On the Gains from Tradable Benefits-in-kind: Evidence for Workfare in India

Widely-used in-kind benefits are often rationed across potential participants in social programmes. This poses an operational challenge in implementation with limited public information. One possible solution is to randomize assignment given the available information. However, as this paper shows, neither the purposive targeting of in-kind benefits nor randomized assignment is likely to yield a competitive equilibrium given private information on diverse personal gains. The paper characterizes equilibrium assignments and the implications for policy and evaluation. The theory is applied to an antipoverty programme providing rationed jobs on rural public works in a poor state of India. Special-purpose survey data reveal large unexploited gains from allowing tradable assignments. The paper shows that the potential gains exceed those from poverty targeting and stylized cash transfer options. Realizing the scope for poverty reduction by such means will undoubtedly require complementary efforts in reducing the influence of other market and institutional factors creating poverty in the first place, including inequalities in access to credit and knowledge.