Happy estimates

Government predictions must not be over-optimistic

The Quick Estimates of India’s economic growth, released by the Ministry of Statistics and Programme Implementation on Tuesday, have revised the GDP growth figure for 2010-11 downwards slightly, to 8.4 per cent from 8.5 per cent. This is the figure for the growth in GDP at constant prices; 2010-11 growth in GDP at current prices is much higher, at 17.5 per cent, reflecting an inflationary environment. The downward revision is, fortunately, relatively marginal. But that makes it an outlier in terms of India’s headline data series recently. The Economic Survey and Budget presented last year, in particular, have been revealed as far too optimistic, projecting growth figures (and restraint in expenditure) that turned out to bear little relation to what actually happened. They estimated growth for 2011-12 at between 8.75 and 9.25 per cent. In actuality, the Reserve Bank of India just revised its estimates for this fiscal year’s growth to seven per cent. A two per cent difference is too large for a standard margin of error.

This takes on particular significance, as much more than arcane details of statistical methodology are at stake. The larger issue of data integrity is of great importance when the government relies on it extensively in the Budgeting process. If the government indulges in overestimation as a matter of course, then Budgets will not respond properly to the country’s output squeeze, or to the state’s expenditure explosion. Last Budget season’s predictions for this year are a case in point. The methodology relied on savings rate calculations, and the proportion of investment devoted to building up infrastructure. A simplistic, Soviet-style calculation of the incremental capital-output ratio or ICOR to estimate growth can then be used. This process is capable of generating innumerable errors and distortions. Long-gestation projects, for example, automatically raise the ICOR, complicating the process of growth estimation. Growth theory has taken many strides since the ICOR was developed, and for good reason; the government’s statisticians and economists need to update their back-of-the-envelope methods of calculation.

It is widely agreed that India is entering a growth slowdown. The International Monetary Fund has estimated India’s 2012 growth to be seven per cent. Growth in 2011, it says, was around 7.4 per cent. The World Bank, meanwhile, expects the Indian economy to grow by just 6.8 per cent in 2011-12. This is significantly lower than the finance ministry’s estimates, which centre on 7.5 per cent. This trend, of a divergence in predictions, has been observed before; and it is not usually the finance ministry that is correct. In exculpation, some claim that nominal growth is as predicted — but citizens will hardly be satisfied with a high-inflation, low-real growth environment. In the Budget for 2012-13, it is to be hoped that this tendency towards over-optimistic predictions comes to an end. Questions will and must be asked about the bases on which any projections are made, especially if they are noticeably out of line with independent, non-governmental estimates of growth in the coming year.