No proof required- GDP debate: RIP

Now that the new GDP data have implicitly received the RBI good housekeeping seal of approval, perhaps we can all move on to more challenging appraisals of the Indian economy

Written by Surjit S Bhalla | Updated: April 23, 2016 4:37:40 am
GDP, indian economy, RBI, Reserv bank of India, GDP data,CSO,  Central Statistics Organisation, Raghuram Rajan, RBI governor Raghuram Rajan, ministry of corporate affairs, express opinion, Illustration: C R Sasikumar

One does not often get a chance to say “I told you so.” As readers of this column know, I have constantly reiterated that the new GDP data are authentic and correct (most recently in ‘Believe it, GDP data is right’, February 22), and there finally appears to be new data, which can put this debate to rest. The controversy over the new GDP data has raged for over a year now, and each month brings a new member of the Doubting Thomas tribe. Briefly, for those lying under a rock, the Central Statistics Organisation (CSO) brought out, in January 2015, estimates of the GDP from 2011-12 onwards using the new method of calculating GDP. Revision of GDP data is a routine exercise, and has been conducted at least four times in the last 50 years. I am sure you never heard of previous revisions and there is a political economy reason for this (more on that below).

The fact remains that this new GDP data have been questioned by all and sundry.

And the doubters are a list of who’s who — the RBI (including Governor Raghuram Rajan), Chief Economic Advisor Arvind Subramanian, and many foreign investors and prestigious publications (including The Economist, The Wall Street Journal and the Financial Times). And the questioning has varied from polite to downright defamatory. The polite version states that the new GDP data, showing a 1 to 2 percentage point increase in the growth rate than previously estimated, is incorrect because it does not feel right.

The major difference between the old and new was that the new data made a significant departure from previous methods by estimating industrial and manufacturing production from balance sheet data of both unlisted and listed firms (ministry of corporate affairs, MCA, database). These data are now available on the RBI website. And we infer that as the RBI has published these data on its own website, the RBI believes in the “accuracy” of the MCA data. So we need to strike the RBI off the list of Cassandras.

What the RBI website states is that the MCA data “have been compiled based on audited annual accounts of 2,37,398 NGNF [non-government non-financial] private limited companies received from ministry of corporate affairs (MCA), accounting for 23.3 per cent of population paid-up capital.” In addition, the results pertaining to 16,923 publicly listed companies in the MCA database have also been posted by the RBI.

The MCA data show that the 58,256 unlisted firms in manufacturing, accounting for a third of total manufacturing sales, have been registering close to double-digit growth, and a mid-teens average growth in value-added. The weighted average growth in value-added for both listed and unlisted firms was a healthy 13.4 per cent in 2014-15. Using IIP data, nominal growth in manufacturing is between 2 to 4 per cent, with the manufacturing price deflator between minus 1 and 3 per cent. Now you decide — is the “feel” of growth provided by balance sheet data of over 2,50,000 firms better or worse than the IIP data covering, on the basis of a survey, the production of a few hundred odd firms? The difference in the feel is a 10 percentage point difference in manufacturing growth.

What is curious is why these experts (and expert journalists) never once investigated whether the IIP data were correct. The IIP data are based on the Indian economy as of 2004-05; in that year, textiles had a weight of 6.2 per cent in the total value of industrial production, and motor vehicles had only two-thirds of the weight of cars, that is, 4 per cent. I could not find a category for “mobile phones”, but the classification “office, accounting and computing machinery” has a total weight of only 0.3 per cent! The MCA database, by definition, has the “correct” weights in production because they are based on balance sheet data.

Textile volumes have grown at a much lower rate (6 per cent per annum since 2004-05) compared to a 10 per cent-plus rate for the volume of cars and two-wheelers produced in India. The bottom line is that if the IIP data were updated to a 2011-12 base, just the “correction” for motor vehicles and textiles would add 0.3 percentage points to annual IIP growth.

There is yet another indicator — even the old, outdated IIP data are showing a marked acceleration over the last five fiscal years. In 2011-12, the IIP grew at 2.9 per cent. The next two years, IIP growth averaged 0.5 per cent; in 2014 and 2015, IIP has averaged 2.7 per cent. Given that industry is 30 per cent of GDP, this implies that GDP growth in 2014-15 and 2015-16 would be 0.6 percentage points above that of 2012-13 and 2013-14, simply on account of higher IIP growth.
When you point out that one indicator of “feel” — volume of auto sales — grew at 7 per cent in FY16, the highest in the last five years, the doubters just shrug their already drooping shoulders.

So now for a political economy explanation for why the doubters rose en masse. Let me make it clear — this explanation does not apply to all serious analysts, just a very large majority. The GDP controversy was ignited by the sharp upward revision to GDP growth for 2013-14 — almost a 2 percentage point increase from 4.7 to 6.6 per cent.

The ruling Congress had suffered a humiliating defeat in the 2014 general election. The common belief or conventional wisdom was that the Congress lost the 2014 election because of two years of the slowest GDP growth in more than a decade — 4.5 and 4.7 per cent in 2012-13 and 2013-14, respectively. Now suddenly, the CSO was reporting that these two years averaged 6.1 per cent growth (5.6 and 6.6 per cent in 2012-13 and 2013-14). That represented very good GDP growth — so why did the Congress lose so badly?

However, there is an alternative interpretation for the election loss of the Congress — while GDP growth was a factor, I believe that the overwhelming reason Narendra Modi’s BJP won a majority in the Lok Sabha was because of high corruption and even higher inflation under Congress-UPA rule. The Congress inherited an average inflation rate of only 4.4 per cent (1998-2004). The Congress left office with an average inflation rate of 7.9 per cent. And the average inflation rate for UPA 2 was 9.8 per cent.

By creating doubts about the GDP data, the hawa-makers hoped to capitalise on the structural change in methodology and prove to all concerned that if only the new GDP data had been released prior to May 2014, the UPA would still be in the saddle. Given that it was not, it must be that the new data are wrong! And the feel-gooders argument goes, the new Modi government was not providing any extra GDP growth. Doubter estimates of GDP growth in 2015-16 — according to the old GDP method — place it around 4.5 to 5 per cent, that is, the same that was registered by the Congress in 2012-13 and 2013-14.

A lot of smart people bought this snake oil explanation over the last year. Given overwhelming evidence to the contrary, and now with the RBI good ousekeeping seal of approval, perhaps the time has come for the Cassandras to get real.

 

The writer is contributing editor, ‘The Indian Express’, and senior India analyst, The Observatory Group, a New York-based macro policy advisory group

For all the latest Opinion News, download Indian Express App

More From Surjit S Bhalla
  • What divides RBI and MoF

    In a modern, middle-income, growing country, high real interest rates hurt the economy — and the people..

  • Whose RBI is it?

    Genuine independence of a central bank has to be earned the old-fashioned way — by performance...

  • Jobs and the pre-election year

    Conventional wisdom of no job growth in 2017-18 is found to be a gross under-estimate by a correct application and interpretation of household survey data…

Advertisement
Advertisement
Advertisement
Advertisement