The recently announced GDP estimates for third quarter of 2024-25, second advance estimates of 2024-25, and revised and final GDP estimates of 2023-24 and 2022-24 show drastic revision in some of the estimates made earlier. For example, the 2023-24 GDP growth estimate has been revised sharply upwards by 100 basis points to 9.2% from the earlier 8.2%, making the growth during the financial year one of the fastest in the last decade.
Similarly, the gross value added (GVA) has been revised upwards from 7.2% to 8.6% for FY24. The second quarter 2024-25 GDP growth numbers were revised upwards by 40 basis points, from 5.4% to 5.8%. There were other revisions as well. The 2022-23 final GDP growth has been revised down by 20 basis points to 7.6%.
Smaller revisions in GDP estimate numbers are a common practice, what baffles analysts and economists are the massive revisions. As Madhavi Arora, economist, Emkay Global Financial Services, quips: “Massive upward revisions to past years and quarters have made the GDP forecast exercise extremely dynamic.”
Projections that miss actual numbers by such a large margin lead to wrong policy decisions by the government as well as private entities. If accuracy of government data becomes a suspect, the effectiveness of policy making comes under scrutiny as well.
“One of the sharpest revisions in recent years makes it difficult to interpret the data appropriately. Data revisions are understandable as the statistics office has access to more data as compared to provisional estimates. Yet quality of data remains a concern,” Kunal Kundu, India Economist at Societe Generale Corporate and Investment Banking.
Kundu sees more issues with such sharp revisions in GDP numbers. He says the discrepancy component in GDP calculation should have moved closer to zero with more relevant data, but it remains unchanged. He points out that despite no major revision in WPI data, the GDP deflator changed significantly in certain periods. “In fact, the skewed deflator makes real GDP appear more robust as compared to virtually a secular decline in nominal GDP in growth terms,” he further adds
Why the revisions
Shubhada Rao, economist and founder QuantEco, an independent economic research firm, explains: “Provisional GDP numbers usually rely on listed firms’ data. The unlisted (unorganised) firms’ data is usually available with a lag. It is likely that unlisted firms fared better and hence such a substantial upward revision.”
According to Manoranjan Sharma, chief economist, Infomerics Ratings, this (revision of GDP numbers) is a multi-layered issue and involves greater data coverage, changes in research methodology and adoption of best practices elsewhere to make the process of data computation more scientific and realistic.
“My limited point here is that such periodic data revisions should not be summarily dismissed but need to be carefully considered in the light of the transforming big picture and the granular details to make an objective assessment of the comprehensive national income growth process,” says Sharma.
Sunil Kumar Sinha, professor of Economics at Institute for Development and Communication, Chandigarh, explains that generally in an upswing, initial estimates tend to underestimate the GDP and hence revisions tend to push GDP upwards and vice versa. He says even during the 2004-2009 period revisions pushed GDP upwards and sharply for a year or two. “Many revisions during 2014 to 2020 were downwards. Primarily revisions tend to be driven by not being able to estimate the economic activities of the unorganised sector correctly,” he adds.
But experts are also of the view that with the use of technology and data available from such vast sources, the efforts should be to limit the quantum of revisions.
“With tech advancements in capturing real time data and also harnessing the GST data, we should strive to minimise the magnitude of revisions. About time, we forecast our past with accuracy,” quips Shubhada Rao of QuantEco.