Exploring Recipient Preferences and Allocation Mechanisms in the Distribution of Development Aid
通过激励相容实验获取潜在受援者对14种减贫措施的估值，发现估值高度异质且难以根据受援者特征预测，进而模拟比较现金转移与投票两种分配机制在产生受援者剩余方面的表现。
Abstract This study uses incentive-compatible techniques to obtain valuations of 14 common poverty reduction interventions from (probable) aid recipients. Recipients’ valuations for these interventions are highly heterogeneous both across interventions and across recipients of the same intervention. Valuation for interventions does not correlate with overall poverty or with perceived need for specific interventions, suggesting that targeting individuals with high valuations based on recipient characteristics is difficult. Through simulations, this study assesses how various allocation mechanisms—cash transfers and voting—compare in generating recipient surplus in the allocation of aid. When markets function and constraints on joint private contributions to public goods do not bind, cash transfers generate considerably more recipient surplus than voting. Even when cash transfers cannot enable public goods and some services, they may still outperform voting at very low resource levels. However, as resource levels increase, voting dominates cash transfers from a surplus-maximization perspective.