Abstract

The decision to carry out a caste-based census in India in order to, among other things, get a more scientific basis for reservations has generated considerable debate. Much of this debate has focused on data quality, without much attention paid to the question of what should be the scientific basis for determining whether a caste is backward. This article explores the concepts used, implicitly or explicitly, in the literatures on caste and reservations to come up with a framework to determine dominance-based caste hierarchies in a milieu of change. We test this framework in an empirical reality drawn from the south Indian state of Karnataka to argue for a more disaggregated and inclusive approach to reservations.