There is an old joke about a drunk man searching under a lamppost for his keys. Did you drop them here? No,but this is the only place where theres enough light to look. Thats what the recent plan to identify the BPL population and include caste and religion information sounds like. Obviously,there are large overlaps in these categories but the point is not to create a composite picture of deprivation but to target poverty as efficiently as possible. Just as we need accurate data on numbers of the poor for more useful welfare schemes,we should also avoid imposing a fixity on their caste identity by putting this information in the official census.
Though colonial administrators attempted to map caste,the post-Independence establishment recoiled at a record of caste identity in the national census. However,after the Mandal Commission inaugurated a phase of competitive politics around caste,for electoral mobilisation and policy benefits,there seemed to be a new rationale in knowing the real numbers. Given how much rides on caste classification,in terms of education,employment and welfare,even political office,some argue that there is a need to assess the dimensions of each claim. However,this argument fails to recognise that officially assigning caste categories can harden these identities. First,it is a more mutable category than most altered by marriage,migration and mobility. It shifts by geographical context,and people respond differently depending on the framework whether they are being socially slotted or making a bid for affirmative action. Unlike religion or ethnicity,caste is a trickier kind of group identity.