The controversy surrounding India’s gross domestic product (GDP) series refuses to die down. Since 2015, GDP numbers have drawn a fair amount of criticism by way of their accuracy and reliability being questioned.
The criticism falls under many categories - such as, the choice of methodology used by the Central Statistical Organisation (CSO), the reliability of the data used, notably MCA 21, dissonance with the falling credit growth, and rising unemployment.
The controversy has escalated further following claims by Dr Arvind Subramanian that GDP growth in the post-2011 period is overestimated by 2.5%. Importantly, this is not the only study on the subject. EPW carried a series of articles on this issue.
Then, Dr Subramanian Swamy in an op-ed in the Hindu
in 2015 questioned the GDP growth on technical grounds: “Finance ministry had chosen to use the Paasche’s Index instead of the Laspeyres Index to calculate the growth rate, which, under the present inflationary conditions, will artificially raise growth rate figures.”
This author tried to connect the dots in his study
“What is the True Growth Rate: Reconciling the Old and New GDP Estimates
” in 2016 which was shared with the government.
Discourse on this topic has been reduced to a cacophony. As a result, what the CSO did and how it impacted the GDP figures, is all but forgotten. Therefore, just to keep the focus; CSO’s exercise of 2011, in effect, brought about two important changes.
The change of the base to 2011-12 was merely statistical, the adoption of the United Nations System of National Accounts (UN SNA) 2008 standard was methodological. The latter introduced a new concept and altered the very prism through which economic activity in India is looked at and understood.
Economists deemed the study of standards as hopelessly insignificant and banal, unworthy of their attention. However, as German philosopher Friedreich Nietzsche said on the power of a representation system: "the reputation, name and appearance, the usual measure and weight of a thing, what it counts for –originally almost always wrong and arbitrary – grows from generations unto generations, merely because people believe in it, until it gradually grows to be a part of the thing and turns into its very body."
Standards are also a type of representation system that fix the world view of an observer. In fact, a standard can be a representational system with an embodied meta-ideology that provides the frame of reference of a discourse and defines the playing field upon which a specific scholarship evolves.
Meta-ideology of standards may not be of serious concern in a technology domain but it does have implications if it is in the social and economic domain. Thus, it is in these technical details of UN SNA that one can locate some of the answers to this controversy.
United Nations Statistics Division (UNSD) has been setting standards on how to compile the national accounts statistics since the Second World War
. SNA is necessary to ensure cross country comparisons. Even the International Monetary Fund (IMF) conditionality insists that countries adopt UN SNA standards.
Since 1947, four major standards have been provided by the UN, namely the 1953 SNA, 1968 SNA, 1993 SNA and the 2008 SNA. The latest SNA is to accommodate structural changes in the world economy due to the rise of the digital economy and other factors.
There is no uniformity in the adoption of the UN SNA among member countries. The pace of adoption is a function of the available financial resources, qualified manpower, political support and availability of the necessary basic economic statistics, etc.
Thus it is pertinent to take stock of the progress of adoption of UN SNA in member countries as it will be useful later in evaluating Dr Subramanian’s findings.
In a presentation “10 years since the adoption of the System of National Accounts 2008: Status of implementation” Ilaria Di Matteo of UNSD notes that 180 (≈93%) member states have more or less complied with 1993/2008 SNA in 2017. In respect of 2008 SNA, 79 (≈41%) member states had fully complied at the end of 2017.
They note that quite a few countries are still lagging in the implementation of SNA. Thirteen member states (≈7%) are still using the 1968 SNA. In respect of the base year, Di Matteo notes that 84 member states (≈44%) are compiling volume measures of GDP using base years prior to 2009, which may result in biased real GDP growth rates.
Whenever a standard is updated, it raises obvious question of backward compatibility. Conscious of this, in 2016, UNSD in a presentation titled “The 2008 SNA: Impact on GDP and implementation status” noted that 62 member states have switched over to 2008 SNA.
Some member states have opted to introduce certain aspects of the 2008 SNA in a gradual manner. For instance: Eurostat, Korea and Singapore implemented the 2008 SNA in 2014, Australia in 2009, India in 2011 and Indonesia in 2015. In the Organisation for Economic Co-operation and Development (OECD) countries the implementation has happened at various points since 2008.
UNSD in their survey covering the Eurostat members, OECD members, Australia, Indonesia, Republic of Korea and Singapore, noted that the total impact of the 2008 SNA on the EU nominal GDP level in 2010 was 3.67%.
Methodological changes accounted for 2.27% and statistical improvements 1.40%. The largest impact in methodological change came from one head, namely, research and development (R&D) at 1.86%. In the OECD countries the average impact of the 2008 SNA is at 3.1%. Real growth rates are generally affected to a much lesser extent in the OECD countries.
But some countries have seen changes in real growth rates such as Australia. In conclusion, there is no definite pattern.
Hence, in the light of the diverse experience of the UN member countries after adopting the 2008 SNA, singling out CSO (or India) and leveling a blanket charge of data cooking is quite unfair and damaging to its credibility. Those who have resorted to this have perhaps not pondered over the ramifications of such a tirade.
One may disagree with the methods employed by the CSO, say for deflation as Dr Swamy noted, or the applicability of the 2008 SNA standards themselves to India, given its current state of development as done by this author; but such a position has scarcely found place in the media discourse or in the working paper of Dr Subramanian.
Coming directly to the assertions made by Dr Subramanian in his working paper
, my contention is that he has used an improper tool to arrive at the estimate of 2.5%. His paper is quite technical and in the interest of making the crux of his argument accessible to an average reader I will only explain his basic methodology.
I will explain this in the second part.
(The writer is an economist in the banking sector. The views are personal)