There is no sleight of hand in the shift in the base year per se, from 1980-81 to 1993-94. With 1980-81 as the base, the GDP growth for 1997-98 was 5.1 per cent; with 1993-94 as the base, last year's growth has been kept at 5 per cent. Yes, the base year change has brought new industries (established since 1980-81) into reckoning; they have been assigned weights at the expense of old and sunset industries. It is possible that the new industries have beenfast-growing. Thus, with the 1980-81 base, manufacturing rose by 5 per cent in 1997-98; with 1993-94 as base (incorporating new industries), manufacturing growth last year was higher at 6.8 per cent. But this declined by nearly a full percentage point to 5.7 per cent in 1998-99. The CSO data do not defy recession

And jobless growth stands out in the CSO's sectoral estimates. With the new base year, the growth of the services sector, estimated at 13.4 per cent for 1997-98 (8.3 per cent on the old base), is down to 5.8 per cent. The sharp drop in the growth of services shows that employment growth suffered a set-back in 1998-99. In its new computation CSO has included, partially (so it is reported) the unorganised sector: this is a highly labour intensive sector. This inclusion makes the decline in industrial growth worrisome. Whatever advantage the government may derive from the better than expected growth of the GDP (to show that the fiscal deficit is not out of control), is nullified by the evidence ofsharply slowing growth of employment. Not just in services and industry, in the other employment intensive sectors too: construction and financial services in particular. The poor were hurt by the slow down; so were middle class job seekers.

What boosted the GDP in 1998-99 is the high growth of agriculture. But this is not as good as it seems, for the growth has been computed over the agricultural output decline of last year. No wonder rural demand is not fuelling the growth of manufactures.