How India Plans to Count Caste Now, What Went Wrong Earlier

The Registrar General of India (RGI) has decided to use an open-ended question on caste in the ongoing Census, without giving respondents a predetermined list of caste names.

The 1931 Census

What Happened After 1931?

The 1931 Census was the last comprehensive caste enumeration in India and revealed major difficulties in accurately identifying and classifying castes. Several caste groups attempted to club themselves under broader or newly created identities to increase their numerical and social strength. In Punjab, sections of leather-working Dalit communities identified themselves as “Ad Dharmis”, presenting it as a new religious identity. Similar identity changes such as Adi Dravida, Adi Andhra and Adi Karnataka appeared in other regions. The Census also recorded attempts by several related communities to consolidate under broader identities such as “Yadava”, making caste classification highly subjective.

The 1941 Census collected caste information, but the data could not be fully tabulated because of World War II. In the 1951 Census, independent India discontinued general caste enumeration under the Nehru government’s emphasis on equality and secularism. The 2011 Socio-Economic and Caste Census (SECC) produced around 46.7 lakh distinct caste names, compared with only about 4,147 castes recorded in 1931. The huge number of entries resulted from spelling variations, surnames, sub-castes and inconsistent self-identification, making the data difficult to use. The Union government consequently withheld the raw caste data of the 2011 exercise. The 2023 Bihar caste survey also raised questions over caste lists, classification methods and survey methodology.

What the Numbers Say

• Government lists already contain around 1,208 Scheduled Castes, with different caste entries varying across States.
• A 2022 government release listed exactly 730 Scheduled Tribes across the country.
• The absence of a similarly standardised national list for all caste groups remains a major challenge for an open-ended caste Census.

The Proposed Solution

• Experts suggest preparing a predetermined and exhaustive list of recognised caste names before conducting caste enumeration.
• Separate standardised lists may be needed for SCs, STs, OBCs and castes falling in the general category.
• Experts estimate that a well-prepared list could reduce classification errors to only around 2–3%.
• Census forms could also include options such as “no caste” and “no religion” for respondents who do not identify with these categories.
• Separate questionnaires have been suggested to capture the social exclusion and deprivation experienced by SC and ST communities.
• Questions on internal hierarchies and unequal deprivation within SC and ST groups could also help debates over sub-categorisation.

An entirely open-ended caste question may reproduce the problems of earlier exercises unless India develops standardised caste lists, clear coding rules and a robust classification methodology.