Former NSC Member and Acting Chairman PC Mohanan on Double Deflation and Doubts Over GDP Data

The public debate following the release of the Q1 GDP numbers for 2026–27 has once again brought to the fore the contrast between official data and perceived reality. The understanding of socio-economic changes through evidence has been clouded by the specific political positioning of those who contest the numbers.

The new GDP series, with 2023–24 as the base year, incorporates several technical refinements to bring the compilation methodology closer to internationally recommended practices. However, appreciation of these refinements is constrained by a lack of methodological transparency and access to the relevant datasets.

The use of double deflation in the new series is one such refinement, though it may not play a significant role in estimating quarterly GDP. According to MoSPI, quarterly GDP is mostly estimated by applying benchmark indicators, where “the estimates computed for the previous financial year are extrapolated using the indicators reflecting the performance of various economic and institutional sectors.”

Understand Double Deflation

Double deflation is a method used to calculate gross value added, or GVA, at constant prices. In simple terms, it separately adjusts both the value of what an industry produces and the cost of the inputs it uses for changes in prices. The difference between the two then gives us the real GVA.

The UN System of National Accounts, which sets global standards for GDP estimation, recommends double deflation for calculating the real GVA of an industry. The practice of deflating only the final GVA of a sector, as was done in the past, does not capture changes in the prices of inputs relative to the prices of output. This can lead to a misleading picture of the actual increase in value added.

To apply double deflation, we need a price index to adjust nominal output for changes in prices, as well as an input price index for the raw materials, services and components used in production. This, in turn, requires prices collected from producers, known as the Producer Price Index (PPI) for goods and services. As with any index, individual product prices also need to be given appropriate weights. The general practice is to use weights from the Supply Table of the national accounts, which reflects each product’s share in an industry’s nominal output during the base year.

The same principle applies to inputs. Since an industry uses inputs from different sources, the input price index needs to incorporate PPI and import price indices, as well as information on taxes, transport costs and trade margins. This is necessary to ensure that inputs are valued at the prices actually paid by purchasers. The weights are derived from the Use Table of the Supply and Use Tables (SUT) framework, which provides a breakdown of how much of each product is used by different industries and sectors.

Room for Doubts

The press release issued by MoSPI on September 2, 2026, in an effort to provide more clarity on some of the doubts, gives an example of how double deflation works in practice. A negative inflation rate in the implicit deflator for manufacturing does not mean that manufacturing prices have fallen.

If input prices rise faster than output prices, the difference in their price movements can result in nominal GVA growing more slowly than real GVA. The implicit GVA deflator, which is derived by comparing nominal GVA with real GVA, can therefore show negative inflation even when both output and input prices are rising.

For example, in the latest press release, GVA growth at constant prices for the manufacturing sector in Q1 2026–27 is 9.2 percent over the corresponding quarter of 2025–26, while at current prices, it is only 7.7 percent.

MoSPI has explained why a negative manufacturing deflator is possible under double deflation. But it has not made clear how the benchmark-indicator method, which is normally important for estimating quarterly GDP when detailed data are unavailable, fits into this particular calculation.

There is another issue. A general input index would assume that all sectors use the same mix of inputs. But this may not give a correct picture of real value addition. For example, if energy prices rise, industries that use relatively little energy would be affected differently from energy-intensive industries.

The UN System therefore recommends applying double deflation at the most granular level of industry and product breakdown possible, rather than at the level of total GDP. The need for double deflation was well known all along, but MoSPI did not attempt it earlier because the necessary disaggregated data were not available. This is where the Supply and Use Tables become relevant.

The Supply Table provides information on the availability of goods and services in the economy at the product level. The Use Table, in turn, shows how these goods and services are used. This includes their use as intermediate inputs by industries, private and government consumption, gross capital formation (GFCF, changes in inventories and valuables), and exports.

In the latest SUT published by MoSPI, there are 155 rows corresponding to industry-product groups. For manufacturing, there are 30 activity groups and 75 items across these groups. The Supply Table is generally prepared at basic prices, with details provided to convert these to purchasers’ prices. The latest SUT is for 2023–24. For manufacturing activities, the Annual Survey of Industries (ASI) provides details of major input-output items.

Even when these sources are used, arriving at an aggregate picture of supply and use requires scaling up the survey estimates. This procedure is also not very clear, especially when Ministry of Corporate Affairs’ MCA21 data are the main input for GDP compilation rather than ASI.

Above all, there is little clarity on the price data used to construct input-output deflators. An appropriate and transparent system of price data collection is essential for compiling any index.

Refinements and Perceptions

The much-awaited shift in the base year has given MoSPI a chance to bring GDP estimation closer to international standards and thus address the poor rating given by agencies like the IMF to India’s National Accounts Statistics. Many of these technical refinements, however, are difficult to communicate when the revision leads to significant changes in the numbers and questions are raised by experts. This is especially so when the datasets used in the new series are not accessible for examination by researchers.

It is not easy to appreciate the double deflation technique when the actual methodology is opaque. The desirable practice would have been to allow these exercises to be carried out before the actual base-year revision, bringing greater transparency to the whole process.

Given the use of numbers in an argumentative polity and public discourse, methodological changes and refinements in key indicators should be introduced using tested datasets and after adequate prior consultation. Few may remember that the Indian Association for Research in National Income and Wealth used to be the forum where such methodological refinements were discussed by experts before their implementation. Though the Association still exists, the current political character of macroeconomic data leaves little space for open debate.

The present debate on GDP has very little statistical context in the absence of clarity on the methodology and datasets used. The loud claims about the economy doing wonders based on the GDP growth number disregard the limitations of GDP as an indicator of economic progress.

PC Mohanan is former Member and Acting Chairman of the National Statistical Commission. The views expressed are personal.