Targeting social assistance in fragile settings: An experiment in community-based targeting

Targeting is an important but challenging process in the design and delivery of social and humanitarian assistance programs. Community-based targeting (CBT) that uses community leaders as targeting agents is often preferred for their local information advantages, especially when data-driven methods are not feasible. However, it remains unclear how different CBT approaches fare under different constraints and contexts. Specifically, how do community leaders target and allocate resources under different resource constraints and discretion? We implemented a cluster-randomized controlled trial among community leaders in 180 villages in Ethiopia to evaluate how community leaders target and allocate resources under different budget constraints and discretion. We find that, when resources are limited, community leaders prefer to increase the number of beneficiaries by reducing the average transfer amounts. Community leaders appear to minimize exclusion errors even at the expense of increased inclusion errors. Consistent with this, we find that granting community leaders some level of discretion allows them to reduce exclusion errors and include households that are materially deprived or living in conflict-affected settings. Finally, community leaders are more vulnerable to favoritism when the budget is larger and they lack discretion. These findings offer important lessons that can inform targeting in the context of increasing budget pressures and cuts.