It was good that Dr Raghuram Rajan, Governor, RBI, has cleared some of the mist — and myth — surrounding the GDP numbers. He gave two simple examples:

* First example: ‘Exports minus imports’ is a contributory factor to GDP. Exports declined in 2015-16, but imports declined even more. So, ‘exports minus imports’ in 2015-16 was greater than ‘exports minus imports’ in 2014-15. That means growth, contributing to the GDP number (in current prices)!

* Second example: ‘Sales minus input costs’ is another contributory factor to GDP. Sales of all manufacturing firms declined by 9.02 per cent in 2015-16 over the previous year, but input costs declined more sharply (18.42 per cent), thanks to the steep fall in oil and commodity prices. Consequently, ‘sales minus input costs’ was greater in 2015-16 than in 2014-15. Result is growth!

Output declines, growth rises!

In both cases, there was no increase in output. Exporters exported less, factories produced less, firms had lower turnovers, there was little or no additional employment, and yet the sectoral ‘growth’ figures of the export sector and the manufacturing sector were positive, leading to the illusion that export and manufacturing sectors had ‘grown’ at a brisk pace. (Here, I assume that productivity remained constant.) Since the two sectors contributed significantly to the overall GDP number, the GDP number got a boost.

The steep decline in oil and commodity prices is a new phenomenon and I think the CSO is not equipped to construct credible growth numbers in such a situation. The disconnect between the numbers and the reality on the ground is very evident. No amount of rationalisation can change the reality that is felt every day by the citizens — CPI inflation upwards of 5 per cent, no new visible investment and no additional jobs on offer. The average person is therefore stunned when he is told that the GDP had grown at 7.6 per cent.

We get a more realistic picture of the economy if we look at the output numbers. Of the eight core industries, coal, refinery products, fertilisers, cement and electricity produced more in 2015-16 while crude oil, natural gas and steel produced less than in the previous year. The growth in the Index of Industrial Production of the eight core industries in 2015-16 was only 2.67 per cent. Nevertheless, the CSO reported high growth in the three sectors (that encompass the eight industries), namely, mining and quarrying (7.44 per cent), manufacturing (9.29) and electricity and gas (6.57)!

Deficiencies in methodology

Ever since the CSO moved to the new method of calculation, there has been growing scepticism. Analysts have pointed to several aspects of the new methodology that are suspect. One of the main deficiencies was the choice of deflators. Gross value additions (GVA) are measured in current prices but in order to make them comparable with measurements based on prices in the base year (2011-12), one must ‘deflate’ them with an appropriate deflator. In the cases of ‘trade, hotels, transport, communication etc’ and ‘financial, real estate and professional services’, the GVA at current prices was 6.6 per cent and 7.4 per cent respectively. The CSO applied negative deflators and reported a growth rate of 9.0 per cent and 10.3 per cent respectively. When the 12-month average of CPI inflation in 2015-16 was 4.9 per cent, it is difficult to believe that inflation in these two sectors was negative! There has been no convincing explanation why negative deflators were chosen for these two service sectors.

The second deficiency was the use of the MCA21 database. It has been admitted that the database requires to be stabilised. We also know, anecdotally, that small and medium firms find it difficult to report the data about their firms periodically in the new format. Besides the MCA21 database has been shared only with the CSO, and no one has been able to verify the correctness of the data.

The GDP growth number is an approximation of the underlying economic activity. Many economic decisions of the government and private investors are based on that number. My analysis leads to the conclusion that the GDP number is not ‘as accurate as possible’ and certainly does not inspire confidence. Some corrections are therefore in order.

Corrections needed

The first correction must be to identify industries and sectors where quantitative output had declined in 2015-16 over 2014-15 (productivity assumed as constant) and assume that growth in such industry/sector was zero. Examples would be exports, crude oil, natural gas and steel.

The second correction would be to re-visit the deflators that were used. There is a huge divergence between WPI inflation and CPI inflation. It makes no sense to use negative deflators for sectors (especially services) where oil and commodity prices have no role to play and where consumer prices have, in fact, increased.

The third correction is to reconstruct the GDP numbers using the old, conventional data and release them alongside the numbers using the MCA21 database until the MCA21 database is believed to have been stabilised.

Finally, there is no harm in publishing, at least for a few years, GDP numbers using the old method and the new method. The CSO should complete the following table and restore credibility to the whole exercise:

My guess is that the the number for each of the question mark-years would be between 5 and 5.5 per cent.