Apples, Oranges and a 7% Hole: Why India’s 2.6% GDP Claim Does Not Survive the Arithmetic

India’s 7.8% GDP growth faces scrutiny as the 2.6% claim falls apart, shifting the debate to deflators, rebasing and revisions.

 India’s latest GDP numbers were meant to be a routine statistical release. Instead, they have triggered a fight over a more fundamental question: did India really grow at 7.8% in the first quarter of FY27? The official answer is yes. Real GDP grew 7.8% in April–June 2026, beating the Reserve Bank of India’s 7% projection, while nominal GDP rose 10.3%. Then came the counterclaim. Former Finance Secretary Subhash Chandra Garg put nominal growth at just 2.6% and argued that real growth was close to zero. The figure quickly became political ammunition, with opposition parties embracing it and the Ministry of Statistics and Programme Implementation (MoSPI) dismissing the comparison as statistically meaningless.  
The dispute is not primarily political. It is a question of what a growth rate is allowed to be. A growth rate compares an economy with itself, measured the same way, one year apart. Garg’s ratio compares this year’s new-series nominal GDP with last year’s old-series first estimate. The ratio exists. It is not India’s Q1 FY27 growth rate. Once both quarters sit on the 2022–23 series, the valid comparison remains 10.3% nominal and 7.8% real. High-frequency volume data sit far closer to that print than to a near-zero real expansion. The residual debate that survives is narrower, and more useful: a low implicit GDP deflator, the mechanics of double deflation, and the quality of statistical communication around a large rebase.

What the official accounts say

MoSPI’s 31 August 2026 release, compiled on the 2022–23 base, gives the following levels:

Measure	Q1 FY26 (new series)	Q1 FY27	Growth
Real GDP (₹ lakh crore)	75.46	81.36	7.8%
Nominal GDP (₹ lakh crore)	80.00	88.27	10.3%
Real GVA	68.21	73.82	8.2%
Nominal GVA	72.24	80.53	11.5%

The exact ratios are:

$$g^{real} = \frac{81.36}{75.46}-1 = 7.82\%$$

$$g^{nom} = \frac{88.27}{80.00}-1 = 10.34\%$$

The implied GDP deflator follows from the two growth rates:

$${\pi }^{GDP}=\frac{1+g^{nom}}{1+g^{real}}-1=\frac{1.1034}{1.0782}-1\approx 2.33\mathrm{\%}$$
At constant prices the composition is uneven. The primary sector grew 2.9% (agriculture 3.6%, mining −2.4%). The secondary sector grew 8.6%, with manufacturing at 9.2%. The tertiary sector grew 10.0%, led by financial, real estate, IT and professional services at 12.1%. On the expenditure side, real gross fixed capital formation rose 11.9%, real private final consumption 7.1%, and real exports 12.0%. That is the internally consistent official story.

Reconstructing the 2.6% claim

Garg’s calculation uses this year’s new-series nominal GDP against last year’s old-series first estimate:
$$g^{mix}=\frac{88.27}{86.05}-1=2.58\mathrm{\%}\approx 2.6\mathrm{\%}$$
The ₹86.05 lakh crore figure is the 29 August 2025 estimate of Q1 FY26 nominal GDP under the superseded 2011–12 series. It is not the current comparable base. MoSPI’s path for that same quarter is:

Date	Series	Q1 FY26 nominal GDP (₹ lakh crore)
August 2025	Old series (2011–12 base)	86.05
February 2026	New series, first estimate (2022–23 base)	80.32
June 2026	New series, provisional update	80.44
August 2026	New series, after updated IIP and PPI	80.00

The collapse from 86.05 to about 80 is almost entirely the series change, not an August 2026 haircut designed to inflate this year’s growth. Within the new series, the Q1 FY26 nominal base moved only from 80.32 to 80.00—about 0.4%. Even a comparison of ₹88.27 lakh crore with the June new-series estimate of ₹80.44 lakh crore yields about 9.7%, not 2.6%.
Let $\delta$ be the rebase-level gap between the old-series August 2025 print and the current new-series base:

$$\delta = 1-\frac{80.00}{86.05} \approx 7.03\%$$
Garg’s mixed growth is then an identity, not a discovery:

$$g^{mix} = (1+g^{nom}_{new})(1-\delta)-1$$

$$=1.1034\times 0.9297-1=2.58\mathrm{\%}$$

Any rebase that lowers last year’s level by 7% will mechanically convert 10.3% same-series growth into about 2.6% mixed-series growth. That is true for India in 2026. It would also have been true in 2015, when the 2011–12 series raised levels relative to 2004–05. The direction of the level shift does not make the mixed ratio a growth rate.
This is the standard national-accounts rule. Growth rates are computed inside one series. When the base year, coverage, source data and deflation method change, historical levels are restated so that $Y_t / Y_{t-1}$compares like with like. Mixing $Y^{new}_t$ with $Y^{old}_{t-1}$ is the definition of a series break. Former Chief Economic Adviser Krishnamurthy V. Subramanian put the same point bluntly: taking a nominal number from the old series, dividing it into a nominal number from the new series, and proclaiming 2.6% growth is comparing apples with oranges.
Garg’s further claim that real growth is “close to zero,” and that manufacturing and consumption show negative growth, follows only if one carries the mixed 2.6% into a second invalid step—deflating it with a CPI-like price index of 2–2.5%, or comparing new-series current-year components with old-series prior-year components. Official same-series real private consumption is +7.1%. Manufacturing real GVA is +9.2%. Those are not negative.

Why the level fell when the series changed

A 3–4% downward revision in annual nominal GDP under the 2022–23 series, and a larger gap for that particular old Q1 print, is not mysterious. The new accounts do several things at once.
They move the price and quantity base from 2011–12 to 2022–23, so weights reflect a services-heavier, digitally denser economy. They replace a large block of proxy indicators with GST, e-Vahan, the Public Financial Management System, updated enterprise data and a new IIP/PPI architecture. They replace single deflation in manufacturing, and in parts of agriculture, with double deflation: real GVA equals real output minus real intermediate consumption. They integrate supply-use tables more tightly, which shrinks the production–expenditure discrepancy over time.
When old proxies overstated activity—as MoSPI has itself said they did for several industries—the level of nominal GDP falls even if growth rates inside the new series remain in the mid-to-high 7% real range. That is a correction of the measuring rod, not proof that Q1 FY27 produced almost no additional output. The rebase also changed fiscal talking points. India’s nominal economy is smaller on the new rod than on the old one. Debt, deficit and tax-to-GDP ratios should be restated on the new series. That is a legitimate public-finance consequence. It is not evidence that year-on-year volume growth vanished.

Double deflation and the −1.5% manufacturing deflator

The most technically serious objection is not 2.6%. It is that manufacturing nominal GVA rose 7.7% while real GVA rose 9.2%, implying an implicit GVA deflator of about −1.5%. That looks odd when energy and metal prices jumped after the West Asia shock.
It is not a statement that factory-gate prices fell. Under double deflation:

$$GVA^{n} = O^{n}-IC^{n},\qquad GVA^{r}=O^{r}-IC^{r}$$

$$d_{GVA}=\frac{GVA^{n}}{GVA^{r}}-1$$

If input prices rise faster than output prices, the intermediate-consumption bill inflates faster than the value of output. Nominal value added is squeezed even as physical throughput rises. Then $GVA^{n}$ can grow more slowly than $GVA^{r}$, and $d_{GVA}$ turns negative. For manufacturing:

$$d_{mfg} = \frac{1.077}{1.092}-1 \approx -1.37\%$$ 
MoSPI points to textiles and cotton ginning, basic metals, and rubber and plastics as the segments where that relative-price pattern appeared. OECD work on double deflation has long noted volatile, sometimes negative, implicit GVA deflators around commodity shocks. Mining is the mirror image: real GVA −2.4% against a large nominal jump, because crude, gas and ore prices spiked while volumes slipped. That is what a price shock looks like in the accounts. It is not evidence that the real manufacturing print of +9.2% is fabricated.
The economy-wide 2.3% GDP deflator is still low relative to CPI near 3.9% and especially WPI/PPI. That gap is expected to some degree. The GDP deflator covers services, government consumption and investment, not just the wholesale commodity basket. It remains the weakest official number in the release. If the true implicit inflation were closer to 4%, real growth on the same nominal 10.3% path would be nearer 6% than 7.8%:
$$g_{alt}^{real}=\frac{1.1034}{1.04}-1\approx 6.1\mathrm{\%}$$
That is a different critique from Garg’s, and it is still open. It does not convert 7.8% into zero.

Do the volume indicators look like 2.6%?

Quarterly GDP in India is a benchmark-indicator construct. The right external check is whether physical and tax indicators look like a 2–3% nominal slump or like a 7–8% real expansion.
MoSPI’s own indicator annex for Q1 FY27 versus Q1 FY26 includes cement production +8.9%, finished steel consumption +8.3%, IIP capital goods +15.2%, IIP electrical equipment +27.0%, commercial vehicle sales +18.3%, three-wheeler sales +29.7%, and household vehicle registration +15.9%. Goods-and-services exports at current prices rose 25.8%. Mining and fuel minerals are the clear weak spots, which the GDP print already records.
Independent high-frequency tracks point the same way. Manufacturing and services PMIs remained in expansion. Auto retail and capital-goods production were strong. Core industries for April–June were modest in aggregate, about 3.6%, because coal, crude, gas and refining contracted, while cement, steel, iron ore and electricity held up. Company-level work on Q1 manufacturing points to volumes holding while margins were hit by input costs—exactly the pattern double deflation is designed to show.
None of that looks like an economy whose nominal output rose 2.6% and whose real output was flat. It looks like an economy in which investment and selected manufacturing and services lines carried growth while the energy complex and parts of the primary sector did not. The indicators reject collapse. They do not pin 7.8% to the first decimal. That distinction matters.

What Garg gets right — and what he does not

Garg is right about three institutional facts. First, the new series lowered historical nominal levels, and the public communication of that level shift was poor relative to the political use of the 7.8% headline. Second, quarterly estimates will be revised; treating the first print as scripture is a mistake on every side. Third, a 2.3% GDP deflator next to a commodity-price shock deserves a published item-level reconciliation, not only a press note.
He is wrong that those facts imply 2.6% growth. That number is the rebase gap dressed up as a growth rate. Once the comparison is forced onto one series, nominal growth is 10.3% and real growth is 7.8%. The claim that manufacturing and consumption contracted in real terms does not survive the same-series tables.
A cleaner way to state residual uncertainty is statistical, not conspiratorial. Let official real growth be $g = 7.8\%$. The relevant risk is revision risk inside the same series as IIP, PPI, company accounts and crop data settle—not a five-percentage-point hole created by reaching back to a discontinued base. Historically, Indian Q1 first estimates do move. They do not routinely collapse from about 8% to about 0%.
Intent cannot be inferred from the identity. MoSPI’s timeline shows that the bulk of the level drop is the February rebase, not an August haircut. That kills the story that last year’s GDP was cut last week to manufacture 10.3%. It does not prove the opposite claim: that every methodological choice was innocent of presentation effects. Base-year revisions change levels and the optics of growth. Poor bridge tables created a vacuum that a mixed-series ratio then filled. That is a governance limitation of the statistical system, not evidence of a fabricated quarter.

Limits of this reading

The analysis is strong as a rejection of an invalid estimator. It is only moderately strong as a validation of 7.8%. Q1 GDP is compiled by a benchmark-indicator method, not a full annual census of production. There is no published standard error. A later revision of 50–80 basis points would not resurrect the mixed-series identity. It would still change the economic reading.
The analysis also cannot re-estimate GDP from micro data. Showing that mixing ₹88.27 lakh crore with ₹86.05 lakh crore is invalid does not prove that every new-series level is correctly measured. A full audit would need item-level output and intermediate-consumption series, the exact PPI/IIP mappings used for double deflation, GST-based corporate allocation, and old-versus-new bridge tables for Q1 FY26. Those are not on the public release in usable form.
Composition, welfare and dollar measures were outside the original question. A strong investment-and-services quarter can still be a weak household quarter. Real GDP per capita, employment intensity and dollar GDP—which moves with the rupee—are separate objects. Garg has separately argued that dollar growth looks weaker once depreciation is counted. That is a different metric, not a repair of the 2.6% identity.

Conclusion

The 2.6% figure makes for a powerful headline, but it does not survive the basic test of comparability. It is the product of placing two different statistical yardsticks side by side and treating the resulting gap as economic growth. On a consistent 2022–23 series, Q1 FY27 nominal GDP growth was 10.3% and real GDP growth was 7.8%. The evidence therefore does not support the claim that India’s economy barely grew in April–June 2026.

That, however, is not the same as declaring the 7.8% number beyond question. The unusually low implicit GDP deflator, the negative manufacturing GVA deflator, the effects of double deflation and the substantial changes introduced with the new series all warrant closer scrutiny. The first-quarter estimate will also be revised as more complete data become available. These are legitimate statistical questions—but they are fundamentally different from the invalid exercise of combining two different GDP series to produce a 2.6% growth rate.

The more useful debate now is about the quality and transparency of the new measurement framework. India needs a GDP series that is not only methodologically defensible, but also sufficiently transparent for economists, investors and the public to understand how the numbers are produced. For now, the data point to a quarter of genuine and relatively strong real expansion, supported by investment and services, even as parts of the primary sector remained weak. The 2.6% controversy should therefore be treated not as evidence that India did not grow, but as a reminder that when the statistical yardstick changes, explaining the change can matter almost as much as the number itself.

IndraStra Global is now available on
Apple News, Google News, Feedly, Flipboard, and  WhatsApp Channels.

DISCLAIMER: This article is for informational purposes only and should not be considered investment advice, solicitation, or portfolio management services. However, the company, its affiliates, or contributors may hold financial interests in certain securities or markets discussed herein. Readers should conduct their own research or consult a qualified financial advisor before making any investment decisions.

COPYRIGHT: This article is published under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://creativecommons.org/licenses/by-nc-nd/4.0/

REPUBLISH: Republish our articles online or in print for free if you follow these guidelines. https://www.indrastra.com/p/republish-us.html