MoSPI Sources and Methods Report: New GDP Series Uses Double Deflation in 28 of 30 Manufacturing Categories

The news

New Delhi. The Ministry of Statistics and Programme Implementation (MoSPI) on Monday, September 21, released ‘Sources and Methods for Compilation of National Accounts Statistics’, the detailed manual behind the new GDP series introduced in February with 2022-23 as the base year, replacing 2011-12. The Indian Express reports that the new series uses ‘double deflation’ in 28 of 30 categories of the manufacturing sector, and that work is on to extend it to the remaining two — processing and preservation of meat, fish, fruit, vegetables, oils and fats; and pharmaceuticals, medicinal chemicals and botanical products — where a high share of imported inputs makes it “challenging” to map inputs to item-level Producer Price Indices. Gross Value Added (GVA) is the value of output minus the value of inputs. To get real GVA, which strips out price changes, double deflation adjusts output and input separately by their own price indices; single deflation divides both by the same index. Earlier, MoSPI used double deflation only for agriculture and for mining and quarrying, relying elsewhere on the Wholesale and Consumer Price Indices — one of the biggest criticisms of Indian data, since input and output prices often move at different rates. The document comes seven months after the series, the shortest gap ever; earlier such manuals took up to three years. It contains no new data and consolidates three reports of sub-committees of the Advisory Committee on National Accounts Statistics published in February. The Economic Times reports that the series makes greater use of corporate filings, limited liability partnership records, GST data and labour surveys, and estimates the household and unincorporated sector directly through the Annual Survey of Unincorporated Sector Enterprises (ASUSE) and the Periodic Labour Force Survey (PLFS). The Indian Express reports that nominal GVA attributed to ‘households’, the proxy for the informal sector, was cut by ₹2.9 lakh crore, or 2.7%, for 2022-23, led by construction, where the household share fell to 59% from 79%; trade and repair, hotels and restaurants, and road transport now look more informal. Former finance secretary Subhash Garg has argued that April-June 2025 nominal GDP was revised down to ₹80 lakh crore from ₹86 lakh crore to make the 7.8% growth of April-June 2026 “look better”; MoSPI says revisions follow the System of National Accounts 2008 and the IMF’s Quarterly National Accounts Manual. The syllabus link is GS3 on growth and its measurement.

Static syllabus linkage

1. GVA and GDP are linked by taxes and subsidies on products. Gross Value Added at basic prices measures the value created by producers: output minus intermediate consumption. GDP at market prices equals GVA at basic prices plus taxes on products minus subsidies on products. Nominal figures are at current prices; real figures are at constant prices of the base year, obtained by deflating with price indices. India has used GVA for sectoral growth and GDP for the headline number since the 2011-12 series was introduced in 2015.
2. The base year is revised to capture a changing economy. A base year is the reference year whose prices and structure are used to compute constant-price estimates. India’s national accounts have moved through several bases, including 2004-05 and 2011-12, and now 2022-23, because old weights misrepresent new industries and new consumption patterns. A base revision also brings in new data sources, which is why past levels of GDP change. The National Statistical Office, created in 2019 by merging the Central Statistics Office and the National Sample Survey Office, compiles the accounts under MoSPI.
3. Single and double deflation give different answers when prices diverge. Under single deflation, nominal GVA is divided by one price index, usually an output price index. Under double deflation, output is deflated by output prices and inputs by input prices, and real GVA is the difference. When input prices, for example of oil, fall faster than output prices, single deflation can overstate real growth, and vice versa. The United Nations’ System of National Accounts 2008 recommends double deflation as the preferred method.
4. The informal sector is measured by surveys, not filings. The household sector in the national accounts includes unincorporated enterprises such as small shops, workshops and self-employed workers, and serves as the proxy for the informal economy. Because these units do not file company accounts, their output is estimated from surveys such as the Annual Survey of Unincorporated Sector Enterprises and the Periodic Labour Force Survey, both conducted by the National Statistical Office. The institutional sectors in Indian accounts are the public sector, the private corporate sector and the household sector.

Why UPSC loves this

1. GS3 names growth and its measurement. Mains questions have asked about the controversy over India’s GDP numbers after the 2015 revision and the credibility of official statistics. A question on the 2022-23 series, its methods and its critics is highly likely, and this document provides the answer’s factual core.
2. Prelims asks definitions from national accounts. UPSC regularly tests the difference between GDP and GVA, factor cost and market prices, nominal and real, and the institutions that compile data. Double deflation and the base year are natural statement-based questions.
3. Data credibility is an ethics and governance issue too. The independence and transparency of statistical systems has come up in the context of delayed survey releases and the 2019 creation of the NSO. Publishing methods quickly is a governance improvement an answer can cite.

Analysis

1. Double deflation answers the most serious technical criticism of the old series. The charge against the 2011-12 series was that manufacturing growth looked strong partly because input prices, especially commodities, fell while output was deflated by the same index. Moving 28 of 30 manufacturing categories to double deflation removes that bias at its source. The two exceptions, food processing and pharmaceuticals, are import-heavy, and imports need their own price indices, which India still lacks at item level. This is a real improvement, but it also makes the new series less comparable with the old, so any growth comparison across the two needs care.
2. A smaller informal sector is a finding, not automatically a flaw. Cutting household GVA by ₹2.9 lakh crore for 2022-23 sounds like the informal economy has been written down. But the change comes largely from construction, where direct survey and corporate data show that more activity is formal than previously assumed. At the same time trade, hotels and road transport are found more informal than earlier estimates. The direction of the change matters less than whether it rests on direct measurement; ASUSE and PLFS are a better base than the old practice of extrapolating from corporate indicators.
3. The Garg criticism is about revisions, not about method. Subhash Garg’s point is arithmetic: if last year’s quarterly GDP is revised down from ₹86 lakh crore to ₹80 lakh crore, this year’s growth rate rises even if this year’s level is unchanged. MoSPI’s answer, that revisions are systematic and follow SNA 2008 and the IMF manual, is correct in principle but does not address the timing. The way to settle such doubts is to publish the full back-series and the reconciliation between old and new estimates. Publishing the manual in seven months is a step in that direction, but data users will judge by whether the back-series follows.
4. Speed of disclosure is itself a credibility asset. In the past, manuals took up to three years, allowing suspicion to fill the gap. Releasing the methods within seven months, and noting it contains no new data, allows independent economists to replicate estimates. Statistical credibility is a public good: investors, the Finance Commission and the RBI all depend on it. The counter-view is that methods without underlying unit-level data still leave room for doubt, since outsiders cannot check the calculations.
5. Better measurement of labour and enterprise is the real gain. Using PLFS for labour input and ASUSE for small enterprises means that the national accounts now see the self-employed worker and the small shop more directly. That helps policy for the labour and entrepreneurship factors of production, since schemes for micro enterprises need an accurate count of their output. It also exposes the size of informality in hotels and transport, sectors where policy has assumed formalisation is progressing faster than it is.

Administrator's brainstorm