GDP growth numbers have generally brought some cheer,especially in recent years,as the Indian economy withstood the global slowdown much better than most other large nations. But early celebrations after the release of the more recent quarterly GDP figures this week,which estimated growth at an better than anticipated 8.8 per cent,dissipated just a day later,when analysts pointed out some glaring inconsistencies in the basic numbers in the Central Statistical Organisations four-page release. The charges were so striking that they cast serious doubts on the accuracy of the GDP numbers,and the very integrity of the organisation.
The analysts arguments were simple. Look they said,at the numbers for GDP growth at factor cost estimated by collating the growth of production in the different sectors of the economy like agriculture,industry and services,and including their various sub-sectors. These numbers,for supply-side growth,were a high 8.8 per cent. Yet there was an immediate need to check their veracity. Why? Because the growth in overall demand in the economy,reflected in the GDP growth in market prices collated from the growth of consumption and investments in both the private sector and government did not fully tally with the supply-side numbers. In fact,it was just less than half,at a meagre 3.7 per cent. It is impossible for supply to grow at double the pace of demand,especially when you have already accounted for the effects of external trade.
But unfortunately such arguments carried little weight,as disparities in the consumer and wholesale price indices were only marginal; both hovered around the double-digit mark. But the differences in the deflators used to move from the raw numbers to real growth were substantial. In fact the CSO numbers indicated that,while the deflator used to estimate supply-side real GDP growth was a realistic 12.9 per cent given the price scenario during the period the deflator used for calculating demand-side real GDP growth was a disproportionately large 21 per cent.
Given such blatant inaccuracies the CSO had no option but to rework the numbers. The excessively large deflator used for estimating real GDP growth at market prices has now been revised down to 14.8 per cent,which is closer to the other deflator. And this has allowed demand-side growth to move close to supply-side growth,with the real GDP growth at market prices going up sharply to a more respectable 10 per cent much higher than the 8.8 per cent real GDP growth at factor cost.
The question then is: why did such large errors go unnoticed in an organisation which in its heyday was compared to the best in the world? Whatever may be the real reason,which may never be fully known,this incident once again highlights the steady deterioration in the CSOs professional capabilities. Data problems have already been observed in other indicators,like the one that measures industrial production.
And some of the early signals of deterioration in the quality of national accounts data have already been documented. For instance,a comparison of the GDP growth rates of India,China,Indonesia,Korea,Taiwan and Thailand by the Asian Development Bank showed that the size of the GDP revisions in India tends to be larger than of the other economies. And another study,by the IMF of 10 Asian countries,showed that the highest quality of national accounts data was that of Japan,followed by Korea,Thailand and Philippines. Indias data quality was lower,similar to that of countries like Indonesia,Pakistan and Mongolia. The study noted that,though the integrity of the data was still an important strength of these countries,methodological soundness,accuracy and reliability had become challenges. So the errors in the quarterly GDP estimates are only the most recent validation of such charges.
The writer is a senior editor at The Financial Express p.raghavan@expressindia.com
