, 24 tweets, 8 min read Read on Twitter
A month after holes in a key database used in GDP estimation were first exposed, the controversy around #India's GDP #data and statistical system has only grown. This #thread is a summary of some answers and many questions on this issue #stats #institutions #economy (1/n)
The latest to question the new GDP series is the former CEA @arvindsubraman who was also among the first to question the new series numbers when they first came out in 2015 (2/n)
His main contention, then and now, is that the new series fails to tally with other economic indicators. Others also have made this point but his recent paper (hks.harvard.edu/centers/cid/pu…) is the most systematic demonstration of that issue (3/n)
As for his 'revised' growth estimates, I think they are indicative rather than definitive. And the definitive answers can only come from an independent review of sources and methods, a fact he also recognizes and recommends (4/n)
One problem that he and others have highlighted relates to the increased reliance on inaccurate deflators in the new series, in which volume based indicators were replaced with value-based indicators (5/n)
The move to value-based estimation in itself is not an improvement as some seem to argue. SNA recommends the use of volume-based indicators to est. real growth, and even if it did not, the use of deflators without first getting and testing the relevant price data series... (6/n)
... is fraught with problems. The economist Rajeshwari Sengupta was among the first to point out these issues in a 2016 Mint oped (livemint.com/Opinion/58qihT…) (7/n)
The economists R Nagaraj and (late) TN Srinivasan had a detailed take on this (and other issues) in this excellent summary paper ncaer.org/events/ipf-201… (they also suggested that several SNA 2008-related changes could have been delayed till better data was collected) (8/n)
The other big problem in the new series relates to the manner in which the MCA-21 corporate database was plugged into the national accounts. The process by which this was implemented... (9/n)
... itself should have raised red flags, given that the independent member on the subcommittee was unaware of the last minute methodological changes (livemint.com/Politics/jKTdO…) (10/n)
But that it did not tells us something... Dr Subramanian is right that the problems with the new series are technical more than political but the fact that these technical problems were not addressed or fixed seems to be a political/governance issue... (11/n)
... If the growth numbers were unfavourable, would they have met the same treatment?
That would have remained a purely hypothetical question had we not witnessed the swift reaction in the case of the #PLFS jobs report, which was immediately criticized and held back. (12/n)
The chickens have come home to roost now, with growing questions on the manner in which the MCA-21 database has been used in GDP estimation after NSSO officials found it to be full of holes (livemint.com/news/india/new…) (13/n)
While official statisticians and govt officials maintain that this does not affect growth rates, they have not revealed details of the methodology or released the raw MCA-21 database being used for GDP estimation, leaving data users unconvinced (14/n)
This piece (livemint.com/politics/polic…) provides an overview of the questions that are being raised on the use of the MCA-21 database, and why some researchers argue that the new method may be overstating growth (15/n)
The debate has led to clarity on one issue: that the use of the MCA-21 database has led to a situation where we don't know how much output is being produced in 'which sector, and in which state' ... (16/n)
... which has serious consequences for sector-specific plans/policy-making as well as for state finances. It has prompted the Finance Commission to step in to 'reconcile' the data (livemint.com/politics/polic…) (17/n)
This also has implications for the deflator problem. If we don't know which part of output is manufacturing and which is services, we would be deflating value of one (apples) using price data for another (oranges) and our real growth estimates will never be accurate. (18/n)
There are other issues. Some are legacy problems, some of which may have got worse in the new series. This data-led explainer provides an overview of the key issues of contention in the new GDP series: livemint.com/news/india/the… (19/n)
As to how we got here, that's a long story involving battles of turf, resources, and egos among technocrats/bureaucrats. And that story is here: livemint.com/news/india/how… (20/n)
For more on this, and how things could get better, here is my piece from two years ago that @CafeEconomics reminded me of livemint.com/Opinion/aMKJ70… (21/n)
In conclusion, I must say that we are not China yet, and there is no evidence yet that growth numbers are getting 'fudged'. However, the lack of competence together with opaque systems and weak oversight have led to numbers that seem out of sync with reality. (22/n)
India's statistical crisis is also an opportunity to reclaim a glorious past. The path we choose now will have important ramifications for our economy and our democracy. The economist and Mint columnist, V Anantha Nageswaran has some wise words on this for the govt... (23/n)
..."Transparent statistical methodology will lend credibility not only to... GDP data, but also to the government. Timely compilation and dissemination of official data is part of good governance," he wrote (livemint.com/opinion/column…).
One hopes they listen to such advice. (24/n)
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Pramit Bhattacharya
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Follow Us on Twitter!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!