A false growth rate

By Prem Shankar Jha

The mystery surrounding India's growth rate deepens every time the Government, or the Central Statistical Organisation announces a new statistic. Only on January 28 week the CSO revised its 'quick' estimate of the growth of GDP in 1998-99 from 5.8 per cent to 6.8 per cent. The deus ex machina was the rain god who helped farmers to reap record autumn (kharif) and spring (rabi) crops. This boosted the rate of growth of agriculture to a record 7.2 per cent. Needless to say, the Government was overjoyed. It was, after all, the first year of the BJP's rule at the Centre.

A week later the CSO announced that its 'advance estimate of growth rate for 1999-2000 was 5.9 per cent. This did not , however, please the Government , for it fell below the minimum acceptable figure of six per cent. Thus only a day later, Mr. Yashwant Sinha, the Finance Minister, expressed the hope that timely winter rains would boost the rabi output and enable growth to touch 6 per cent after all.

It would therefore seem that despite all domestic and external setbacks the Indian economy has performed exceedingly well since the mid-Nineties. After a single year of post crisis stagnation, when the GDP grew by only 0.86 percentage point, the growth rate picked up to 5.3 per cent in 1992-93, 6.2 per cent in 1993-94, 7.8 per cent in 1994-95, 7.2 per cent in 1995-96, and 7.5 per cent in 1996-97. Even in the recession year of 1997-98 the growth rate did not fall below 5.1 per cent - a full 1.5 percentage point above the average growth rate for the two decades from 1956-57 to 1974-75. The average rate of growth for the post crisis period therefore comes to a very healthy 6.5 per cent.

This impression flies in the face of a host of other statistics, notably those of industrial growth. Till 1996-97, the rise and fall of the GDP growth rate was largely a reflection of the rise and fall of industrial growth and to a lesser extent, that of agriculture. From 1996-97, the two went completely different ways. Thus the rate of growth of GDP was within one per cent of the growth in agriculture and industry from 1990-91, till 1994- 95, and again in 1996-97. Only in 1995-96 and then from 1997-98 onwards has the GDP growth rate systematically exceeded the average for agriculture and industry. It did so by 3.5 per cent in 1995-96, 3.3 per cent in 1997-98, 1.2 per cent in 1998-99 and bids fair to exceed it by two per cent in 1999-2000. This has been made possible by colossal rates of growth in the services sector ranging from 8 to 8.5 per cent , year after year, and as high as 10.4 per cent in 1995-96.

It would be tedious to quote still more figures and turn this article into a statistical exercise. Suffice it to say that the GDP figures for the Nineties reveal an unmistakable tendency to overestimate the growth in the services sector, which has imparted an upward bias to the growth rate throughout the last three decades. This has become particularly pronounced in the three most recent years, 1997-98 to 1999-2000.

The upward bias is revealed by a strange anomaly in the service sector's contribution to the GDP. Between 1970-71 and 1996-97 this grew by more than ten percentage points at current prices 10.04 per cent from 32.91 per cent to 43.03 per cent. But it grew by nearly 11 per cent at constant prices during the same period. While such a sharp rise in the share of services in the GDP at current prices in a bare two and a half decades is within the realms of possibility, an equally sharp rise in its share at constant prices is almost impossible. What makes nonsense of economics is that the share of services in the GDP at constant prices should rise faster than the share at current prices. The reason is that in every single industrialising country in the world, the share of the services sector in the GDP has risen fairly sharply over time at current prices but the share at constant prices has tended to remain unchanged or to rise very slowly. The rise at current prices occurs because productivity in the services sector rises much more slowly than in the manufacturing sector so more manpower is needed to store, distribute and perform other services related to the national output, than simply to produce it. But wage rates in the services sector tend to get pulled up more or less in tandem with industrial wages. This ensures a rise in the share of services in the GDP at current prices.

If statisticians had developed a perfect way of measuring the value of services at constant prices this 'inflation' would have disappeared completely. But it is extremely difficult to measure the 'real' value of services. What the anomaly in the data reveal is that the Central Statistical Organisation has virtually stopped even trying to do so. That is why it has uncritically taken the 76 per cent increase in Central Government pay packets in 1997-98 and 1998-99 and the Rs. 15,000 crores of pension arrears as increases in real GDP. Worse still, since the State governments are only now paying out the salary and pension increments and arrears, this service sector inflation of GDP will continue for the next two or three years.

The only way to get a realistic assessment of the actual rise in national output therefore is to take the weighted average of the growth in agriculture and industry, allow for a half per cent per annum higher growth in the services sector, and then calculate the GDP as the weighted average of growth in the three sectors. This will give the actual growth rate as 3 per cent in 1997-98, 6.1 per cent in 1998-99, and if the Government's forecast for agriculture and industry this year comes true, 4.4 per cent in 1999-2000.

Send this article to Friends by E-Mail

Section  : Business
Previous : Hyundai to position Sonata in Rs. 12 lakh category
Next     : Zydus Cadila to tap market