This paper posits a political-economy model a la Lindbeck and Weibull (1987) to study how mandated political representation of a disadvantaged minority group affects targeting within the group — particularly, the effect on the poor in that group. The model demonstrates how such af?rmative action can result in reduced transfers to the poor when voters are biased towards candidates from their own caste. Since the 1950s, India has implemented af?rmative action programs for the upliftment of Scheduled Castes (SCs) in different spheres; a prominent form of which is mandated political representation in the legislature. Previous ?ndings in the literature have been that such mandated political representation biases policies in favor of the minority group as a whole. However, the question of who gains and who loses within the minority group has not been studied before. Data on household level consumer expenditure collected by the National Sample Survey Organization (NSSO) combined with data on mandated representation in the Indian legislature are used to verify the predictions of the model. There is evidence of lesser implementation of anti-poverty programmes in areas reserved for SC legislators, in accordance with the theory. There is also evidence suggesting that political reservation has had a positive impact on Scheduled Caste inequality. These ?ndings suggest that mandated political representation has been detrimental to the poor SCs and could be aggravating inequality within the SCs.