The Micro-Macro Paradox of Aid Revisited

Abstract The goal of development aid used to be development as proxied by economic growth. This goal was operationalized in cost benefit analysis. However, aid effectiveness has always been in doubt. The doubts have had two consequences: Aid is falling, and its goal has been widened to make it non-operational. This paper looks at the reasons for the doubt, the micro-macro paradox of aid: (i) Micro project evaluations find a fair efficiency. The cost-benefit tools used to evaluate aid projects should aggregate to the macro. (ii) Univariate macro estimates find a zero-correlation result, and that lags solve the causality problem. (iii) Multivariate macro estimates find highly variable results with a small meta-average. A list of possible explanations that may reduce the paradox is provided, but the effect-sizes of these possibilities are hard to assess. It is argued that as the operational goal of aid has vanished, so has the support for aid.