Aid Allocation and Poverty Reduction

Abstract

In the efficient allocation of aid, aid is targeted disproportionately to countries with severe poverty and adequate policies. For a given level of poverty, aid tapers in with policy reform. In the actual allocation of aid, aid tapers out with reform. Aid now lifts about 30 million people a year out of absolute poverty. With a poverty-efficient allocation, the same amount of aid would lift about 80 million people out of poverty. Collier and Dollar derive a poverty-efficient allocation of aid and compare it with actual aid allocations. They build the poverty-efficient allocation in two stages. First they use new World Bank ratings of 20 different aspects of national policy to establish the current relationship between aid, policies, and growth. Onto that, they add a mapping from growth to poverty reduction, which reflects the level and distribution of income. They compare the effects of using headcount and poverty-gap measures of poverty.

They find the actual allocation of aid to be radically different from the poverty-efficient allocation. In the efficient allocation, for a given level of poverty, aid tapers in with policy reform. In the actual allocation, aid tapers out with reform. In the efficient allocation, aid is targeted disproportionately to countries with severe poverty and adequate policies - the type of country where 74 percent of the world's poor live. In the actual allocation, such countries receive a much smaller share of aid (56 percent) than their share of the world's poor. With the present allocation, aid is effective in sustainably lifting about 30 million people a year out of absolute poverty. With a poverty-efficient allocation, this would increase to about 80 million people. Even with political constraints introduced to keep allocations for India and China constant, poverty reduction would increase to about 60 million. Reallocating aid is politically difficult, but it may be considerably less difficult than quadrupling aid budgets, which is what the authors estimate would be necessary to achieve the same impact on poverty reduction with existing aid allocations.

This paper - a joint product of the Office of the Director, and Macroeconomics and Growth, Development Research Group - is part of a larger effort in the group to examine aid effectiveness.