Slow growth? We’ll raise it

If the findings of a recent study on the country’s national accounts are anything to go by, neither the Prime Minister nor his Finance Minister need to be unduly worried about the year’s poor economic growth. Given the wide, and increasing, discrepancies in estimates, it’s relatively easy to present almost any kind of number for GDP, investment, consumption, or anything else that you desire. And while it would clearly be unfair to assume that the Central Statistical Organisation (CSO) will hike its estimates upwards deliberately, or that it is even being asked to do so, it is true that the CSO’s estimates in the recent past have been the subject of great controversy. It is well-known, for example, that when the CSO put out its estimates of GDP growth of 6.8 per cent for last year, other agencies such as the private sector Centre for Monitoring Indian Economy’s (CMIE) estimate was a much lower 6 per cent. Few private sector economists, in fact, thought the CSO numbers were accurate. It is also worth mentioning that the CSO revised it’s own estimates for GDP growth for 1994-95 from 5.3 per cent initially, to 6.3 per cent and finally to an incredible 7.2 per cent. Figures for 1995-96 also changed with almost the same frequency. At the time the CSO put out its revised estimates, in fact, several private sector economists had pointed out that there were serious flaws in the estimates made by the CSO some like S.P. Gupta of the Indian Council for Research on International Economic Relations (ICRIER) were even on the panel of groups formed by the CSO to help make these estimations. But they were obviously not been consulted while such numbers were put out.

Gupta’s arguments, in fact, were summarised in a news-article in The Indian Express as far back as January 31. Citing just one of them will illustrate the serious nature the discrepancies in the estimates. Gupta pointed out, for example, that the CSO estimates for 1995-96 implied that households were keeping 55 per cent of their savings in the form of physical assets like houses and gold. This, however, flies in the face of the fact that the share of physical savings has been falling constantly and was a mere 40 per cent in 1994-95 ! A study done by the Economic and Political Weekly’s Research Foundation (EPWRF) a couple of weeks ago lends further credence to arguments made by economists such as Gupta and cites countless examples of discrepancies in the various estimates made by the CSO and the National Sample Survery (NSS).

Frankly, if the implications of their findings weren’t so serious, you’d want to laugh. Estimates of the public sector output in the mining sector, for example, are higher than the estimates of total production in the country ! For 1994-95, for instance, public sector output in the mining sector has been estimated at Rs 22,750 crore (at current prices) for the economy as a whole, the production, however, is estimated at a lower Rs 19,669 crore. Discrepancies in the figures for expenditure on consumption are even higher, and are increasing with each passing year. In the ’70s, NSS data showed roughly 5 to 12 per cent lower consumption than CSO’s did. This difference increased to 21 per cent in 1983-84 and further to 27 per cent in 1987-88. The going gets better, if you can call it that, as one goes along.

Story continues below this ad

Capital formation — what the average person calls investment — varies by as much as Rs 37,670 crore, or around 3.4 per cent of GDP for 1995-96. And no, this difference is not just due to the fact that two separate agencies are doing the estimation. This difference often arises from the same data-set. So, if one is to add up the gross capital formation of various industries, you get one figure. If, however, you add up the value of the assets created in the same period, you get a different number. As in the case of consumption, this difference has also been rising — from 1.9 per cent of GDP in 1980-81 to 2.9 per cent in 1990-91.

Similar increasing discrepancies are also seen in other numbers such as the GDP. In 1980-81, for instance, the GDP estimate based on production estimates of various sectors was roughly 0.7 per cent higher than the estimates based on expenditure. This rose to 3 per cent by 1990-91 and to as high as 6.6 per cent in 1994-95. In fact, if one is to take the higher estimate of GDP, and use this to calculate the investment, this then shoots up to a whopping 32.6 per cent of GDP for 1994-95 instead of 26 per cent as reported by the CSO.

Given this background, it is almost certain that this year’s CSO estimates will be greeted with the same degree of scepticism as last year’s and the year before that. What’s worse, is that this problem continues despite the fact that several experts have given their recommendations on what needs to be done. Ironically, the CSO is currently working on its database with experts from the IMF, but this pertains to just reducing the time lag with which data is available, and not to the quality of the data itself.