The new intellectual inputs that went into the computation of the new series of GDP with terms like basic prices and distinction between product and production taxes being thrown in, has added to the level of confusion when interpreting these numbers. Nobody can pin point as to what is amiss in the methodology, and hence it must be right. But every time the estimates come out, there is scepticism as the numbers do not quite gel with the ground reality. The Q3 results add to this debate as it demonetisation did not quite matter in terms of impact on growth—the number is really good at 7%. No agency or economist expected such a positive scenario.
To be fair, the CSO numbers do tell us that compared with FY16 when growth was 7.9%, FY17 would be inferior at 7.1% which means that 0.8% has been shaved off due to demonetisation as there was nothing else amiss in the economy. Q3 and Q4 were to be the leading quarters, and hence it is possible to argue that at 7% in Q3, growth would have been higher in case there was no demonetisation. In fact, at the start of the year, it was assumed that growth would be around 7.6% and with a bit of luck could touch 8%. Hence, the economy has definitely come lower this year.
Further, the segments that were affected quite sharply were real estate, consumer goods and construction, which is what the Q3 growth numbers do reveal. Consumer goods, however, remain an enigma as manufacturing has grown at a high rate of 8.3%.
Two issues need to be addressed by the CSO. First, it is essential to create a series for the earlier years prior to FY12, so that one can benchmark these growth rates. The present series which keeps GDP growth broadly in the range of 6.5-7.5% for the last 5 years or so, may have had much higher numbers in the good pre-2011 years when the FY05 base year yielded growth rates in the region of 8-9%. A conjecture could be that this could be 9-10%, in which case it is possible to position the growth rates of 7% in the present context.
The other is to match growth in physical numbers with value added, which holds especially in industry where there are clear corresponding numbers available. For instance, value added in mining, manufacturing, electricity and construction grew by 7.5%, 8.3%, 6.8% and 2.7%, respectively in Q3-FY17. In physical terms, IIP shows that growth rates were 5.2%, 0.20%, 5.25% for the three major segments, and steel (12.5%), cement (-0.85%) which are taken to be synonymous with construction. The difference for the three leading sectors is 2.3%, 8.78% and 1.5%, respectively. So, low production lead to high value addition. The argument that manufacturing is producing more high value items which even on low physical production numbers yields high value addition no longer holds as this would be the case for all the years and does not reflect a single year phenomenon as the base for comparison would be progressively higher. As value addition is sum of profits and salaries, does this mean that while jobs have not increased, value addition has and led to disproportionate gains for owners of capital? This would be of interest to Thomas Piketty.
Another area which has to be addressed is the case of constant revisions to GDP estimates, which has become a norm. The FY17 numbers released in February 2017 would be different from those in May 2017 when the first annual estimate is provided which will then further undergo a change when the first advance estimates for FY18 come in January 2018 with revised FY17 numbers.
For FY17, the GVA numbers for the first two quarters have been revised downwards by up to 0.4%, while GDP growth numbers have improved. In FY16, the variation in GVA between the first set of estimates and the ones provided this time was as high as 0.9% in Q2, while GDP strayed by 0.6% during this period. Such deviations are serious as they could change the judgement from being good to ordinary or the other way round. Based on such experiences, it could be possible to work on the assumption that there could be up to 0.5% change in the number when the final estimates arrive.
The issue to be debated is whether or not the CSO should be in a hurry to bring out high frequency data which is susceptible to wide swings. The counter argument would be that having some indicative numbers is better than not having any such guidance and if all operators work on the principle of ‘plus/minus 0.5%’, things will still be manageable.
The major challenge for the CSO is the existence of a very large unorganised sector. Agriculture is tough because even the ministry works on guesses all the time as a large part of the output never enters the market as the marketable surplus is lower and has to be estimated. Further, all sales transactions are not recorded which makes it difficult to arrive at the output numbers. Satellite imaging helps because there is no alternative.
You might also want to see this:
In both manufacturing and services, unorganised sector is large. SMEs do not have reporting statements. This also holds for unorganised retail, transport, professionals, hospitality, where there could be under-reporting. Monetising the economy is a way out and introduction of GST should make things more transparent.
There are evidently no easy solutions as the economy is complex and largely unorganised. Getting in the National Agricultural Market and GST which involves online recording will help to diminish the noise elements in the data. The push given to e-filing is also useful. The local bodies especially the panchayats need to be involved aggressively and harness technology to improve the systems of reporting. This surely will be a long process, but should be expedited with a time frame kept in mind while bringing all levels of the government on the same platform.
The author is chief economist, CARE Ratings. Views are personal