About three-and-a-half years ago, I commented on the just-released new GDP series (‘The new GDP is for real’, IE, April 18, 2015). Unlike several times in the past when the base-year had been changed (1970-71, 1980-81, 1993-94, 1998-1999, 2004-5), the new base year adoption of 2011-12 met with howls from sections of academia and the political Opposition. The academic types, correctly, demanded that the back-series of the data should be released before the final acceptability of the new-series. While I agreed with the principle of releasing the back-series data, I accepted the new-series as “correct” for the simple reason that it had undergone changes under the guidance of the IMF, and the UN system of accounts.
Constant introduction of new methods and series is an expected development in emerging economies. Why? Because such economies are going through major structural changes and, therefore, new estimation is mandatory.
Unfortunately, for the first time, revisions to GDP growth rates became a political football. In one sense, this is a tribute to India moving from a poor developing economy status to a transforming lower middle-income country. Maybe it is also a tribute to the fact that caste and other traditional (assumed) determinants of voting are becoming less important as voters are voting with their pocket-books. Hence, economic data — particularly real GDP growth and inflation — are assuming centrestage.
This reality coincided with the overwhelming defeat of the Congress in 2014; moments before the release of the election result, I had a bet with several fellow pundits about the outcome of the election. These pundits, whose sympathies lay with the Congress, believed that the 2014 result lead to a hung Parliament. And since the BJP (Narendra Modi) did not have any allies, the Congress would form the government.
But the Congress lost spectacularly and this seismic defeat had to have consequences. The 2014 election campaign was centred around low GDP growth rate, high inflation, and high corruption during UPA-II. Hence, any “right” thinking member of the Congress party, and/or a “right” thinking member of the influential English media, would want to overturn the conventional wisdom about UPA being bad managers of the economy.
The new GDP series added more fuel to the conclusion that UPA-I managed the economy badly. The level of GDP in the new base year (2011/12) was lower by about 2 per cent than the previous estimate for 2011/12 based on the old (2004/5) data. In other words, the cumulative growth prior to 2011/12 was lower by 2 percentage points (ppt). Over a gap of 10 years or so, this is small potatoes in terms of differences in growth rates, so the rejection of the new GDP series was somewhat inexplicable. What made this rejection even less understandable was the fact that the old nominal and real GDP in 2013/14 was identical to the new nominal and real GDP. And even less than less understandable about the opposition to the new GDP series was the fact that for the year prior to the election — that is 2013/14 — the new series showed GDP growth to be 6.9 per cent versus the old estimate of 5 per cent.
This extended discussion has been provided to alert the reader to the fact that the new GDP series was real and should have been accepted without controversy. Frankly, I never understood the controversy.
Before we proceed, one explanation for why the new back was not provided at the time of the new series release in 2015. This was because of a major change in the collection of industrial production data; the new series did it through data provided by the Ministry of Corporate Affairs (MCA) while the old series was based on index of industrial production (IIP). Unfortunately, MCA data was only collected from 2006 onwards — hence, the maximum backward series that could have been supplied was from 2006.
Nevertheless, there was a justified demand for the back series, and a committee was set up under the aegis of the National Statistical Commission. A draft of their back-series report has just been released and is available on the MOSPI website series (Report of the Committee on Real Sector Statistics; hereafter, the Report). And before one could say politics, the Congress, and former Finance Minister P Chidambaram (and the Twitter army of the Congress) was agog with the claim that the back-series report showed that in two years (2007/8 and 2010/11) India clocked double digit growth, 10.2 per cent and 10.8 per cent respectively — hence, it was unfair, and a canard, to state that the UPA had mismanaged the economy. As it happens, the 10.8 per cent figure was there according to the old GDP series; for 2007/8 the earlier estimate was 9.8 per cent, now enhanced to 10.2 per cent. Not much to be concerned about.
But there is a lot to be concerned about with the back-series estimates, regardless of one’s ideology or training in economics. As I had pointed out in my 2015 article, and it bears repetition, “how come after all the revisions etc, real GDP was 2 per cent lower in the base year of revision, FY 2012? This goes to the heart of the differences between the old and the new. Two sectors of the economy undergo fundamental changes in the revision — wholesale and retail trade (WRT) and manufacturing. The former sees a decline in its share of GDP by approximately 5.5 percentage points, and manufacturing witnesses a 4.5 percentage point increase. The new shares are about equal for both (18 percent of GDP). Which means that two-thirds of old and new GDP is not witness to much change.”
The change in manufacturing is due to the use of MCA data rather than IIP. I want to concentrate on WRT (represented by the national accounts category, “Trade, Hotels, transport and communication”; hereafter referred to as THTC). Note that this is about 18 per cent of the economy, and is the sector for which the information available is minimum. It appears that the nominal rate of change in THTC GDP was proxied, in the national accounts, by the rate of growth in state sales tax collections — this was made formal in the new GDP series.
For the five low inflation NDA-I years (1998-2003), nominal GDP growth in the new THTC (back-series) averages 9.1 per cent per annum, state sales tax revenue averages 11.4 per cent and the old THTC series averages 11.5 per cent. Everything is nicely consistent. In the next five UPA-I years (2004/5-2008/9), state sales tax growth averages 14.9 per cent, the old THTC growth averages 14.9 per cent but the new THTC back-series averages 22.8 per cent per year!
WPI inflation was higher by 1 percentage point (ppt) in UPA-I compared to NDA-I; hence, the 5.8 ppt difference in THTC growth (old series) means that real THTC growth was higher by nearly 5 ppt in the UPA-I period. This reality is there in the old series.
In real terms, THTC grew at a (9.1-4.7) or 4.3 per cent per annum rate in NDA-I; in UPA-I, old real THTC growth was (14.9-5.7) or 9.1 per cent per year. The new back-series real THTC growth in UPA-I: (22.8-5.7) or 17.1 per cent per annum. Think about it — wholesale trade expanded at 17 per cent real per year for five years in UPA-I?
I have looked at many series for India and many series for other economies and for many years; but I haven’t ever come close to this kind of a real “explosion”. Briefly put, the back-series does not pass this first “smell” test.
The difference between the old and new series is an extra 8 ppt of THTC in each year of UPA-I; with an 18 per cent share in GDP, that is an extra growth of (.18*8) 1.4 per cent per annum. Correcting the new back-series for this THTC “mishap” means that the new back-series, rather than averaging GDP growth of 8 per cent per annum for 2004-2008, actually averages 5.6 per cent per annum.
More research needs to be done on this important subject. THTC growth is the real “unknown”; one reasonable proxy for this is sales tax revenue growth. But the new series departs radically from this assumption, and does so starting with UPA-I.
What is clear is that the back-series report provides very little evidence for better economic management by UPA-I than previously thought; one unintended consequence of the back-series report is to downgrade GDP growth in UPA-I by over 1 per cent per annum. If that is done, then UPA-I growth does not appear to be too different than NDA-I. Is that the real message of the Report?