Hanna, Rema, Vivi Alatas, Abhijit Banerjee, Benjamin Olken, Ririn Purnamasari, and Matthew Wai-Poi. 2016. “Self-Targeting: Evidence From A Field Experiment In Indonesia”. The Journal Of Political Economy 124 (2).
In this paper, we show that adding a small application cost to a social assistance program can substantially improve targeting because of the self-selection it induces. We conduct a randomized experiment within Indonesia’s Conditional Cash Transfer program that compares two of the most common methods of targeting welfare programs in the developing world: in one, beneficiaries first need to apply for the program, and then an enumerator visits them at home and determines their eligibility based on a proxy-means asset test; in the other, they are visited directly by the enumerator and automatically enrolled if they qualify based on the same proxy-means test. When applications were required, we find that the poor are more likely to apply than the rich, even conditional on whether they would pass the asset test. On net, the villages where applications were required have a much poorer group of beneficiaries than automatic enrollment villages. However, marginally increasing the cost of applying does not necessarily improve targeting: while experimentally increasing the distance to the application site reduces the number of applicants, it screens out both rich and poor in roughly equal proportions. Estimating our model of the enrollment choice suggests that our results are largely driven by the rich forecasting that they have a very small likelihood of passing the asset test, and so not bothering to apply, which in aggregate substantially improves targeting efficiency. The results suggest that the combination of the small cost and the final screening gives this class of mechanisms the ability to achieve many of the benefits of self-selection without imposing onerous ordeals on program beneficiaries.