Samaritan agents? : on the delegation of aid policy

Abstract

Should an aid donor delegate the responsibility for allocating its budget to an agent less averse to inequality than itself in order to alleviate the Samaritan's Dilemma it is facing? Despite the intuitive appeal of this proposition, I show that the optimal type of agent depends on whether or not committing to a greater share for recipients where the productivity of aid is low is efficiency-enhancing. This is the case for donors not too concerned with redistribution. They would therefore benefit from delegating the determination of the discretionary allocation rule to agents more sensitive to distributional issues than themselves.