Sources and methods

Questions about GDP calculation refuse to go away

The controversy about India’s gross domestic product (GDP) calculations refuses to go away. If anything, the blow to employment, demand, and output caused by the effects of the pandemic and associated lockdowns has reignited the conversation about whether GDP numbers accurately reflect the possible distress on the ground, particularly in the informal sector. A recent presentation by former chief economic advisor Arvind Subramanian and former International Monetary Fund representative in India Josh Felman, as reported in this newspaper, in particular has sparked discussion, given that the two authors asserted that GDP might in fact have contracted in the pre-pandemic year of 2019-20. They argue that what is being observed is a “collapse” in India’s growth, based on various indicators such as the investments shown by the index of industrial production, consumption, imports, tax revenues, and bank credit. There was a clear difference in these indicators as compared to their levels during previous periods of GDP growth. Messrs Subramanian and Felman argue that therefore there must be a difference also in the overall GDP numbers.

This is not a universally held view. Speaking to this newspaper, former chief statistician Pronab Sen pointed out that the “new” GDP calculation followed global best practices in seeking to build up a picture of GDP from value-added calculations rather than through accounting for actual physical output growth. There is no question that this method is not only more appropriate for GDP calculation but also that the new GDP series was constructed with this end in mind. It is possible, Dr Sen argues, that the value added might have increased — for example, through an increase in the value embedded in the production of high-end automobiles — while the physical amount of what has been produced might have decreased. This might also be the product of a shift in income distributions and an increase in inequality. Yet it is equally true that, if the case is that the value added is increasing, it should be effectively reflected in some other indicators — for example, overall corporate earnings. Those who disagree with the Subramanian-Felman thesis have not effectively identified these other indicators.
 

The broader point here is that India cannot afford basic questions about its most cited data series to linger for this long. Part of the problem is that good faith questions about the GDP numbers are too often discussed in a vacuum of genuine information as to the decisions made in constructing the series. What is surveyed, what is counted, what is estimated, and what is extrapolated? How are these indicators put together, and are the deflators appropriate? What is the systematic process for revisions as more data is known? How is the informal economy being assessed? It is incumbent upon the National Statistical Office (NSO) and its nodal ministry to take more effective steps to address these questions. Restoring faith in the statistics is a necessary first step towards the formulation of an appropriate post-pandemic economic policy. It has now been almost a decade since the last comprehensive study of sources and methods for GDP estimation was published by the NSO, in 2012. The year 2022 should see another such publication that might lay some of these questions to rest.