Why India's Caste Census Debate Should Ask the Right Questions |
Listen to this article:
When the Telangana government released the findings of its Socio-Economic, Educational, Employment, Political and Caste (SEEEPC) Survey in early 2026, it did something India’s policy establishment has long resisted, to put numbers to the structure of inequality that millions navigate every day. Covering more than 35.4 million individuals across 242 castes, roughly 97% of the state’s population, it is the most granular caste deprivation dataset independent India has ever produced.
The survey’s Composite Backwardness Index (CBI), built on 42 multidimensional parameters spanning education, land ownership, housing, sanitation, occupation, debt, and social integration, shows that Scheduled Castes (SC) score 96 and Scheduled Tribes 95 on a scale where higher means more deprived, while upper-caste Other Castes (OCs) score just 31. The average SC or Scheduled Tribe (ST) household is, by the state’s own multidimensional measure, approximately three times as backward as the average upper-caste household.
This data triggered the need for national conversation about caste census and reservations. Which communities will demand larger quotas? What happens to the Other Backward Classes (OBC) share? While, these are not unimportant questions, they are the wrong ones to begin with.
A census that proves the problem is much bigger
The single most important finding in the SEEEPC Survey is not about quotas. It is about the nature of deprivation itself. The CBI was constructed precisely because single indicators like especially income miss how disadvantage accumulates and persists across domains and generations. What the index reveals, across all 242 castes, is a multidimensional hierarchy that operates in housing, sanitation, schooling quality, land access, labour market position, debt burden, and spatial location simultaneously. Poverty, for the most backward castes, is not just one which no single welfare instrument can address.
Consider what the data shows about domains that reservations do not touch. Only 8% of SC households own agricultural land, compared to 31% of OC households. Among daily wage workers, 46% of SCs and 41% of STs survive in the precarious informal labour market, against just 11% of OCs. Private school enrolment stands at 10% for SCs and 8% for STs, compared to 30% for OCs. Social closure through endogamy, which reproduces both advantage and disadvantage across generations, remains near-universal, with inter-caste marriages at roughly 5 to 6%. None of these gaps, in land, in informal labour, in schooling quality, in spatial concentration, can solely be resolved by adjusting the percentage of seats reserved in a central university.
The reservation system is not being indicted here. It has been effective, within its scope, at increasing representation in formal public institutions. The problem is that the public debate has allowed reservation to stand in for the entire architecture of caste redressal, only relying as a seat in a government college is an answer to a family living on daily agricultural wages in a rain-dependent district with no pucca home and no toilet. The Telangana survey questions and demands that we expand the frame and rethink the dimensions.
The SEEEPC Survey also settles, at the level of state-wide empirical evidence, a question that the Supreme Court left open in its 2022 judgment upholding the 103rd Constitutional Amendment, whether the Economically Weaker Section (EWS) reservation has an empirical basis.
It does not. The EWS quota was introduced in 2019 without any........