Thread by @JustinSandefur

Is India’s Consumption Really *Falling*?

New op-ed by @FelmanJosh @arvindsubraman @duggan_julian and me.

business-standard.com/article/opinio…

Ungated + data/code:
cgdev.org/blog/indias-co…

1/

@FelmanJosh @arvindsubraman @duggan_julian A bombshell report a few weeks ago, with leaked results from the 2017/18 NSS round -- which the government has been withholding for months -- showed that real per capita consumption across India had fallen by 3.7% since 2004/05 (!)

2/

@FelmanJosh @arvindsubraman @duggan_julian That's negative real consumption growth over the same period that official national accounts figures report 7% consumption growth per annum.

Somebody is very wrong.

3/

The debate has broken into two camps:

Some see the NSS survey as proof that India's economy is cratering, and that the combination of demonetization and GST just before the 2017/18 survey took an immense toll on Indian households.

4/

Others -- more defensive of India's headline economic growth rate -- dismiss the survey data, claiming that that false-response rates are growing, and the NSS no longer captures a meaningful share of Indian consumption.

5/

In this piece, we basically argue for a middle ground. Nat'l accounts and surveys have diverged enormously since 2011/12, and both are hard to reconcile with other economic indicators.

The truth probably lies somewhere in between.

6/

.@arvindsubraman has written extensively about the over-optimism of the national accounts here:
hks.harvard.edu/centers/cid/pu…
and
hks.harvard.edu/centers/cid/pu…

And @duggan_julian and I wrote about the same trouble with new 2019/20 numbers here:
cgdev.org/blog/if-india-…

7/

@arvindsubraman @duggan_julian But that doesn't mean we should embrace the leaked NSS consumption numbers uncritically.

We calculated what C growth over this period might be if @arvindsubraman was right, and GDP growth was 2.5% lower than officially reported. Answer: C still growing at +3.7%.

8/

In sum, there is no doubt that India's growth has slowed -- perhaps much more so, and much earlier than official figures attest. But it seems premature to conclude consumption is in absolute decline.

9/

We draw only one firm conclusion:

The govt should release the 2017/18 survey, no matter how inconvenient. Let researchers pore over the micro data. There's a lot to learn from imperfect data -- both about the economy retrospectively, and better measurement going forward.

10/10

Typos:

In #3, NSS showed negative real per capita C growth since 2011/12 not 04/05.

In #4, to clarify: -3.7% is per capita. We calculate total C growth in NSS at +1%, which is the comparable number to natl accounts 7%.