The Union government gives special grants only for those states or regions that are deemed ‘backward’ for their development to bring them on par with others. But what determines if a state is backward?
Economic metrics such as gross domestic product and per-capita income are used to measure the relative backwardness of states. Sounds logical, right? Similarly, shouldn’t social justice schemes also focus more on castes that are more backward than others? If yes, how does one measure the relative backwardness of a caste?
For the first time in India, a composite backwardness index (CBI) has been constructed in Telangana to objectively measure the relative backwardness of a caste. The analytical framework is methodologically modest in description and conceptually radical in implication: it produces a single, comparable number for the relative backwardness of each of 242 caste groups, derived from the responses of 3.55 crore people across 75 fields of information.
The CBI is built on 42 equally-weighted parameters spanning education, occupation, income, land and assets, living conditions, access to finance, gender and social discrimination. Each caste is scored on a scale of 0 to 126, where a higher score denotes greater backwardness. The least backward caste in Telangana, the Kapu community, scores 12. The most backward, the scheduled caste Dakkal, scores 116. The state average is 81. Above that line lie 135 of the 242 castes, accounting for two-thirds of the population. Below it lie 107, including all 18 general castes. The numbers, for the first time, settle questions that have so far been the province of assertion.
The Mandal Commission, whose 1980 report shaped four decades of reservation policy, used 11 parameters and relied largely on field visits and judgement to classify social and educational backwardness. It was a serious effort with the tools then available. The CBI uses 42 parameters and 3.55 crore observations, distinguishes rural and urban living standards within each parameter, and produces a continuous score rather than a binary classification.
The CBI provides empirical evidence of the established ugly truth of Indian society—people belonging to SC and ST communities are three times and the backward classes 2.7 times more backward than the general castes. A whopping two-thirds of Telangana’s population belong to really oppressed castes that badly need help.
The index also changes the paradigm of welfare design, which has so far been built around broad clusters such as BC, SC and ST. The data shows that even within these groups, inequality is sharp. The SC Madiga community is more than double the population of SC Malas and 15 percent more backward. Shaik Muslims are five times more numerous than general-caste Muslims and 40 percent more backward. People from the Padmasali BC are actually less backward than the state average. Aggregating these communities into a single category obscures more than it reveals.
Nearly a third of the beneficiaries of Telangana’s flagship welfare schemes belong to castes less backward than the state average. The CBI reframes the welfare question itself: instead of filling a common well from which marginalised groups draw in proportion to their population, the state can now deliver resources in proportion to each caste’s backwardness.
The CBI also reveals the specific drivers of backwardness. Contrary to popular belief, it is not land ownership but access to English-medium education that is the strongest predictor of where a caste sits on the scale. People from the Lambadi tribe and Mudiraj backward group rank well on irrigated land ownership, but remain among the more backward groups. Meanwhile, the Gouds, goldsmiths and BC Christians score better than the average because of school access.
The CBI also shatters the myth that caste ceases to matter among the very poor. Among households earning under `1 lakh a year, the gap in scores across general castes, BCs, SCs and STs mirrors the gap among the rich. Even at this income level, 34 percent of children from general castes attend private school, against just 5 percent of SC and ST children. The general castes carry advantages—networks, language, expectations—that money alone does not buy. A third of all private-sector professionals in Telangana are from the general castes; only 4 percent are STs, despite the two groups being similar in population.
Taken together, the findings argue for a recalibration rather than a rejection of the existing social justice framework. India’s policy architecture was built around the village economy, where land, water and access to common spaces were binding constraints. Those have not disappeared, but they have been joined or overtaken by access to the institutions that determine modern economic outcomes—English-medium schools, private healthcare and formal-sector employment.
Three caveats. First, the CBI is a relative measure within Telangana; it ranks castes against each other, not against the rest of the country. Second, like all self-reported survey data, it carries biases on land, income and occupation that smooth-out at the aggregate level but should not be over-read at the margin. Third, the choice of 42 parameters and equal weights is a defensible starting point, not a final answer; future iterations will refine both. None of these undermines the analytical framework itself.
The Telangana ‘CBI’ has finished the investigation. The questions now are if the polity has the intent and expertise to render justice based on the verdict, and will the rest of the nation follow Telangana’s lead.
Praveen Chakravarty | Convenor of the Expert Group on Telangana Caste Census & co-developer of the CBI framework
(Views are personal)
