GDP: WHAT’S IN A NUMBER?

The Central Statistical Organisation (CSO) of the Government of India has come out with some nice GDP growth numbers for the third quarter of 2016-17. Prime Minister Narendra Modi has taken the numbers to heart and is taking credit for great economic management by his team. While Modi’s “hard work” is competing with the Opposition’s “Harvard”, those in the government who cannot take credit for either are keeping mum. So what is the real situation?

The main issue is not whether the GDP growth figures are actually higher or lower, but that common everyday experience is not reflected in the numbers released by India’s statistical establishment. In such situations, it becomes the duty of the government’s experts to provide credible reasons, but they haven’t done so. Demonetisation could have theoretically led to higher growth, just as theoretically it could have led to lower growth. Either way, reasons—backed by numbers—need to be given. Over a period of time, it is credible reasoning that builds faith and trust in the government and its statistics.

It took many decades of number-crunching to build a credible reputation for government data. This reputation is being destroyed by the silence and mechanical practices of government statisticians and economists. Again, the point is that India and the world need to understand what happened, and how. A simple, mechanically produced number is not what the world needs. And so, in hushed tones, people in Mumbai are talking about India’s numbers being as difficult to believe as China’s. Those talking in Delhi are even more hushed.

The prime minister is rightly taking credit for the numbers, as any good politician would—and should. After all, by taking credit, he is also taking responsibility for India’s economic performance. But I wonder whether even he knows what is happening.

A large share of GDP growth comes from sectors where data is not available in time for the CSO to include in its quarterly estimates. Therefore, it takes the previous year or quarters’ data to impute estimates for the latest quarter. This is simply a form of imputation, and some may say is even standard. A proactive group of statisticians would not blindly use older parameters; they would instead incorporate estimates from quick dipstick studies. But that is not how India’s Central Statistical Organisation functions—it is more mechanical.

Typically, the units whose numbers are so imputed are those in the unorganised sector, as well as smaller units. If demonetisation and digitisation had helped such units, growth would be higher than the current report suggests. If they hurt these units, growth would be lower than reported. Demonetisation and the consequent digitisation of transactions probably did hurt, as structural shifts of the scale that occurred tend to hurt in the short run. So, after a few quarters, the third quarter’s GDP growth numbers will most certainly be revised downwards.

Now here is the kicker—even after the growth numbers are revised, there is no way to tell whether demonetisation helped or harmed India’s economic growth. This is a structural shift. India is most certainly on a new path of high-intensity digital transactions, but many other factors also play a role, making it difficult to figure out the net effect in the long run.

Either way, it is the government’s duty to share the truth with the public. I would have loved it if the statisticians and advisors to the prime minister and government had said, “These are early numbers—the actual picture of quarterly growth will be revealed later, and we expect it to be lower”, before they released these mechanically calculated estimates. That is not the prime minister or finance minister’s job, it is the statisticians and economists who have to take on the responsibility.

The writer is an economist and director, Indicus Foundation

It took many decades to build a credible reputation for government data. This is now being destroyed