Aid, institutions and economic growth: Heterogeneous parameters and heterogeneous donors

Hassen Wako


Aid effectiveness has been a subject of long-sustained debate. This study contributes to this debate using panel data from 43 Sub-Saharan African countries. Its novelty lies in assessing the intermediary role of institutional quality between aid and growth, and in taking a disaggregated view of aid (at the level of a donor). Using estimation techniques which allow for recipient-specific (slope) parameters and suit the context of non-stationary and cross-sectionally dependent panels, the study finds that the relationship between aid and growth is characterised by heterogeneous (or recipient-specific) short-run parameters but a shared long-run coefficients. In the long-run, the direct growth effect of (aggregate) aid from 'traditional' donors is robustly non-positive, and the indirect effect is negative and robust to different specifications. Disaggregation reveals that there is heterogeneity in aid-effectiveness from the donor side as well: there are cases of 'good' aid (four donors), 'bad' aid (ten donors), 'neutral' aid (three donors) as well as cases where the total effect of aid is 'indeterminate' (four donors). With a lesser confidence, attributed to smaller sample size and less reliable quality of data, Chinese aid to Sub-Saharan Africa has a positive direct growth effect, a negative institutional effect, and thus an indeterminate total effect. The short-run relationships are generally not robust to alternative specifications. Comparison of the behaviour of donors with differing degrees of aid-effectiveness suggests that the future of aid would benefit more from focusing on its quality than quantity. In particular, two quality aspects - reduced fragmentation (or better specialisation) and better donor alignment (with recipient country's policy and system) - deserve much more attention.

JEL Classification: F35; O43; F63; F43

Keywords: aid; economic growth; institutions; donor/recipient heterogeneity

Download the working paper