The National Statistical Office (NSO) has reportedly “tested” the goods and services tax (GST) data to compute the gross domestic product (GDP) for July-September and October-December quarters of the current fiscal year. This is a prelude to the NSO formally making use of the information on the key indirect tax levied on a wide gamut of transactions to estimate private consumption, the largest expenditure component of the GDP. Consumption is not directly estimated at present, but calculated from production data. In recent periods, this has particularly raised concerns, given the wide divergence between the GDP and consumption figures. It is equally important that granular GST data is made available, so that financial health of businesses as well as economic conditions of different geographies could be better understood and analysed. Also, creditors could use disaggregated GST data to vet loan applications, and for cash flow-based lending. The NSO’s move is part of a comprehensive exercise to reinforce India’s economic statistics. Revision of the base years of GDP and key price and industrial output indices (wholesale price index, consumer price index, and index of industrial production) and construction of a producer price index too are under way. All this is welcome and much needed to make the official data more credible.
It is undeniable that India’s statistical system is fraught with multiple infirmities and rigidities. In 2019, a Harvard University working paper by former chief economic adviser Arvind Subramanian noted with “robust evidence” that after India changed its data sources and methodology for estimating real GDP for the period since 2011-12, growth has been significantly overestimated. While official estimates placed annual average GDP expansion between 2011-12 and 2016-17 at about 7%, the paper reckoned that actual growth may have been “about 4.5% with a 95% confidence interval of 3.5-5.5%”. If that were the case, something might have been seriously amiss, and policy or business decisions shouldn’t rely on such data. To cite a lapse, after the GDP series based on 2011-12 prices was brought out, the NSO stopped publishing re-based data for the previous series (2004-05) for the years since 2008-09. This has made rigorous historical comparisons implausible.
There are also increasing instances of various sets of official data showing dichotomy with the information sets gathered by private agencies and anecdotal narratives. The issues being cited in informed circles include “lack of transparency”, absence of enough granular data, “outdated” methodologies for sample collection, and surveys being not consistent enough. Timely changes in data computation techniques, updated sampling methodologies for surveys, and constant broad-basing of data sources are vital for statistical integrity. While this holds true worldwide, India’s economic statistics has its intrinsic set of problems too, with the economy being largely services-driven, with a very large informal sector.
The recent spate of large revisions of national income data, like a 100 basis point upgrade for real GDP growth in FY24, has puzzled many, including seasoned statisticians. The only — arguably inadequate — explanation for this is that unlike in the past when unregistered entities’ data came with a frequency of once in five years, the same has been captured annually since 2021-22. The government would do well to ensure statistical methods are dynamically updated, and the relevant mechanisms are well-equipped with technology and manpower. Independent oversight of statistical practices by professional bodies is a must. The easier way to ensure this may be to adequately empower the National Statistical Commission.
