Census 2027 Will Ask Your Caste - But Not Give You a List. India Tried This Before.

An Old Census Mistake, Rescheduled for 2027 In 1931,

colonial census officials in Punjab found that 418,789 people - nearly as many as the province's entire Christian population - had suddenly started calling themselves "Aad Dharmis," a caste category that had not existed ten years earlier.

Elsewhere, herding communities calling themselves Ahirs, Goalas, Gopis and Idaiyans folded themselves into a single new label, "Yadava." The 1931 Census was not just counting caste; it was watching caste identity reshape itself in real time, because the question it asked was open-ended - write in whatever you call yourself - with no fixed list to check against.

Census 2027 has just made the same choice.

The Pattern Repeating This is not

India's first attempt to relearn this lesson. After 1931, caste was collected but dropped from final tabulation in 1941, then abandoned entirely in 1951 on Nehruvian grounds of secular equality. When the UPA government revived caste counting through the 2011 Socio-Economic and Caste Census, it again used open-ended self-reporting - and produced 46.7 lakh distinct caste names, more than eleven thousand times the 4,147 recorded in 1931.

The data was unusable. The Centre withheld it. Where the parallel breaks down is scale and stakes. In 1931, a colonial state with limited administrative reach could tolerate an untidy caste column because it wasn't feeding welfare eligibility.

In 2027, caste data will sit alongside the Ministry of Social Justice and Empowerment's list of 1,208 Scheduled Castes and a 2022 count of 730 Scheduled Tribes - categories that already determine reservation and scheme access.

An open-ended answer that spells "Rajput" one way in Bundelkhand and another way ten kilometres over, where it might actually denote the OBC Lodh community, isn't just messy data anymore. It's data that misroutes entitlements.

Why India Keeps Choosing the Harder Path Former

ICSSR chairman S.K. Thorat's proposed fix - a predetermined general-category list modelled on existing SC/ST/OBC lists, accepting a 2-3% margin of error - is the same logic that eventually stabilised SC and ST enumeration decades ago.

Bihar's 2023 caste survey used exactly this approach: officials compiled a master list first, precisely to avoid the SECC's chaos, though it drew its own disputes over classification methodology. That India keeps returning to open-ended enumeration despite this repeated failure suggests the resistance isn't technical - pre-built lists exist for SC and ST already - but political.

A predetermined general-category list requires the state to formally decide, in advance, which caste names count and which don't, a decision every 1931-style consolidation movement shows communities themselves are still actively negotiating.

Census 2027 has chosen to let that negotiation happen inside the census form rather than before it. History suggests that choice will cost the exercise its usability, not its accuracy.