One of the untenable arguments against reservation is to question the Mandal estimation, by projection from the 1931 census and drawing on the National Sample Survey Organisation’s survey for 1999-2000, that 52 per cent of population belongs to backward castes (BC). The charge was led by Karan Thapar as Devil’s Advocate in his May 21 interview of Union HRD Minister Arjun Singh and in his article ‘Mandal Muddle’. Others have followed, breezily asserting in their ignorance of historical facts that “no votary of caste-only-criterion is anxious to update these figures.”

The crucial historical fact missed by anti-reservationists is that BC commissions and other bodies have been pressing for inclusion of BC-related population data over the last half-century.

Anti-reservationists have just woken up to this, with the sole Canutian aim of rolling back the inexorable wave of reservation.

NSSO’s sample survey cannot be a substitute for or supersede census, current or old, so far as quantifying, as close as possible, the total population and of any section, including BC population. Otherwise, why should India and other nations persist with censuses when a less-expensive and sample survey method would do for estimating population.

Therefore, the Mandal methodology of projection from the latest available census, is more acceptable and closer to ground reality than sample surveys, which may help in other ways, such as understanding details of many quantifiable and measurable variables between social, economic, and other classes.

There is evidence to show the limitations of the NSSO’s methodology. For example, its surveys of 1999-2000 and 2004-05 show STs in Punjab, Haryana, Delhi and Chandigarh (partly corrected in the latter round) where no STs are recognised by the presidential orders. Similarly, the survey showed SCs in Nagaland, where there are none. One important reason for these aberrations is that classification of social groups by NSSO was entirely based upon the household head’s declaration.

... contd.