The gap in manufacturing sector GVA

The manufacturing sector includes the organised sector comprising registered factories and companies, and the unincorporated sector comprising small factories and workshops outside the corporate sector.

Early this year, the National Statistical Office (NSO) released the new series of the National Accounts Statistics (NAS). It shows that the manufacturing sector’s gross value added (GVA) is ₹38.6 lakh crore (₹38.6 trillion), constituting 14.7 per cent of GDP (gross domestic product) for the year 2023-24 at current prices. How reliable is the official estimate? This article investigates.

The manufacturing sector has two parts: One comprises all registered factories employing 10+/20+ workers with/without power, including registered companies (defined as the organised/formal sector). The other comprises unincorporated/ informal/ household units consisting of small factories/workshops falling outside the corporate/factory sector. The Annual Survey of Industries (ASI) reports the production accounts of the factory sector, while the Annual Survey of Unincorporated Sector Enterprises (ASUSE) covers the informal sector. The combined output (GVA) of the factory and unincorporated sectors almost completely represents total manufacturing output.

For 2023-24, the sum of the GVA of units covered by the ASI and ASUSE is ₹27.4 lakh crore (the Alternative Estimate or, AE). The official figure, as per the NAS, is ₹38.6 lakh crore. It is higher than the AE by a whopping 40.9 per cent. Let’s call this the GAP. What could account for it?

While minor variations between the two estimates is understandable due to methodological or definitional issues, such a substantial GAP between the two estimates using official data sources surely raises many questions.

What are the data sources used for estimating the official GVA? For the unincorporated sector, it is ASUSE – the same as what we have used to estimate the AE. Hence, it cannot account for the GAP noted above. Moreover, the unincorporated sector’s share in total manufacturing GVA is a mere 13.9 per cent.

Hence, the reason for the GAP must lie in the estimation of organised manufacturing output. The NAS uses company balance-sheet data sourced from the Ministry of Corporate Affairs’ database (MCA-21), compiled using the annual statutory filings by registered companies. This practice of using the MCA data began with the previous NAS revision (base year 2011-12), partially replacing the ASI. The procedure has continued in the latest revision, with minor modifications.

A standard way to validate the GVA is to use employment data to estimate the sector’s potential output by applying appropriate “technical ratios” drawn from the ASI and ASUSE datasets. The official Periodic Labour Force Survey (PLFS) for 2023-24, estimated that the manufacturing sector employed 697.5 lakh workers.

However, as per the ASI and ASUSE datasets, only 532.9 lakh workers were employed to produce the official GVA. Thus, quite possibly, the contribution of the remaining 164.6 lakh “residual workers” — 697.5 minus 532.9 — may account for the GAP in the GVA reported above.

The definitions of employment in the three surveys are not the same, and their data collection methods differ. However, for a validation exercise, the PLFS estimates provide a useful reference point. Official agencies also use similar methods.

Some of this GAP is contributed by the residual 2,72,534 MCA companies, not covered in the 78,618 “private companies” captured in the ASI data. A majority of the residual companies are likely to be to non-factory private companies. The left over residual workers are likely to belong to the unincorporated sector not covered in ASUSE survey because of their small size.

Then, applying the “technical ratios” of the appropriate segments of manufacturing as derived from unit-wise ASI and ASUSE data, the potential GVA of the residual workers is estimated to be ₹3.6 lakh crore.

Then adding the potential GVA to the AE of ₹27.4 lakh crore reported earlier, the likely/potential overall manufacturing GVA could be ₹31.0 lakh crore.

This figure, however, still falls short of the official GVA estimate (of ₹38.6 lakh crore) by 24.5 per cent. In other words, the potential GVA of all workers employed in the manufacturing sector in companies and unincorporated sector enterprises would at best account for 80.3 per cent of the official estimate. It still leaves ₹7.6 lakh crore (or 19.7%) worth of NAS manufacturing GVA “unaccounted” for or “unexplained” (Figure 1). This is the real puzzle of the new GDP figures for the manufacturing sector.

How can this GAP be reconciled with the best alternative estimate obtained using widely used data sources? The NSO documents say that, being an establishment-based survey, the ASI reportedly fails to capture value addition taking place within an enterprise, but outside of the factory premises (such as head office, marketing and distribution, or R&D activities). Is this really true? Probably not. Available evidence does not seem to support the official view (Dholakia, Nagaraj and Pandya, Economic and Political Weekly, 2018).

Alternatively, if the ASI-based GVA estimate did not underestimate production, how could such a large GAP arise. Is it because of the NSO methodology of scaling up of sample estimates of active companies for the universe of companies whose size and composition are hazy and unverified?

As per the latest NAS, for 2023-24, manufacturing GVA is ₹38.6 lakh crore at current prices. An alternative estimate, using time-tested official ASI and ASUSE data sets, finds the official figure to be higher than the alternative estimate by 40.9 per cent.

Even after accounting for the contribution of residual companies and workers, the unexplained GAP remains at 24.5 per cent. Whether the official estimate represents a “fuller description of ground reality” through the use of corporate data, or amounts to an overestimation of output, remains a matter of contention. The statistical issue can only be resolved if the MCA data and NSO’s methodologies are made public for independent verification.

(Jatinder S. Bedi is a Professor of Economics at the Institute for Development and Communication, Chandigarh. R. Nagaraj was formerly with the Indira Gandhi Institute of Development Research (IGIDR), Mumbai)

Text and Context / The Hindu Explains / economy (general)