Checkbox caste: On the counting of caste, Census 2027

The rehearsal for the second phase of Census 2027 , under way in 16 States and Union Territories since July 6, carries a key feature: an “open column” where respondents can state their caste , which the enumerator will record. Unlike the 2011 Socio-Economic and Caste Census (SECC) , which also had this feature, this counting of caste in the Census itself has statutory backing. The pre-test ends on July 20, and the government says that it will then finalise the methodology for counting caste. The hope is that the pre-test findings will corroborate what is known: an open-ended response on caste yields only unwieldy data, as seen in the 2011 SECC, which eventually proved unusable. It is not difficult to understand why. The method led the 2011 SECC to return more than 46 lakh “caste names”, against the 4,147 in the 1931 Census, the last to tabulate caste. Respondents entered surnames, sub-castes and clan names as if interchangeable, inflating the count into incoherence. The Centre told the Supreme Court in 2021 that the SECC figures were too error-ridden to be relied upon for reservation. The pre-test should instead point to a better method — using the digital Census’s hand-held devices, pre-loaded with a curated list of castes and sub-castes, so that the enumerator selects the ‘correct’ entry after asking the respondent. Mistakes and mismatches will happen, but as the 2022-23 Bihar caste survey revealed, this method could return more usable data.