What a Caste Census Can and Cannot Settle

When the Bihar government released the results of its caste-based survey in October 2023, the headline number, that Other Backward Classes and Extremely Backward Classes together made up roughly sixty-three percent of the state's population, was treated by supporters as vindication of a long-standing demand and by critics as a political stunt timed for electoral advantage. Both reactions missed a more interesting question. India last counted caste comprehensively in 1931, under colonial administration, and has relied since on extrapolations, sample surveys and the Mandal Commission's contested 1980 estimate of fifty-two percent OBC population ever since. Bihar's survey was not really a technical exercise in headcounting. It was a demand for the state to see its own society clearly, made by parties who suspected that decades of policy had been built on guesswork that happened to suit whoever was in power.

Why the demand has grown louder

The push for a national caste census, led most vocally by Bihar's political class across the RJD-JD(U)-Congress spectrum and echoed by opposition parties nationally, rests on a reasonable grievance. The Socio Economic and Caste Census of 2011 collected caste data alongside economic indicators but the caste component was never officially released, reportedly because of inconsistencies and duplication in the raw data, leaving India's largest attempt at caste enumeration in independent history buried in unpublished files. Meanwhile every major welfare and reservation decision, from OBC quotas in central government jobs and educational institutions to state-level sub-quotas, has proceeded on the Mandal Commission's forty-year-old estimate, updated only through unofficial projections. Advocates argue this is indefensible: a state that measures literacy, employment and consumption expenditure with statistical rigour every few years has no comparable rigour on caste, the single variable that determines eligibility for one of its largest redistributive programmes.

What better numbers would actually change

A credible national count would do at least three useful things. It would allow policymakers to check whether reservation quotas, currently capped in aggregate for OBCs at twenty-seven percent nationally, correspond even roughly to population share, since if OBC groups are undercounted relative to their real numbers, the case for revisiting that ceiling strengthens considerably. It would also expose the internal inequality within the OBC category itself, where a small number of politically dominant castes are widely believed to capture a disproportionate share of quota benefits while hundreds of smaller, more backward communities see little. This is precisely the finding that commissions such as the Rohini Commission, constituted in 2017 to examine OBC sub-categorisation and finally submitting its report in 2023, were set up to investigate, and caste-disaggregated population data would make their recommendations far more defensible than the patchwork administrative data they were forced to rely on. Third, granular data would sharpen the targeting of welfare schemes that already use caste as an eligibility marker, from scholarships to housing subsidies, reducing both under-inclusion of genuinely disadvantaged groups and leakage to better-off ones within nominally backward categories.

What the count cannot resolve

The strongest argument against treating a caste census as a panacea is that India's reservation disputes are not, at bottom, disputes about missing data. The Supreme Court's 1992 Indra Sawhney judgment capped total reservation at fifty percent except in extraordinary circumstances, a limit that has held despite state after state, from Tamil Nadu's decades-old sixty-nine percent quota to Maharashtra's Maratha reservation attempts, seeking to breach it. No caste census, however accurate, mechanically overturns a constitutional ceiling grounded in concerns about administrative efficiency and non-quota candidates' rights; that requires either a constitutional amendment of the kind used for the ten percent Economically Weaker Sections quota upheld in the 2022 Janhit Abhiyan judgment, or a fresh constitutional bench willing to revisit Indra Sawhney itself. Nor does a census settle who should count as backward in the first place, a question that has produced decades of litigation over the inclusion of politically assertive but economically diverse communities such as Jats, Marathas and Patels, none of which is resolved by knowing their exact population share. A community's numerical strength does not automatically establish the social and educational backwardness that the Constitution requires as the basis for reservation, and courts have repeatedly said so.

The political economy problem

There is also a harder, less discussed risk. Once caste numbers are published, they become permanent ammunition in electoral mobilisation, hardening identities that many social reformers had hoped would gradually blur through urbanisation, inter-caste marriage and economic mobility. Bihar's own survey was met with immediate demands from various caste associations for either increased quotas or reclassification, suggesting that the data will be used less to guide considered policy revision than to arm the next round of quota politics with more precise numbers to fight over. This is not an argument against transparency, since obscuring inconvenient social facts has never served justice, but it is a reason to be sober about what enumeration achieves. Data can inform a debate; it cannot substitute for the political consensus, or the judicial reinterpretation, required to actually change the rules.

A modest, necessary reform

The sensible position lies between uncritical enthusiasm and dismissal. India should conduct a rigorous, methodologically sound national caste count, ideally folded into the long-delayed decennial Census whose 2021 round has still not been conducted, with the same statistical care applied to consumption surveys. The results should feed directly into sub-categorisation exercises like the Rohini Commission's, into periodic review of the creamy layer income ceiling that determines OBC eligibility, and into an honest, evidence-based conversation about whether the fifty percent ceiling still serves its original purpose seventy years after independence. What the count should not be sold as is a shortcut past the genuinely difficult constitutional and political choices that reservation policy requires. Bihar's survey has done the country a service by proving the exercise is administratively feasible and politically survivable. Whether the rest of India uses the resulting clarity to build fairer, more targeted policy, or simply to intensify the scramble for a larger slice of a capped quota, will depend on choices no dataset can make for it.