Perils of comparing GDP from different base years

Why in the News

The Ministry of Statistics and Programme Implementation (MoSPI) has released output data for the first quarter of 2026-27, showing gross domestic product (GDP) growth of 7.8 per cent in real terms and 10.3 per cent in nominal terms. A former Finance Secretary alleged that the corresponding quarter of the previous year had been revised down to produce a flattering comparison, and computed nominal growth of only 2.6 per cent. That computation takes its numerator from the new 2022-23 base year series and its denominator from the discontinued 2011-12 series. The two series rest on different data sources and a different methodology, so no growth rate can be derived across them. The live question is no longer the arithmetic. It is the size of the downward revision the new base produced, and whether the statistical ministry will account for it rupee by rupee.

What does a base year revision do?

1. The base year anchors the price comparison: A base year is the reference year whose prices are used to strip inflation out of output, so that real growth measures volume rather than price change.
2. Revision is routine and was overdue: Every economy revises its base year, normally once in about five years. The absence of a revision was itself a reason India’s GDP was losing credibility.
3. It is an opportunity to rebuild the estimate: A revision lets the government bring in new data sources, improve methodology and capture an economy that has changed since the last base.
4. It changes real GDP measurement first: Nominal GDP is measured at current prices, so a change of base year does not by itself explain a fall in the nominal series.

What did the first quarter data show?

1. Growth beat the expectation set at the start of the quarter: Most economists expected about 7.5 per cent for April to June. The official figure came in at 7.8 per cent in real terms.
2. The quarter opened in the middle of a war: The West Asia conflict was disrupting output across the world, and India’s heavy dependence on West Asian energy imports was expected to slow growth further.
3. The world did not contract either: The International Monetary Fund (IMF) expects world growth of 3.0 per cent in 2026 against 2.9 per cent in the previous year, so an economy withstanding the shock is not by itself anomalous.

Why is the 2.6 per cent claim invalid?

1. The rollback happened before the war, not after the result: The new series was unveiled on 27 February 2026, one day before the United States went to war with Iran. Nominal GDP for the first quarter of 2025-26 was rolled down that day from Rs 86.1 trillion on the old series to Rs 80.3 trillion on the new one.
2. Later revisions were marginal: The same quarter was estimated at Rs 80.4 trillion in June and Rs 80.0 trillion on 31 August, against Rs 88.3 trillion for the first quarter of 2026-27.
3. The sequence rules out reverse engineering: The base was rolled down six months before the current quarter’s number existed, so the previous year’s figure was not cut to flatter it.
4. The same method produces an absurd result on real GDP: Applied to the real series, mixing the old denominator with the new numerator implies growth of almost 70 per cent in the quarter.

What question does the revision genuinely leave open?

1. The first half of 2025-26 lost about Rs 11 lakh crore: Nominal GDP for the first two quarters fell from Rs 171.30 lakh crore on the old series to roughly Rs 160 lakh crore on the new one, a cut of about 6.5 per cent concentrated in those two quarters.
2. There is nothing left to reconcile against: The old series was discontinued before comparable third and fourth quarter estimates for 2025-26 were published, so no complete old series year exists to match quarter by quarter.
3. The demand is for a reconciliation bridge: The revision should be broken down in rupees into revised source data, changed sectoral coverage, methodological changes, revised taxes and subsidies, and changed price indices and deflators, for GVA as well as for GDP.
4. The long run picture is comparable: Nominal GDP rose about 32.8 per cent under the old series and 32.3 per cent under the new one over 2022-23 to 2025-26, and cumulative real growth is broadly similar.
5. A downward revision is not lost output: The economy did not shrink by Rs 11 lakh crore. Better data can move a historical estimate down.
6. The annual number moved too: Nominal GDP for 2025-26 was revised from Rs 357 trillion on the old series to Rs 345 trillion on the new one.

Challenges to India’s national income estimation

1. Informality is estimated rather than counted: A large share of output comes from unregistered units that no annual return captures, so their contribution is inferred from proxies. Eg. The unincorporated sector is covered by a sample survey, and its output after the 2020 lockdown was derived from indicators rather than enumerated.
   The Fix: Link the enterprise surveys to Goods and Services Tax and Udyam registration data to build a live frame for small units.
2. Deflators historically overstated value addition: Single deflation applies one price index to output without separately deflating inputs, so a squeeze on firms’ margins is recorded as extra production. Eg. Manufacturing GVA in the 2011-12 series was criticised for a decade on exactly this ground.
   The Fix: The 2022-23 series abolished single deflation, and producer price indices published from June 2026 must now be extended to services.
3. No back series accompanies the new base: Users cannot compare the new estimates with earlier decades without a consistent recomputed history. Eg. The back series produced for the 2011-12 base was itself contested and withdrawn from circulation.
   The Fix: Publish a full recomputed back series alongside the new base rather than after a lag.
4. Credibility is contested politically rather than statistically: Each release is judged as a verdict on the government instead of as an estimate with a stated method, which crowds out technical scrutiny. Eg. The IMF has previously raised issues with India’s national income estimates.
   The Fix: Restore a fixed publication calendar for the National Statistical Commission’s own review reports, so scrutiny is institutional rather than episodic.

Conclusion

The methodological point is settled and the credibility point is not. A series can be more accurate than the one it replaced and still be harder to interrogate, because the comparison the public used to make has been withdrawn. Confidence in official statistics is built by letting an independent reader reproduce the numbers, not by asserting that the method was correct. The larger unresolved problem sits behind the estimate: output is growing fast and is not generating enough good quality jobs, which is how a demographic dividend turns into a demographic burden.

Matching Previous Year Question

“[2021, GS3, 10 marks] Explain the difference between computing methodology of India’s Gross Domestic Product(GDP) before the year 2015 and after the year 2015.”