The Irrelevance of Caste Census

Sonalde Desai in “Caste and Census: A Forward Looking Strategy” (EPW, 17 July) suggests that we need a caste census because there is adequate evidence of “caste based inequalities refl ecting social processes that deserve policy attention” and the bulk of her article suggests a way of dealing with the problems of counting and categorisation. I would contest the assumption that counting castes will improve policy outcomes and argue that what we should be debating is not the exact size of any particular caste group, but rather why policy processes are unable to cover the last mile.

What we are getting wrong may be something far more fundamental than not having the right data. Just as an illustration of my point, data on education has, for many years, revealed the overlapping vulnerabilities of location, caste and gender– shown in the NCAER Human Development Indicators 1994, for example. The policy decisions, such as special focus on girls in educationally backward blocks since 2003, are a product of data analysis. In 2010, we are still noting the same issues and still trying to work out strategies for reaching girl children from some social groups and communities and in some parts of the country. If what we need are multiple strategies which are decentralised, innovative and locale specific (which I would argue in favour of), we need numbers and analysis at very local levels. We do not need a national caste census (which raises other troubling questions around identity).