Caste Census

A caste census is the enumeration of a population disaggregated by caste, recording each respondent's specific jāti as a data field alongside age, sex, religion, and occupation. In India the practice is rooted in the colonial decennial censuses conducted under the Census of India operations from 1881 onward, when British administrators recorded caste comprehensively as an organising category of imperial governance. The full caste enumeration of all groups was last completed in the 1931 Census under Census Commissioner J. H. Hutton; the 1941 caste data was collected but, owing to the Second World War, never fully tabulated. After independence, the Government of India decided in 1951 to discontinue the enumeration of caste except for the Scheduled Castes and Scheduled Tribes, whose count is constitutionally necessary under Articles 330 and 332 for the reservation of seats in the Lok Sabha and state legislative assemblies. Consequently, every census from 1951 onward has counted SCs and STs but not Other Backward Classes (OBCs) or the wider universe of jātis.

Procedurally, the decennial census is conducted by the Office of the Registrar General and Census Commissioner of India under the Census Act, 1948, which empowers enumerators to collect information and binds respondents to answer truthfully while guaranteeing the confidentiality of individual records under Section 15. Enumeration proceeds in two phases—houselisting and population enumeration—with caste, where recorded, captured as a free-text or coded response. The difficulty of a comprehensive caste census lies precisely in this coding: India has thousands of jātis with overlapping, regionally variant, and synonymous names, no single authoritative master list, and no constitutional definition of "backward class." The 1931 exercise grappled with the same classification problem, which is one reason post-independence governments cited administrative impracticability for abandoning it.

A distinct and instructive variant is the Socio-Economic and Caste Census (SECC) of 2011, conducted not under the Census Act but by the Ministry of Rural Development and the Ministry of Housing and Urban Poverty Alleviation, with caste data collected by the then Ministry of Social Justice and Empowerment. The SECC was the first caste-inclusive enumeration since 1931, but its raw caste data—reportedly listing some 46 lakh distinct caste and sub-caste entries—was deemed riddled with errors and never officially published. An Expert Group under former NITI Aayog Vice-Chairman Arvind Panagariya was constituted in 2015 to classify and validate the SECC caste returns; its work did not yield a usable, released caste count. The SECC's socio-economic deprivation data, by contrast, was released and used for beneficiary targeting.

The contemporary politics of the caste census intensified after several state governments commissioned their own surveys. Bihar conducted a caste-based survey in 2022–2023, releasing results in October 2023 that enumerated the state population and reported Extremely Backward Classes and OBCs together at roughly 63 percent. Karnataka commissioned a Socio-Economic and Educational Survey in 2015 under the H. Kantharaj-led State Backward Classes Commission, whose findings remained contested and largely unreleased for years. Telangana conducted its own caste survey in 2024. In a significant policy reversal, the Union Government announced on 30 April 2025, through the Cabinet Committee on Political Affairs, that caste enumeration would be included in the forthcoming national Census, ending the post-1951 exclusion of comprehensive caste data.

A caste census must be distinguished from the reservation system it is invoked to inform: reservation is the constitutional mechanism of affirmative action under Articles 15(4), 16(4), and 46, whereas the census is merely a data-gathering instrument. It must also be distinguished from the SC/ST enumeration, which the regular census already performs, and from the Mandal Commission estimate of 1980, which derived the figure of 52 percent OBC population not from direct enumeration but by extrapolating from the 1931 caste data and other surveys. The absence of a current OBC headcount means reservation policy and welfare targeting rest on decades-old projections, which is the core empirical grievance driving demands for a fresh count.

The controversies are substantial. Proponents argue that without disaggregated data, the proportionality and adequacy of reservations cannot be assessed, that the 50 percent ceiling on reservations affirmed in Indra Sawhney v. Union of India (1992) cannot be rationally revisited, and that policy operates blindly. Critics warn that enumerating caste reifies and entrenches identity, risks inflating community claims, and may harden social cleavages for electoral mobilisation. There are also federalism questions: the Census is a Union List subject (Entry 69), so state surveys occupy contested constitutional space, a point litigated in challenges to the Bihar exercise that the Patna High Court and Supreme Court declined to stay. Data accuracy, self-reporting bias, and the politicisation of sub-caste claims remain unresolved.

For the working practitioner—the policy researcher, the civil services aspirant, or the desk officer tracking Indian social policy—the caste census is a fault line where empirical governance, constitutional affirmative action, and electoral politics converge. Mastery of the topic requires holding three threads together: the legal basis (Census Act 1948, Articles 330/332, the Indra Sawhney ceiling), the data lineage (1931 to SECC 2011 to the 2025 decision), and the live federal contest between Union and state enumerations. The 2025 inclusion of caste in the national Census, if implemented, will produce the first authoritative all-India caste dataset in over nine decades, with direct consequences for reservation quanta, sub-categorisation of OBCs, and welfare delivery.