India's Economy in September 2026: What the 7.8% Actually Measures

Speaking at the centenary of Shri Ram College of Commerce on 5 September 2026, the Prime Minister returned to a stage he had used in 2013, and invited the audience to compare the two moments. "Growth was down on the ground and inflation was sky high," he said of 2013. "The world talked about policy paralysis in India and counted it among the 'fragile five' economies, but now views it as the fastest-growing major economy." Of the latest number: "The 7.8% growth has enthused the country. This comes at a time when a war is on in West Asia."

It is a fair invitation, and this post accepts it. The comparison is run here across every major indicator rather than the four that suit a speech, using the government's own releases. Some of what follows is better than the speech claims. Some of it is not. A little of it is genuinely strange, and the strangeness is in the official tables, not in anyone's model.

Three ground rules. Every figure is attributed to the body that published it, with its release date. Where two official sources diverge, both appear with the basis of each. Political claims are quoted and attributed, never adopted.

Start with what has genuinely changed for the better

Three things are not close, and an honest reading says so before anything else.

Bank balance sheets have been transformed. Gross non-performing assets of scheduled commercial banks stood at 1.8% in March 2026, a multi-decadal low, with capital adequacy at 17.7% and CET1 at 15.3%, both multi-decadal highs (RBI, Financial Stability Report, June 2026). In 2013 the banking system was at the front edge of a bad-loan cycle that would take gross NPAs above 11% by 2018 and consume a decade of policy attention. That cycle is closed. It is the single largest structural improvement in the Indian economy since 2013, and it is a large part of why the system absorbed the shocks of 2026 without a credit event.

The inflation regime is different. CPI inflation was 4.45% in July 2026 (MoSPI, released 12 August 2026). In 2013 it ran near 10%. India now has an inflation-targeting framework, a statutory Monetary Policy Committee, and a policy rate of 5.25%, held unanimously on 5 August 2026. Whatever else is arguable, the price environment of 2013 is not the price environment of today.

The investment rate is high and rising. Gross fixed capital formation grew 11.9% in real terms in Q1 FY27 and stood at 34.3% of GDP at current prices (MoSPI, 31 August 2026). That is a strong number by any historical standard.

Hold those three in mind. They are the load-bearing part of the government's case, and they survive scrutiny.

1. The 7.8% is the first print on a new measuring stick

The number the Prime Minister cited comes from MoSPI's press note of 31 August 2026: real GDP grew 7.8% in Q1 FY2026-27 against 6.9% a year earlier, and nominal GDP grew 10.3%.

What the speech does not mention is that this is an early reading on a new national accounts series. MoSPI released the new series with base year 2022-23 on 27 February 2026, replacing the 2011-12 base. In its clarification of 2 September 2026, the ministry confirmed that the prior year's quarterly level was restated from roughly ₹86 lakh crore to ₹80 lakh crore, and that figures across the two series "cannot be directly compared."

The CPI was rebased in the same cycle, to 2024=100, and the IIP to 2022-23.

This matters for the comparison being invited. A 2013 figure sits on the 2004-05 base, revised onto 2011-12 in 2015, and the present figure sits on 2022-23. The "then versus now" spans two base revisions. That does not make the comparison meaningless, but it does mean nobody should quote a 2013 growth rate and a 2026 growth rate as though they came off the same instrument.

2. Where the 7.8% came from

This is the part worth slowing down for, and every number in it is arithmetic on MoSPI's own Statement 2.

Decomposing real GDP growth into its expenditure contributions:

Computed from MoSPI, Press Note on GDP Estimates for Q1 2026-27, Statement 2, constant 2022-23 prices, 31 August 2026.

Two lines in that table deserve more attention than they have received.

Net exports supplied nearly two-fifths of the headline. Real imports fell 1.1% while nominal imports rose 30.9%. Because imports are subtracted in the expenditure identity, imports falling in real terms adds to measured GDP.

The residual is carrying 3.2 points. The statistical discrepancy swung from +₹1.37 lakh crore to −₹1.06 lakh crore in a single year. The production-side and expenditure-side estimates of the same economy disagree by a large and rapidly moving margin. The 7.8% is a production-side number; the expenditure side closes only because the residual absorbs the gap.

Why real imports fell while the import bill exploded

The gap between real imports (−1.1%) and nominal imports (+30.9%) implies an import deflator of about +32.3%, against an export deflator of about +12.3%. India's terms of trade deteriorated sharply.

The cause is in the annexure to the same press note. Producer prices for crude petroleum and natural gas rose 58.0% year on year in Q1 FY27. Independently: India's crude import bill was $63.4 billion in April-July 2026 against $40.5 billion a year earlier, a rise of 56.5%, on broadly unchanged volumes. Brent traded near $96 a barrel on 4 September 2026, up roughly 47% year on year, having finished Q1 2026 at about $118 after the military action of 28 February and the effective closure of the Strait of Hormuz.

So: the same quantity of oil, at a far higher price.

Here is the point that neither the speech nor most of the criticism makes. When import prices spike, real imports fall, and measured real GDP rises, even though the country is unambiguously worse off. Real GDP measures output, not purchasing power. The measure that captures a terms-of-trade loss is real gross domestic income, and it has not moved the way GDP has.

The Prime Minister presented the 7.8% as an achievement delivered despite the war in West Asia. On the arithmetic of the government's own tables, a meaningful part of it is delivered through that war, as a statistical consequence of paying much more for the same barrels. This is not a scandal and not a manipulation. It is what the national accounts identity does under an oil shock, and it is one of the more useful things a student of the subject can learn from this quarter.

A secondary factor: under the February 2026 trade framework India moved away from discounted Russian crude toward costlier US and other grades. On the evidence of flat volumes against a 47% rise in Brent, the dominant driver is the global price, not the switch.

Two criticisms that do not hold, and one that does

Much of the commentary has attacked the negative manufacturing deflator. Real manufacturing GVA grew 9.2% while nominal grew 7.7%, implying a deflator of about −1.4%, at a time when the producer price index for manufactured products rose 10.7%. This looks impossible. It is not.

The new series adopts double deflation for manufacturing, deflating output and intermediate consumption separately rather than with a single index. That is a methodological improvement economists pressed for over many years. Under double deflation, when input prices rise faster than output prices, the implicit value-added deflator can fall or turn negative. With crude and gas producer prices up 58%, that is precisely the configuration. MoSPI explains this on page 7 of the press note itself, before anyone had objected.

The second weak criticism is that the GST cut inflated the number. It did the opposite. Nominal net indirect taxes fell 0.4%, and net CGST growth slowed to 4.8% from 7.8% a year earlier. Since GDP equals GVA plus net taxes, the tax cut pulled the headline down: real GVA grew 8.2% against real GDP's 7.8%. On the government's own preferred reading, underlying activity was stronger than the number it is being congratulated for.

The criticism that does hold is narrower and harder to answer. The Sources and Methods volume, which documents the new series in full, was "scheduled for its release by September 2026" as of the 31 August press note, having earlier been slated for August. It is not out. Until it is, the double-deflation results, the new producer price index, and the deflator choices cannot be independently audited by anyone outside the ministry. A government asking to be believed on a contested number should publish the method first. That is a process failure rather than evidence of bad faith, and it is entirely fixable.

3. Prices

CPI inflation was 4.45% in July 2026, against 4.38% in June (MoSPI, 12 August 2026, base 2024=100). Beneath the headline:

Source: MoSPI, Press Release of CPI for July 2026.

The headline is close to target and far below 2013. But food is running above it, rural above urban, and inflation is currently accelerating rather than decelerating, with the acceleration concentrated in food and fuel, which is to say in the part of the basket that poorer households cannot substitute away from. June 2026 was the month CPI moved back above the 4% target after sixteen consecutive months below it.

The RBI's own projection, in its Monetary Policy Statement of 5 June 2026, put FY27 CPI inflation at 5.1%, peaking at 5.9% in Q3 (October to December 2026) before easing to 5.4%, with core inflation at 4.7%. At its meeting of 3 to 5 August 2026 the MPC held the repo rate at 5.25% unanimously, kept the stance neutral, and raised its FY27 real GDP growth projection by 10 basis points to 6.7%.

That last figure is worth pausing on. India's central bank, having seen the same data, expects the full year to come in a full percentage point below the quarter the Prime Minister was celebrating. Either growth decelerates sharply from here, or the RBI does not read Q1 as the whole story.

4. Jobs, and Raghuram Rajan's question

The former RBI Governor's objection to the 7.8%, widely reported in the first days of September 2026, was not primarily statistical. He asked why growth this strong is not producing good jobs, private investment or FDI.

The labour data (MoSPI, PLFS Monthly Bulletin, June 2026, Current Weekly Status):

Source: MoSPI, Press Note on PLFS Monthly Bulletin, June 2026, 15 July 2026, Current Weekly Status.

An economy growing at 7.8% is moving its participation rate by 0.2 percentage points a year and its unemployment rate by 0.1. The clearest bright spot is female participation, up 0.7 points to 32.7%, though that remains among the lowest rates of any major economy. This is the disconnect Rajan is pointing at, and it is real. Note also that the monthly PLFS series only began in January 2025, so there is no comparable 2013 monthly benchmark. On this front the honest answer is that we have better labour data than we have ever had, and it is not showing the response a 7.8% economy should produce.

Rural wages complicate it further. Labour Bureau data showed a 17% jump in male rural wages in March 2026, but that reading follows a sampling break, and independent analysts adjusting for the break put underlying growth near 4.3%, which against 4 to 5% CPI implies roughly flat real wages. The official series and the adjusted series tell different stories, and the government has not reconciled them.

One structural change belongs here too. The rural employment guarantee is no longer MGNREGA: the 2026-27 Budget funds VB-G RAM G, the scheme that replaced it, with allocations 42.8% higher than the 2025-26 revised estimate (PRS Legislative Research, Union Budget Analysis 2026-27). A larger allocation is a real commitment. Whether the new scheme's guarantee is as strong as the statutory one it replaced is a separate question, and it will show up in rural wage data before it shows up anywhere else.

5. Investment, and the FDI hole

The domestic investment picture is strong, as noted at the top: real GFCF up 11.9%, and 34.3% of GDP in nominal terms. Central government capital expenditure was budgeted at ₹12,21,821 crore for FY27, an increase of 11.5% over the 2025-26 revised estimate. Within that, capital outlay rose 6.3% to ₹9,43,042 crore while loans and advances rose 33.8% to ₹2,78,780 crore, so a large share of the increase is lending to states rather than direct asset creation by the Centre (Union Budget 2026-27, presented 1 February 2026, as analysed by PRS Legislative Research).

Foreign investment is the weak column, and it is stark:

Gross FDI inflows hit a record $94.84 billion in FY26 against $80.61 billion in FY25. But repatriation by foreign investors reached about $53.58 billion and outward direct investment about $54.04 billion, and between them they consume almost the entire gross inflow.

Both halves of that are true and both should be said. Record gross inflows are a real signal that India remains an attractive place to deploy capital. Near-zero net inflows are a real signal that foreign capital is leaving about as fast as it arrives. FY26's $6.95 billion is a recovery from FY25's near-zero, and still some 75% below FY23.

6. The external account

Exports grew 25.8% in nominal terms in Q1 FY27, which is surprising given that the United States had imposed tariffs of up to 50% on Indian goods through much of the preceding year.

The tariff position changed materially. On 2 February 2026 the US and India announced a trade framework, and the White House fact sheet of 9 February 2026 records the reciprocal tariff on Indian goods lowered from 25% to 18%, with the separate 25% penalty tariff linked to Russian oil purchases removed. Textiles, leather and footwear, plastics, organic chemicals and certain machinery sit at 18%, while generic pharmaceuticals, gems and diamonds, and aircraft parts were exempted. India committed to purchases of over $500 billion of US energy, technology, agricultural and other products, and to curbing Russian oil imports.

The tariff relief is real and the export recovery is consistent with it. The cost side is the oil bill described in section 2. That is the trade that was made, and it is defensible. It is simply not free, and the price of it is showing up inside the growth number as a terms-of-trade loss.

7. The rupee and the reserves: the 2013 playbook, run again

This is where the Prime Minister's chosen comparison becomes most uncomfortable, because the resemblance runs the wrong way.

In 2013, facing a collapsing rupee after the taper tantrum, the RBI under Raghuram Rajan opened special swap windows on 4 September 2013 for FCNR(B) deposits and bank foreign-currency borrowings. They raised about $34 billion in three months.

In 2026, the rupee hit a record low just below 97 to the dollar in May, and traded near 95.2 in early September 2026. In June 2026 the RBI opened a special USD-INR swap facility. As of 21 August 2026, authorised dealer banks reported $72.85 billion of eligible inflows under it: $65.40 billion in FCNR(B) deposits, $4.86 billion in overseas foreign-currency borrowings, and $2.59 billion in external commercial borrowings. The response was strong enough that the RBI advanced the deposit deadline to 31 August from 30 September.

Foreign exchange reserves reached a record $729.33 billion in the week ended 21 August 2026, up about $63 billion in eight weeks.

Both facts belong in the same sentence. Reserves are at an all-time high, and the great majority of the recent build-up is borrowed foreign currency with a repayment date, mobilised through the same emergency instrument used in 2013, at more than twice the 2013 scale. Reserves accumulated from FCNR(B) deposits are a liability, not a surplus. That does not make the policy wrong. In 2013 it worked, and it appears to be working now. It does make "record reserves" a considerably weaker boast than it sounds, and it makes the 2013 contrast harder to draw than the speech suggests.

8. The fisc

Source: Union Budget 2026-27, presented 1 February 2026.

Consolidation is real but slow, and the debt ratio remains well above its 2013 level. Against that, the deficit is being reduced while capital spending rises, which is the harder and the better way to do it.

9. Households

This is the least flattering column, and it comes from the RBI.

Gross household financial assets stood at ₹490.3 lakh crore, or 141.6% of GDP, in March 2026, against gross financial liabilities of ₹158.5 lakh crore, or 45.8% of GDP (RBI Bulletin, August 2026). Over the sixteen quarters from June 2022 to March 2026, liabilities rose by 9.4 percentage points of GDP while assets rose by 7.4. Households are levering up faster than they are accumulating. Non-housing retail loans account for 58.4% of household debt, up from about 50% in 2019-20, which means the borrowing is increasingly for consumption rather than for asset creation.

Net household financial savings did improve, to 7% of gross national disposable income in FY25 from 5.8% a year earlier. Consumption inequality measured on HCES 2023-24 gives a Gini of about 0.29, and household surveys of this kind systematically under-capture the top of the distribution.

Set that against private consumption contributing 3.95 points of growth, and a plausible reading is that a meaningful part of the consumption strength is debt-financed.

10. The scorecard

What to take from it

The government's strongest claims are true. Banks are fixed, inflation has a framework and is inside it, and the investment rate is high. Those are not small things, and no honest account can wave them away.

The government's headline claim is more complicated than the speech allows. The 7.8% is a real number, produced by a defensible and in one respect improved methodology, on a series whose full documentation has not yet been published. About three of its points come from a fall in real imports caused by an oil price shock that has made the country poorer, not richer. Another three points of arithmetic are absorbed by a statistical residual that swung by nearly two and a half lakh crore in a year.

And the parts of the economy that a growth rate is ultimately supposed to describe, which is to say jobs, wages, foreign investment and household balance sheets, are not moving the way 7.8% would imply. That is Rajan's question, and it has not been answered.

In 2013 the Prime Minister told this same college that a glass could be seen as half full, half empty, or full to the brim with water and air. It remains a good metaphor. The task for anyone reading a GDP release is to work out how much of the glass is water and how much is air. This quarter, unusually, the government's own tables let you do the measurement yourself.

Sources and verification

All figures below were retrieved and checked on 6 September 2026.

• MoSPI, Press Note on Quarterly Estimates of GDP for Q1 (April-June) 2026-27, 31 August 2026. Headline growth, sectoral and expenditure tables, the double-deflation note, and the producer price annexure. The growth-contribution decomposition and all implicit deflators in section 2 are computed from Statements 1 to 4 of this document.
• MoSPI, Press Note on New Series of GDP Estimates with Base Year 2022-23, 27 February 2026.
• MoSPI, Press Release of Consumer Price Index on Base 2024=100 for July 2026, 12 August 2026.
• MoSPI, Press Note on the PLFS Monthly Bulletin, June 2026, 15 July 2026.
• RBI, Financial Stability Report, June 2026 (released 30 June 2026), for bank GNPA, CRAR, CET1 and household debt.
• RBI, Monetary Policy Statement, 2026-27, 5 June 2026, for the FY27 inflation path and core projection; and the MPC decision of 3 to 5 August 2026 for the repo rate and the revised FY27 growth projection.
• RBI Bulletin, August 2026, for household financial assets and liabilities.
• The White House, Fact Sheet: The United States and India Announce Historic Trade Deal, 9 February 2026.
• Union Budget 2026-27, presented 1 February 2026, and PRS Legislative Research, Union Budget Analysis 2026-27, for the deficit and debt targets, the expenditure tables and the VB-G RAM G allocation.
• Petroleum Planning and Analysis Cell data on crude import value for April-July 2026, as reported in the financial press, and the Brent price as of 4 September 2026.
• Morgan Stanley's "Fragile Five" note of August 2013 grouped India with Indonesia, Brazil, Turkey and South Africa. The rupee's 2013 record low of 68.85 was set on 28 August 2013. The RBI's FCNR(B) and bank-borrowing swap windows were announced on 4 September 2013 and raised about $34 billion by 30 November 2013.
• Quotations from the Prime Minister's SRCC address are taken from The Indian Express, 5 September 2026.

Noted for correction if superseded. The August 2026 CPI print and the Q2 FY27 GDP estimate (due 30 November 2026) are not yet available. The MoSPI Sources and Methods volume for the 2022-23 series had not been published as of 6 September 2026; when it appears, the deflator discussion in section 2 should be re-examined against it. The rural wage adjustment cited in section 4 is an independent analyst estimate rather than an official figure, and is labelled as such above.