So Reserve Bank of India governor Raghuram Rajan has clarified that he did not actually question the gross domestic product (GDP) statistics put out by the Central Statistics Office (CSO) based on the new methodology accepted in January 2015. This methodology calculates growth by gross value added at basic prices; the earlier method measured it by GDP at factor cost.
On Thursday Rajan was quoted as having done precisely that in the course of an address at the Indira Gandhi Institute of Development Research. The following day, delivering the C. D. Deshmukh memorial lecture in Delhi, he said he had been talking in a general sense and quoting a story widely used by economists for decades (about what happens to GDP when two mothers baby sit each other’s children) and that his words had been blown out of proportion.
Rajan may indeed have had his words twisted out of context by a controversy-hungry media. But the new methodology of calculating GDP has attracted a lot of scepticism ever since it was unveiled last year, a fact that the Mid-Year Economic Analysis 2015-16 acknowledges. Chief economic adviser Arvind Subramaniam has publicly said he finds the GDP numbers puzzling, though he has emphasised simultaneously that he does not question the integrity of the statisticians who put out these numbers. The most common criticism is that it appears to be overestimating growth especially that of industry.
So should this methodology be junked and work start on a new one? Even economists who have raised questions about the robustness of the data based on the new methodology are quick to say this is not the solution. There is nothing wrong with the methodology, some glitches need to be ironed out, goes the general consensus. “The framework of the methodology is in line with the best global practices,” says D. K. Srivastava, chief policy advisor at Ernst & Young.
Devendra Pant, chief economist at India Ratings echoes that view, even as he points to some problem areas. Whenever there is a shift to a new base year, the size of the economy generally goes up because newer developments are now being captured. But in the case of the new series, the size of the economy was lower in 2011-12 and 2012-13 than under the earlier series. In 2013-14, the economy was only very marginally higher in the new series. He is also not convinced that the new series is able to capture the effect of deflation on corporate performance.
But why was this change necessary? India had been one of the very few countries calculating GDP at factor cost. The shift to the basic prices approach was mandated by the United Nations System of National Accounts in 2008. India tarried because, as Pronab Sen, chairman of the National Statistical Commission pointed out in this interview last year, it disrupts comparability. But work on it started at least couple of years before it was finally adopted, so there is really no room for silly insinuations (including by the Congress during whose rule the work was initiated) that the numbers are designed to help the current government.
“This methodology is way superior to the way GDP was being calculated earlier; the changes made are fundamental,” says Sen, who heads a committee that the government set up in June to review the problems in the new series. Teething problems are inevitable, he says, pointing out that the earlier series had the stability of age. The data flow was the same but it was not as accurate as what is being captured now. It is not as if the earlier methodology was never reviewed or modified, but that involved only tinkering that the margins, he says.
Since the methodology is dealing with new data sets, there will be problems with classification, Sen admits. Take, for example, firms which may be into both manufacturing and trading. They are classified on the basis of the largest source of revenue in a particular year, so may end up switching between categories. (This may not happen with large conglomerates which have separate companies for different activities.)
There will, he says, be large revisions in data over the next couple of years. Indeed, the 2014-15 GDP data that was released on Friday, saw the figures for that fiscal being revised downwards. While the primary and tertiary sector GVA figures saw an upward revision (the former significantly and the latter marginally), the secondary sector figures were revised downwards. Quarterly data is based, says Sen, on advance filings; that is the only data available. This is not final data and could change when audited results come in.
It is the data relating to manufacturing that has invited the maximum criticism. The index of industrial production (IIP) figures show a weak manufacturing recovery, something that does not come out in the quarterly GVA figures.
The IIP, explains Sen, measures output or volume. Value added has to be inferred from that. The IIP also has a fixed base – the output of only selected companies in a particular base year is measured. It does not take into account the output of new companies.
There are also some conceptual problems, he concedes. In the earlier series, data from the Annual Survey of Industries (ASI) was used for estimates on manufacturing. This gave both value and volume, so there was no need for price deflators. The new methodology uses the MCA-21 database of the ministry of corporate affairs, which gives only value data. That means, says Sen, one has to be sure that the price deflator being used is accurate. “The price deflator should be linked to the MCA-21 database. That link doesn’t exist.”
“There is a disconnect between the macro numbers and what is happening on the ground, especially in the manufacturing and construction sectors,” says D. K. Joshi, chief economist of Crisil, but he is confident that this will disappear in a year or two. All countries face these initial problems with data, he notes, and it always takes some years for a series to stabilise.
Srivastava feels the CSO must do a back-series calculation based on the new methodology. This, he says, will throw up consistency issues on the basis of which the problems can be sorted out.
All that is work in progress. “We are on the way to a much better series,” says Sen.
Unfortunately, with growth numbers becoming a hugely political issue, controversies are inevitable. It might be best for ruling party politicians and Modi fans to refrain from jingoistic flaunting the numbers as a proof the promise of achche din, as well as opposition politicians and Modi’s many detractors to refrain from completely rubbishing them, in the process calling into question the integrity of government statisticians. But that may be asking for too much, sadly.