Sunday, Nov 27, 2022

In Defense Of India's New GDP Numbers

While the new numbers are robust, they would make it more difficult for policy makers to even forecast one-step ahead.

File - AP Photo/Ajit Solanki

Recently the Central Statistical Office (CSO) has revised the estimation of GDP in India. While such revisions are undertaken on regular intervals, say once in five years, the extent of revisions could differ. 

The current revision is based on the 2011-12 prices compared to the old numbers, which were based on 2004-05 prices. The change in base years are normally undertaken for price corrections. Of late with the availability of more and robust data bases and also with the changing international practices, the revisions in data not only undertakes price changes but also undertakes change in the production function itself. 

In the latest round, the CSO has made changes to the prices to 2011-12 and also brought in some conceptual changes that are part of UN System of National Accounts (2008), which is followed by most of the countries in the world. 

These revisions also include adoption of latest national accounts classification system that disaggregates output into 11 sub-categories, and also uses recent data bases. One significant difference in the current revisions in terms of databases is the usage of MCA21 (Ministry of Corporate Affairs) database that provides information based on annual accounts of large part of corporate sector (including manufacturing and services) at the enterprise level. It has also used extensively the sales/service tax data in the case of service sector. 

As per the international best practice, the new revision suggest that 'GDP at market prices' as the reference GDP of the country (compared to GDP at Factor cost that was used in the past). Further, it discontinues the estimation of concept of Gross Value added at Factor cost, although the analysts could still estimate it from the taxes and subsidies available at the macro level. 

Based on the new estimates, the share of 'manufacturing' sector has increased substantially while the share of 'trade' group has declined. This is largely due to shifting the manufacturing trade from 'trade' group to 'manufacturing' group. The share of other segments (except agriculture) has also seen some changes that overall has brought down the share of services sector (including construction) from about 65% to 59.5%. These revisions have also resulted in change in the growth rates of GDP. 

For the years 2012-13 and 2013-14, the GDP (based on GVA at basic prices with 2011-12 base) are estimated to be at 4.9% and 6.6% compared to 4.5% and 4.7% respectively in the 2004-05 base. However, the reference GDP, which is the GDP at market prices, for 2014-15 is estimated at 6.9%, higher compared to 5.1% growth in 2012-13, indicating a recovery in the overall economic activity in 2014-15. Going by the disaggregated data also, one could find such recovery even when GDP figures are estimated with the old base. 

What are the implications of such revisions in GDP? First and foremost, it is important to understand that the revisions are based on robust concepts and methodologies adopted at the international level. In the past, given that relevant databases were not available, GDP was estimated at factor cost, which had its own limitations. Second, it has implications on the other macro indictors that we refer in relation to GDP such as fiscal deficit, CAD, investment rate, etc. 

These indicators also need some revisions. However, as the change in the new GDP levels in comparison to old series is not very much, these ratios may not change substantially and can be easily used for policy purpose. Third, predictability of economic behaviour in the near future would be difficult until and unless the CSO brings out back series that is consistent with the present concepts and methodology. 

In that sense, the new GDP numbers could make it more difficult for policy makers to even forecast one-step ahead, leave alone the long term policy simulations exercise. This would also pose problems for the overall budget making process. 

Now that Finance Ministry is preparing the Union Budget for the year 2015-16, one crucial input that they need is the nominal GDP growth for 2015-16. With the new GDP numbers, this would be extremely difficult to estimate. Lastly, applicability of this methodology at the state level is going to be much more difficult.

As one is expecting to see large policy actions at the state level in terms of attracting investments (both foreign and domestic) a reliable GDP indicator at the state level is crucial for long term investors. The new GDP methodology could make states almost impossible to revise their GSDP estimation as the basic data sources relevant for such exercise are limited. 

Overall, while the revisions in GDP was very much needed, it is also necessary to improve the quality of data bases at all levels so as to derive consistent, comparable, and relevant output numbers both at the all-India as well as at the State level.

Prof Bhanumurthy is a senior economist with National Institute of Public Finance and Policy.