Cash and Capabilities

The capabilities approach claims that the freedom to achieve alternative ends is of primary moral importance to policy evaluation. We use a new Bayesian framework for estimating capabilities to conduct a capabilities-based impact evaluation of a prominent poverty-alleviation policy: cash transfers. Specifically, we evaluate a cluster-randomized controlled trial of Kenya's flagship unconditional cash-transfer scheme, Cash Transfer for Orphans and Vulnerable Children (CT-OVC). Our method entails a decomposition of more traditional estimates of impact into a capabilities effect and a choice effect. We find that CT-OVC expanded capabilities across a variety of dimensions (e.g., consumption of micronutrient-rich foods) and that the effects are often larger for women. We also find that choice plays an important contributing role, which we argue has implications for understanding the mechanisms of impact. This paper represents the first capabilities-based impact evaluation, which we claim is a more general approach to evaluation.