In new GDP back series data, average growth during UPA regime is down from previous estimates while growth during NDA is pegged higher than during UPA. Why do findings differ from one report to another?
Why the back series?
When the Central Statistics Office (CSO) moved to a new base year of 2011-12 for national accounts in January 2015, it was faced with a peculiar situation. The CSO faced issues in evaluating GDP with the new base year for years preceding 2011-12 due to lack of availability of the MCA-21 database.
MCA-21, an e-governance initiative of the Ministry of Company Affairs was launched in 2006, to allow firms to electronically file their financial results. With the shift to the new base year 2011-12 from 2004-05, wherein the CSO did away with Gross Domestic Product (GDP) at factor cost, and adopted the international practice of valuing industry-wise estimates as gross value added (GVA) at basic prices, the MCA-21 database also got used in addition to the volume index of Index of Industrial Production (IIP) and establishment-based dataset of Annual Survey of Industries (ASI).
On Wednesday, three years after the shift to the new base year of 2011-12, the CSO and NITI Aayog, in a join press conference, released the back series detailing growth numbers for 2005-06 to 2011-12.
What were the findings of the newly released back series?
The back series released Wednesday has trimmed the growth numbers for the UPA government’s nine years (2005-06 to 2013-14), with the Indian economy growing at an average 6.7% in four years of the UPA’s first term (2005-06 to 2008-09) as well as the UPA’s second term (2009-10 to 2013-14), which are lower than the earlier estimates of 8.1% and 7.0% (2004-05 base) respectively. These growth rates compare with an average 7.4% growth rate (2011-12 base year) seen during the first four years of the present NDA government.
Barring two of the years, 2012-13 and 2013-14, the back series data released for years preceding 2011-12 scaled down growth rates for 2005-06 to 2013-14 by 0.8 to 2.1 percentage points. For 2012-13, the back series based on the new base year (2011-12) revised the GDP growth rate upwards to 5.5% from 4.7% estimated earlier (2004-05 base year), while for 2013-14, the GDP growth rate was revised up to 6.4% from 5.0% estimated earlier.
Sharp downward revisions were seen particularly for two years, 2007-08 and 2010-11. For 2010-11, the growth got revised downwards from a double-digit rate of 10.3% to 8.5%. The 8.5% cent growth in 2010-11 is the highest growth rate in the back series dating back to 2005-06.
Was there not another report on growth submitted recently?
Three months ago, the Committee on Real Sector Statistics submitted its report to the National Statistical Commission (NSC) offering an alternative. It stated that the economy grew at a faster pace under the UPA government from 2004-05 to 2013-14, compared with the average growth during the first four years of the current government. The average GDP at market prices was 8.37% during UPA-I (2004-05 to 2008-09), and 7.69% during UPA-II (2009-10 to 2013-14), the Committee had said. The government had, at that time, termed the estimates as “unofficial”.
The Committee had carried out an econometric exercise using the production shift method to convert the old GDP series consistent with the new series, wherein a quantum correction factor was used instead of the usual splicing method. Simply put, the difference in output at 2011-12 between the two series occurred due to an upward or downward shift in the production curve, which included new economic activities and/or excluded older activities that became extinct. To get the new back series on the base year 2011-12, it was assumed that this shift in production did not occur in a single year and the difference in output was then redistributed asymptotically backward with an annual declining rate up to the year 1993-94.
How was the new back series calculated?
The methodology for preparing the back series estimates for the years 2004-05 to 2011-12 is largely the same as the methodology followed in the new base 2011-12 series, the CSO said in its statement. “In certain cases, owing to the limitations of the availability of data, either splicing method or ratios observed in the estimates in base year 2011-12 have been applied,” it said.
For 2004-05 to 2011-12, various components were re-estimated using a hybrid approach, it said. Changes in GVA growth rates of primary, secondary and tertiary sectors as against the earlier 2004-05 series were attributed to tweaking of data sources. For instance, changes in the growth rates of the primary sector were attributed to agriculture and mining, where the growth rates were lower. In both the new and old series of Wholesale Price Index (WPI), the index for petroleum and natural gas mining used for deflating the output of the sector was lower in 2010-11 than in 2011-12, the CSO said.
In the secondary sector, the differences in rates was attributed to changed data sources and methodologies. The Annual Survey of Industries for 2011-12 was conducted in 2012-13 and finalised in 2015. “This revised data was not available during preparation of the 2004-05 series, but has been used in the computation of the back-series,” it said.
Several changes in data sources were also made in the tertiary sector: in unorganised trade, latest round of NSS, sales tax index and new series of WPI have been used for preparing the back-series as against usage of constant GTI (Gross Trading Income) Index in 2004-05 series. In the communication sector, the new series was computed with minutes of usage rather than telecom subscriber growth. For financial services, the contribution of the Reserve Bank of India (RBI) has been computed by non-market approach as compared to considering the banking division of the RBI it as a market enterprise in the earlier series, reducing its contribution.
Are there any grey areas?
There is not enough explanation for the choice of datasets and proxies, especially those datasets that didn’t exist before 2011-12. Though the CSO release mentioned usage of several proxies, there were no details about why those were selected over other datasets. For instance, for years preceding 2006, when the MCA-21 database did not exist, the CSO has used ASI data for estimating manufacturing growth whereas economists say there could have been other indicators for the same metric.
The role of the NITI Aayog in the release of the statistical exercise of CSO, which comes under Ministry of Statistics and Programme Implementation (MoSPI), has also been questioned.
Also, the use of volume data for calculation of the back series, which could have potentially underestimated growth, has been questioned by economists. International Growth Centre’s India Director and former Chief Statistician of India Pronab Sen said, “The current series is based on company data, MCA-21, which is the balance sheet data, whereas the back series is based on volume data. So in the case of secondary sector, it was based essentially on the Annual Survey of Industries. We do know that ASI was underestimating growth in manufacturing, so I had expected to be revised upwards. But that has been revised downwards as well.”
He added, “The System of National Accounts prefers to go with volume indices, so if the new series had been volume index based then that would have been OK. Unfortunately, the new series is not. And the big difference between the volume index approach and the financial data approach is that the financial data captures changes in quality which the volume approach does not. So if a substantial part of the growth has been coming from quality, then if you take the volume approach you would underestimate the growth.”