Development Effectiveness: What have We Learnt?
指出“贫困效率”援助分配只是一个基准指南，当捐助者缺乏国家信息或无法改变政府偏好时，实际分配应偏离基准，因为政府偏好和资金可替代性限制了援助效果。
We suggest that the "poverty-efficiency" aid allocation is merely a benchmark guide if a donor lacks other information about the country and also the power to change or prevail over government preferences. We argue that in most circumstances donors have only limited scope for the latter and that, while high aid dependence may reduce the fungibility problem and the use of NGOs can by-pass it altogether, such circumstances are not very common. Hence, the main reasons for departing from the benchmark are when the donor has additional information about likely poverty impact, or if poverty reduction is not the objective.