The effect is that the methodology produces great deviation for BC population, because BC listing is very recent at the Central level (1993) and in northern states. Many BC persons are unfamiliar with it. This is compounded by the slowness of listing by some states. That is why West Bengal’s BC percentage is as low as 6.2 per cent, which is unlikely considering the much higher percentages in neighbouring states.

In the identification of BC Muslims (specific castes and tribes of all religions, including Islam and Christianity, have always been included in BC and ST lists), the NSSO’s figure of 31.7 per cent in 1999-2000 and partly corrected to 40.7 per cent in 2004-05 is much lower than reality.

For example, how can it justify 56.8 per cent for Muslim OBCs in 1999-2000 going down to 52.7 per cent in 2004-05 for Karnataka when, as noted by Sachar, Karnataka, like Kerala, has included the entire Muslim population as a sub-category of BCs and the Central List has excluded only nine castes/communities of Muslims who are not socially backward.

The wide variations between the two rounds are evidence of the yet incomplete process of correction of NSSO’s errors. This is also true of the NSSO’s overall BC figure tom-tommed by Karan Thapar and others. When errors are eliminated and maximum precision is introduced, the figure will rise to that estimated by Mandal.

The matter can be settled once for all only by an up-to-date census. The situation needs an immediate special census operation, without waiting for 2011. Evasion of this necessity cannot be visited upon the BCs by seeking to deny them their right to reservation.

... contd.