View: There are more questions than answers on back series GDP data
Three years ago, India rebased its GDP statistics from base year 2004-05 to base year 2011-12, normally a boring exercise.
But it generated huge controversy, as somewhat magically the Central Statistics Office (CSO) upped the GDP growth rate by 2 percentage points. This made India — not China — the world’s fastest major economy, and the Narendra Modi government lapped up the accolades. There were rumblings of disbelief as the uptick did not match any other real indicators. But the world moved on.
Now, after a delay of three years, CSO, under the guidance of NITI Aayog, has rebased the past GDP series, and the GDP growth rate of the four years of the Modi administration is now higher than even that under the boom years of UPA-.
CSO argues that it has not just rebased the GDP series, but has upgraded the methodology with new data sources to meet UN standards. The new methodology is, no doubt, better. But the assumptions it has had to make to approximate some data, and the slicing and dicing required to recast the ‘back-series’, has left it open to questions. Without full access to all of CSO’s data bases, it’s impossible to know exactly where the problem lies. But it’s not difficult to point to some obvious inconsistencies.
First, the investment rate — defined as gross fixed capital formation over GDP — has declined from a high of around 36% in 2007-08 and averaging 33.4% during the UPA period (2004-05 to 2013-14) to a low of 28.5% in 2017-18 and averaging around 29% during the NDA period (2014-15 to 2017-18).
Second, the rebased series don’t match any of the other economic numbers.
Whether we look at corporate sales, profit and investment numbers, direct tax revenue, credit growth, exports and imports, the performance during the UPA period was willy-nilly much better than in the NDA period and reflective of a much faster-growing economy.
According to the International Labour Organisation (ILO), between 2008 and 2015, India also saw real wage growth of 5.5% a year, the fastest in South Asia. The latest poverty data from the multi-dimensional poverty index, produced by the United Nations Development Programme (UNDP) and Oxford University, show a huge decline in poverty — both relative and absolute —in that high wage growth period.
Third, what is curious about the way the new methodology has been used is that GDP growth rate is higher in 2011-12 — the base year, and every year after that — and is lower in every year prior to that in the rebased series compared to the old series. GoI’s press release does not provide the GDP growth rates under the old series after 2011-12, but these are readily available.
Fourth, we find that the big differences in the growth rates are mainly in the tertiary sector, not so much in agriculture or in manufacturing. The service sector real GDP growth rates are 2-3% percentage points lower for every year since 2005-06. There are some curious assumptions that could have produced this result.
For example, for the unorganised trade sector, sales tax — notoriously underestimated — has been used, instead of the general trade index (GTI), and a new wholesale price index (WPI) has been used to deflate the nominal GDP. In the communication sector, telecom subscriber growth was used earlier, but has been replaced by minutes of usage. It isn’t clear which is a better indicator.
Statistics Don’t Click
Given the controversy over these numbers, it’s best to establish an independent panel to review the entire exercise. Perhaps the venerable statistician, Prof C R Rao, could be asked to come back to head such a panel and re-establish the credibility of CSO and India’s statistics.
In the meantime, let us focus on the fact that investment is down, and future growth will be affected if it does not revive.
The writer is visiting scholar, Institute of International Economic Policy, George Washington University, US