Ever since the new GDP data was released in January 2015, most economists, in the government and outside it, have been more than sceptical. With the release of the advance estimates of the GDP for fiscal year 2015-16 (FY 2016), which showed the GDP growing at 7.6 per cent, the crescendo of criticism has reached new heights.
I will show below that this criticism and/ or scepticism is completely unwarranted, and that those who have seriously questioned the GDP data owe apologies to the Central Statistical Office (CSO) for doubting the authenticity of the data and the integrity of the statisticians.
The setting up of a GDP review exercise in India (under the “guidance” of the UN and the IMF) is a pre-planned exercise and conducted every decade or so. In May 2014, India had a national election in which there were only two themes — first, high inflation and low growth, and second, corruption. One of the major results of the new GDP series was that the growth in the year preceding the election, FY14, was jacked up from a previous 4.7 per cent estimate to a respectable 6.9 per cent (later revised to 6.6 per cent).
There are four reasons that explain the presence of the doubting Thomases. First, India’s growth could not be just inches below China’s; it had to be several feet below. Second, India couldn’t, by definition, grow faster than China. There has been a rapid decline in China’s GDP growth rate, from double-digit levels that prevailed for 32 years (1980-2011) to a sudden drop since FY12. The new GDP numbers for India are 6.6 per cent in FY14 (below China’s 7.3 per cent) and 7.2 per cent in FY15 (way above China’s 6.9 per cent). Hence, the only safe conclusion was that the data was not telling the truth.
The third important reason for doubt was purely political. The CSO was being told, according to some doubters, by the Narendra Modi government, in that ever so conspiratorial way, to massage the data.
The fourth reason may be the most important explanation. For quite some time now, the world has been gripped by deflation, or at least declining inflation. India’s GDP deflator (implicit price index), too, has been falling. It touched a historic low of negative 1.0 per cent in FY16. Given this fact, herewith some strong evidence that GDP growth for FY16 may even be higher than that claimed by the CSO.
Nominal non-food bank credit has been creeping up in FY16 and places the real credit growth in the first quarter of FY16 as the eighth fastest rate in the last nine years and the highest rate since Q2 of FY11. Further, few have marked that quantity of oil imports is up a healthy 8.3 per cent. This is the seventh highest April-December rate since the beginning of the high growth period in FY04, and near identical to the average growth rate of 8.5 per cent for the same period.
None of these statistics however, are convincing for the real doubter. The Wall Street Journal, in an article dated February 10, boldly headlined “5 charts that show India might be overstating its growth”. The article asserted that growth was “nowhere near as strong as the data suggest [7.6 per cent] and that other performance indicators show the economy is still struggling to gain momentum. The government made some major changes in the way it measures GDP in January last year, which resulted in a bump in growth rates and presented a much rosier economic picture”.
A major item on everyone’s list (excluding the lazy economist’s citation of low export growth, again, an item messed up by disinflation) is that manufacturing is much weaker than what the data indicate. Prior to the GDP revision, India relied on the Index of Industrial Production (IIP) to estimate growth in manufacturing. Since last year, the CSO has moved to the Ministry of Corporate Affairs (MCA) balance sheet database for an aggregate of over 5,00,000 companies (versus a few hundred contained in the IIP data). Surely, a big improvement in coverage.
However, till date, there is no explanation for this, and I believe there cannot be an adequate explanation why the CSO has refused to release these data. Unless the MCA database is released, regardless of the evidence, the CSO data will be (correctly) treated with suspicion: Hah, you are not releasing the data because you have something to hide.
The IIP data shows a 3 per cent real growth for manufacturing between April and December while the CSO data suggests that real growth was 9.5 per cent and nominal growth was 8.1 per cent. Growth in value added is what the GDP accounts measure — and nominal value added growth is approximately equal to a weighted average of profits and wages. Labour compensation accounts for a significant proportion (almost 90 per cent) of total output, and this has grown at an 8.5 per cent rate in FY16; median profits for a sample of 120 manufacturing companies (BSE 500) rose by 18 per cent. This suggests that value added, in nominal terms, has grown at a 9.5 per cent rate.
The CSO estimate is based on 70 per cent balance sheet growth for organised manufacturing and 30 per cent for the unorganised sector; for this latter component, IIP growth is the proxy used by the CSO. The CSO price deflator for manufacturing is minus 1.4 per cent. Hence, our estimate of real manufacturing growth is (0.7 x (9.5 + 1.4) + (.30 x 3)) or 8.5 per cent vs CSO 9.5 per cent — close!
So I end this discussion with two pleas. First, the CSO should release the MCA database. Second, the harsh critics of the new GDP data should tender an apology to the CSO.