Opinion Politically opportune data

Not surprisingly, this has been seized upon by politicians. The prime minister quickly used this in the UP election speeches to mock the Harvard and Oxford economists (read Amartya Sen and Manmohan Singh) who had predicted negative growth impact. The Congress has questioned the veracity of the data, bringing up one more example of an institution that really should be above the political fray getting involved in it.

Advance estimates are, by definition, rough exercises, based on only a subset of the usual indicators, and typically get revised, often drastically. Placing too much significance on them is not a wise idea, although it may be a politically fruitful one. Also, ever since the CSO changed its methodology for deriving the GDP figures, there have been questions about its comparability with the past data and whether this is the correct way of capturing the reality. These concerns are intensified by the lack of back series of national income data based on the new method.

Some part of the estimates is uncontroversial. For example, agriculture was expected to do much better following the excellent monsoon after a succession of bad years; if anything the projected growth rate of 4.4 per cent is less than anticipated. Similarly, a big part of the GDP increase has come because of public administration, defence and other spending, which grew at 11.9 per cent in the third quarter, and contributed nearly a quarter of the year-on-year increase in GDP.

But industry — particularly manufacturing — is a surprise. According to the CSO, manufacturing output grew by as much as 7.7 per cent in constant prices and 10.3 per cent in current prices in October-December 2016, compared to the same quarter of the previous year. What about all the stories of depressed demand, small factories closing and workers losing jobs, then? Well, the CSO’s calculations do not look directly at informal output. In fact, with the new methodology, the CSO does not derive its data from the Index of Industrial Production as it did earlier, but relies on gross value added from company reports, provided by the Ministry of Company Affairs. The quick estimates use an even smaller subset, using data only for listed companies. This excludes all unregistered companies and informal manufacturing producers, along with a substantial number of registered firms that are not listed on the stock market.

Clearly, this is not representative of the entire manufacturing sector. And it does not square up with the estimates of industrial production produced by the same CSO. The Index of Industrial Production shows that manufacturing output in the period October to December 2016 increased by only 0.2 per cent over the previous year. Yet the CSO finds more than 10 per cent increase in nominal output for the same manufacturing sector.

The general public does not have access to the data used by the CSO in its calculations, or to the list of the manufacturing companies it has used to generate these higher output figures. But company results published for the third quarter suggest different conclusions. According to the Dalal Street Investment Journal, year-on-year sales of 4,220 companies (including companies in energy sectors that performed better than manufacturing) in the third quarter of this year increased by only 5.74 per cent — significantly lower than the 10.3 per cent increase in nominal production declared by the CSO.

There is a surprising claim of revival of investment, which is said to have increased by 3.5 per cent after several quarters of decline. The expenditure quick estimates are actually derived directly, but from the production side. But the production of capital goods actually declined by 7.8 per cent over the same period. Similarly, real private consumption expenditure supposedly increased by more than 10 per cent year on year in the third quarter, but production of consumer goods fell by nearly 10 per cent according to the Index of Industrial Production.

The consensus among many observers is that advance estimates are bound to be inaccurate, and so it is better to wait for the revised estimates to get a real sense of what is going on. But the advance estimates are what get used politically — no one other than boring economists really pays attention to the revised estimates, which in any case come out much later. A cynic would predict a downward revision of the current advance estimates that would hardly get noticed a year from now. But it would provide a politically convenient lower base, from which to project a high growth rate in the next round of advance estimates for the coming year.

Post-truth now comes in many different forms, perhaps the use of official data could be just one of these. But for the health of both the society and the economy, it is important to insulate our statistical agencies from even the possibility of such manipulation.