The key issue in the GDP debate is the use of double deflation and the use of producer/input prices. What new datasets has MoSPI obtained that make these methods feasible?

The transition to double deflation was not sudden. It was built on extensive experiments and detailed analysis using historical time-series data from ASI spanning the period of 2011-12 to 2023-24, Supply and Use Tables (SUT) and the WPI of the 2011-12 series under the guidance of ACNAS. The stability of output PPI from 2022-23 onwards was assessed and MoSPI examined the feasibility of using the output PPI in national accounts prior to its official launch.

The DPIIT had initiated the exercise for compilation of the Output Producer Price Index and hence, the base revision of GDP was an opportune time to introduce double deflation. Since output PPI was not released by February 27, 2026, double deflation was carried out by using available item-level WPI (2011-12 series). Later, on the release of Output PPI in June 2026, following Output PPI-based double deflation, was implemented and updated estimates were released on August 31, 2026.

What are the sector-wise price indices and weights producing the aggregate deflator? Given the big divergence between the GDP deflator and WPI/CPI, can MoSPI provide a reconciliation showing why the GDP deflator behaves so differently?

In the new base year, the deflation strategy in various sectors has been volume extrapolation, single extrapolation and double deflation; use of item-specific price indices has been emphasised in the new base compared to aggregate price indicators applied at the sectoral level.

Since in the old series on applying single deflation (the output and input are deflated by the same deflator), the movement in IPD (Implicit Price Deflator, commonly called the GDP deflator) of National Accounts Statistics was close to the WPI/CPI. However, in the new series, the approaches used in constant price estimation are more sophisticated and hence do not have a one-to-one correspondence with output PPI and CPI.

Under India’s new nationalaccounts methodology, manufacturing uses double deflation: output and intermediate inputs are separately deflated. Therefore, manufacturing GVA’s implicit price change need not equal manufacturing PPI/WPI inflation when the output and input prices vary differently and this fact is addressed in double deflation.

Similarly, CPI reflects price changes of a specific basket curated based on the household consumption of goods and services, whereas GDP by demand-side covers government consumption, investment, export, import too, for which altogether separate price deflators have been adopted.

The GDP deflator is economy-wide: it captures price changes across all domestically produced final goods and services. WPI, by contrast, is a wholesale-price index for goods, with no comprehensive coverage of services. MoSPI explicitly notes that the GDP deflator need not move in line with WPI/PPI/CPI because of differences in their coverage and use of item-level price indices.

Why was the new methodology not run in parallel with the old methodology before becoming the basis for the official series?

Several data sources, such as ASUSE, PLFS and output PPI, apart from methodological changes such as segregation of GVA of multi-activity enterprises, have been introduced in an integrated fashion in the new series. Bifurcation of growth in GVA to the one due to double deflation and that due to new data sources, coverage and other methodological changes is not feasible. Further, releasing two official growths of GDP by running the old and new series concurrently would add to more ambiguity and confusion among the users and, therefore. None of the countries run the parallel series together. India is planning to release the back series for years prior to 2022-23 by the end of this year.

Can MoSPI demonstrate that the improvement is not merely a change in methodology but also an improvement in the quality and availability of the underlying data?

In earlier base series (e.g., 2011–12), the GVA of the unincorporated sector relied heavily on the Enterprise Survey conducted in the base year and the quinquennial round of the Employment Unemployment Survey. In subsequent years, the benchmark/base year estimates were extrapolated using relevant proxy indicators such as growth in sales tax for extrapolating the base year GVA of trade. MoSPI now conducts the Annual Survey of Unincorporated Sector Enterprises (ASUSE) alongside the Periodic Labour Force Survey (PLFS) for labour force estimation on a continuous basis. Thus, the use of proxy indicators is not required.

Then, PFMS for various states were not available in the previous base. GST, which is one of the important administrative data sources for quarterly GDP estimation, was launched in 2017-18. Eventually, the GSTN database has strengthened substantially over a couple of years.

Also, reconciling production-side GVA and expenditure-side GDP often resulted in sizeable statistical discrepancies due to asymmetric data availability. In the new series, the compilation of GDP by production and expenditure approaches has been integrated with the SUT framework. SUT acts as a balancing matrix, ensuring that input-output balances and commodity flow match across all sectors rather than relying on mathematical smoothing or unverified residual balances.

Finally, from ₹86 lakh crore to ₹80 lakh crore is a massive downward revision in the 2026 nominal GDP and that has been being questioned. What do you have to say?

A. Our nominal GDP in the new series has been revised down by about 2.7 per cent in 2022-23, 3.5 per cent in 2023-24 and 3.8 per cent during 2024-25. Such changes in base revision are due to the incorporation of better data sources on its availability and improved methodology.  

In this new series, several methodological improvements have taken place. However, the major factor, taking the levels down, is mainly the new data for estimating the unincorporated sector. Results of ASUSE and PLFS data provide a more direct basis for measurement of unincorporated sector contribution to value added on a regular basis in the new base year.

However, a proper assessment of the sector-wise picture shows revisions have happened both upward and downward. Agriculture and allied activities GVA were revised up by about 3.8–5.9 per cent; financial services and real estate, professional services and ownership of dwellings were revised up by approximately 7.8–9 per cent over the comparable years.

In contrast, trade, hotels and restaurants, and transport and storage saw sharp downward revisions of around 23–26 per cent. Within this category, GVA of trade services was revised downwards by 36 per cent, and road transport by 16.9 per cent, whereas hotels’ and restaurants’ GVA was revised upwards by 5.7 per cent, mainly on account of revised estimates of the unincorporated sector. Since quarterisation is based on annual benchmarks, revision at the annual level have sequentially impacted Q1 and Q2 of FY 2025-26.