Conditional cash transfers to mothers, intrahousehold allocations: the role of unobservability

Over the past few decades, conditional cash transfers (CCT) to households have become an increasingly important tool to generate socially desirable outcomes in human capital development programs. Subsequently, there has been a spur of literature evaluating the impacts of CCTs on poor households, mostly showing a positive impact. Yet, we argue that these impacts are not independent of intrahousehold processes of resource allocation and without a proper understanding of this process, the policy objectives might fail. We particularly look at the case of CCT to mothers and its impact on child human capital accumulation when there is intrahousehold uncertainty. Using an experimental setup in Karnataka, South India, we find that, given an endowment, when mothers cannot observe spousal investment preferences in children, they respond by lowering their willingness to invest in children and increasing precautionary savings. We argue that in a noncooperative environment, without information sharing and unobservability, uncertainties within marriage might result in inefficient allocation of resources. In other words, with the unobservability of spousal preference and resulting uncertainties, resources tend to be under-allocated in household public goods such as children failing the policy objectives of CCT.