GDP.

Nageswaran has used that reset as a defence of honesty rather than of a juiced rate. He said that if the new series had lifted the level of GDP, say from ₹354 lakh crore to ₹384 lakh crore, critics would have accepted it, because that is what many countries do. “In fact, we are the only country which brought it down,” he said.

Anil Sood, a professor and co-founder of the Institute of Advanced Studies in Complex Choices in Mumbai, made the same point from the other side of the argument. After the new series, he said, previous GDP and GVA totals were revised down, so “the estimated size of the Indian economy was smaller than what was projected under the old series.” A smaller level is a poor way to fake a boom. It is also, as this week showed, an excellent way to lift a year-on-year growth rate if last year’s quarter is the one that shrank.

What Last Year’s GDP Looked Like Before February

The ministry’s September 2 note, issued after Garg’s television rounds, laid out four dates. It said the move from ₹86.05 lakh crore to ₹80.00 lakh crore was the result of successive revisions to the GDP series, and that it is “incorrect to interpret the difference as a deliberate downward revision of last year’s GDP to mechanically increase the current year’s growth rate.”

THE PATH FROM ₹86.05 LAKH CRORE TO ₹80 LAKH CRORE

1. August 29, 2025: Q1 FY2025-26 nominal GDP is first published at ₹86.05 lakh crore under the 2011-12 series.
2. February 27, 2026: The 2022-23 series recasts that same quarter at ₹80.32 lakh crore, a cut of ₹5.73 lakh crore before this year’s print exists.
3. June 5, 2026: Provisional estimates for FY2025-26 nudge the quarter up to ₹80.44 lakh crore as indicators are updated.
4. August 31, 2026: New industrial output and producer-price series pull it to ₹80.00 lakh crore, against ₹88.27 lakh crore for Q1 FY2026-27.

Most of Garg’s ₹6 lakh crore, about $63.5 billion on his round six-trillion-rupee figure, therefore left the books in February, six months before anyone knew this quarter would print at 7.8%. On September 4, the Press Information Bureau restated that same path and said the revisions reflect the new base, updated data, and better measurement, not a last-minute squeeze.

Saurabh Garg has told interviewers to “compare apples to apples”: about ₹74 lakh crore in Q1 FY2024-25, ₹80 lakh crore in Q1 FY2025-26, and ₹88 lakh crore in Q1 FY2026-27, all inside the new series. Inside that yardstick, nominal growth is 10.3% and real growth is 7.8%. Cross the yardstick with the discontinued series, and the 2.6% phantom appears.

Household Spending Trailed a Capex Surge

The composition of the quarter is less contested than the base, and it is less flattering than the headline. Gross fixed capital formation, the investment line, rose 11.9% in real terms after 5.8% a year earlier, more than double. Exports of goods and services rose 12.0% after 6.0%. Private final consumption, still the largest slice of demand, rose 7.1% after 6.8%. Government consumption rose 4.3%.

A finance-ministry backgrounder called this the highest Q1 real GDP growth in the four years from 2023-24 to 2026-27, and it beat the Reserve Bank’s 7.0% call for the quarter. It also cooled from 8.6% in January-March. The investment share of nominal GDP rose to 34.3%.

REAL EXPENDITURE GROWTH, Q1 ON Q1

On the production side, services rose 10.0%, led by financial, real estate, IT and professional services at 12.1%. Manufacturing rose 9.2% after 8.3%. Construction rose 7.7%. Electricity, gas and water rebounded 8.9% after a 1.8% decline. Agriculture rose 3.6%. Mining contracted 2.4% after a 12.4% jump a year earlier, a high-base problem as much as a new slump.

The February rebasing had already rewritten the demand mix in a deeper way. A Keio University note on the 2022-23 benchmark found that household consumption share fell to 57.1% from 61.5% in that year, a drop of about 10% in the nominal consumption level, while the investment share rose from 31.2% to 32.4%. India’s recent growth, on the new books, looks less like a consumption boom and more like a capex cycle than the old series suggested.

Anubhuti Sahay, head of India economic research at Standard Chartered Bank, has said the first-quarter figures are “looking better” mainly because of the better method, and that base revisions did not do “much” of the work. “It is not that the GDP number is only froth,” she said. She also said growth is not even, and that quality jobs and the rural monsoon remain open risks. Jaydeep Mukherjee, a professor of economics at Great Lakes Institute of Management in Chennai, has called the investment-and-export mix “not sustainable” while geopolitical risks stay this high. Morgan Stanley and Citi have both put FY2026-27 growth at 7.3%.

Manufacturing’s Price Gauge Went Negative

A second technical fight sat under Garg’s 2.6%. The implied GDP deflator, nominal growth minus real growth, is 2.5% for the quarter. Consumer price inflation was 3.9%. Wholesale inflation was over 9%. Congress has asked any household running a kitchen whether prices rose only 2.5%. The ministry’s answer is that the deflator is not a kitchen index.

The GDP deflator is the price of value added across the whole economy, including government, investment, exports, and services such as banking and IT. When raw-material prices run hot and some service prices run cold, the average can sit well below both CPI and WPI. That is a coverage point, not a proof that households faced 2.5% inflation.

Manufacturing makes the gap vivid. For the first time in this quarterly series, factory GVA is built with double deflation: output and inputs are price-adjusted separately on the new producer-price index. In April-June, input prices rose faster than output prices in textiles and cotton ginning, basic metals, and rubber and plastics. Nominal manufacturing GVA grew 7.7% while real manufacturing GVA grew 9.2%, so the implicit GVA deflator came in at -1.5%.

WHAT THE 2022-23 SERIES CHANGED

• Base year: The benchmark moved from 2011-12 to 2022-23, a lag the IMF had flagged as too long.
• Factory prices: Manufacturing GVA now uses double deflation on granular producer-price indices instead of a single wholesale deflator on both output and inputs.
• New feeds: GST returns, e-Vahan registrations, the Public Financial Management System, listed-company results, and updated unincorporated-sector and labour surveys enter the accounts.
• Still estimated: Reema Bhattacharya, head of Asia research at Verisk Maplecroft, notes that formal-sector company data still carry a lot of the load, while a large part of the informal economy has to be estimated.

The ministry is due to publish a full “Sources and Methods” volume by the end of September 2026. Until that document is out, critics can keep asking which line of last year’s production vanished, and official economists can keep answering that the vanished rupees were a change of measuring stick, not a change of factories.

Congress Counts ₹43 Lakh Crore Off Four Years

On September 3, Congress communications chief Jairam Ramesh put four questions to Finance Minister Nirmala Sitharaman. He said the April-June 2025 total had been revised “not once but four times” from about ₹86 lakh crore to about ₹80 lakh crore, and that “if you keep shrinking the base you’re comparing against, this year’s number will automatically look much bigger than it really is.” That, he said, “is simple arithmetic, not economic growth.”

https://x.com/Jairam_Ramesh/status/2095459781129728240

Ramesh also said the new series cut nominal GDP in almost every quarter since FY2022-23, by ₹8 lakh crore to ₹12 lakh crore a year, for a cumulative ₹43 lakh crore ($455 billion). The implication, in the party’s telling, is that goods and services of that value had been counted and have now been removed. The government has not published a four-year rupee reconciliation that matches that running total line by line.

If you had not revised last year’s GDP, the growth in current prices would have been only 2.6%.

Subhash Chandra Garg, former finance secretary, in a television interview

Prime Minister Narendra Modi called 7.8% a “herculean feat” and said the country had delivered it “despite oil price shocks and supply chain issues.” Sitharaman credited “the people of India and their hard work.” Those lines treat the print as a score. Garg and Ramesh treat the denominator as the story. Both can be true at once only if the 2022-23 series is used on both years, which is what the ministry did, and only if users accept a 2.5% GDP deflator next to 9% wholesale inflation, which many will not.

Cement, Steel and Credit Still Match the Print

The ministry’s own high-frequency annex is the nearest thing to an outside check that does not depend on last year’s rewritten total. Cement output rose 8.9% in the quarter. Finished steel consumption rose 8.3%. Commercial-vehicle sales rose 18.3%, and three-wheeler sales rose 29.7%. Electricity generation, on the industrial-output index, rose 9.3%. Port cargo rose 6.2%. Mining on that index fell 1.2%, and natural-gas use fell 2.6%, in line with the GVA contraction in mining.

CHECKS OUTSIDE THE GDP TOTAL

• July industry: Industrial production rose 6.7% in July 2026, and 6.3% in April-July, after 4.0% a year earlier.
• Bank credit: Credit to industry rose 20.0% in July, credit to services 22.9%, and credit to agriculture 17.0%.
• Trade: Combined merchandise and services exports in April-July were $316.42 billion, up 13.16% from $279.63 billion a year earlier.
• Annual revisions: Real GDP growth was revised up, not down, for three full years, to 7.3% in 2023-24, 7.2% in 2024-25, and 7.8% in 2025-26.

Those volume and credit series will not settle a fight about the rupee level of last year’s GDP. They do make it harder to treat 7.8% as a number with no factories behind it. S&P Global Ratings, in August 2026, kept India’s sovereign rating at BBB/A-2 with a stable outlook, after lifting the long-term grade to BBB in 2025.

The next quarterly GDP release, for July-September, is scheduled for November 30, 2026, and the ministry has already warned that estimates will move again as source agencies revise their inputs. Garg’s missing ₹6.05 lakh crore will still be missing on that date, because it left a discontinued series. What will not have left is the reason the fight caught: India rebuilt the books to look more trustworthy, then asked the country to celebrate a rate whose base had just been written down.