We must buy India’s growth numbers with a pinch of salt

Past quarterly GDP revisions being announced by the government frequently requires that we read current GDP data with caution

India’s economy is estimated to grow at 4.2 per cent in FY20, recording its slowest growth since the 2008-09 global financial crisis. The quarterly GDP growth was estimated at 3.1 per cent for Q4-FY20, the lowest since March 2009. The slowdown has links to the nationwide lockdown and limited economic activity in the second fortnight of March 2020 owing to the COVID-19 pandemic coupled with the lingering domestic and global concerns prior to the COVID-19 obstacle. The economy in FY20 has been largely driven by the agriculture sector and government spending while manufacturing, construction and the services segment have been the laggards.

The data for the quarter and FY20 needs to be read with caution as the press release highlights the paucity of data during compilation amidst the global pandemic, consequent lockdown and extension of regulatory compliance. This entails revisions in these numbers with the current data based on some degree of extrapolation.

However, what is interesting is not the expected decline in GDP growth in Q4-FY20 but the revisions in the quarterly growth numbers released yesterday vis-à-vis the ones released earlier for the same quarters. These revisions can be on account of latest data used in compilation across the previous quarters leading to the “base effect phenomenon”, which simply means increase in the current growth due to downward revision in the growth of previous period.

It is important to note that when quarterly data is not available across the sectors of GDP, the Ministry of Statistics and Programme Implementation computes the growth for the entire year based on past trends and then apportions quarterly. For example, in case of construction activity, the value added for the referenced quarter (say Q1-FY20) is estimated by extrapolation of the estimate value of the same quarter for the previous year (Q1-FY19) with observed growth in the production of key construction materials like cement, iron and steel among others. Similarly for unaccounted construction, the value is computed for the full year using past trends and apportioned equally into 4 quarters. Similar methodology is used for other sectors as well.

Also, the data for listed companies for a given year is available on quarterly basis but private companies file their regulatory accounts with a lag of around six months from the end of the previous year. As fresh data is compiled after regulatory filings by private companies, there are revisions in GDP numbers of the previous years as well.

A common trend seen during FY19-FY20 is that utilisation of fresh data in previous quarters manifested base effect revisions in current year quarters. A downward revision in GDP growth by 80-100 bps in Q1, Q2, Q3 of FY19 resulted in an increase of nearly 60 bps in the quarterly estimates of FY20 as per the February release. To illustrate, in Q2-FY20, there was upward revision in agriculture (100 bps), manufacturing (60 bps), and financial services (130 bps) partly on account of downward revision of 240 bps, 130 bps and 50 bps respectively in growth numbers in Q2-FY19. A similar revision was recorded in June numbers as well. Contrary to the February release, data from yesterday (May 29) showed a downward revision in Q1, Q2, Q3 numbers of 35-70 bps largely owing to new data/changes in extrapolation assumptions for the current year as the last year quarterly growth numbers remained unchanged.

The revision in estimates owing to fresh data is acceptable till the deviation is nominal. Wide upward and downward revisions in estimates build uncertainty for market participants and policymakers, hampering future forecasts as well. Former Chief Statistician, Pronob Sen has raised a red flag that the GDP numbers for Q4-FY20 could be overestimated by around Rs 2 lakh crs. Therefore, till further revisions, we will have to buy the growth numbers with a pinch of salt.

(Sushant Hede is an associate economist with CARE Ratings. Views are personal)

Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH.