Development thresholds of foreign aid effectiveness in Africa

The purpose of this paper is to examine whether initial levels in GDP growth, GDP per capita growth and inequality adjusted human development matter in the impact of aid on development. In substance its object is to assess if threshold development conditions are necessary for the effectiveness of foreign aid in Africa.

The panel quantile regression technique enables us to investigate if the relationship between development dynamics and development assistance differs throughout the distributions of development dynamics.

Three main findings are established. First, with slight exceptions, the effectiveness of aid in economic prosperity (at the macro level) increases in positive magnitude across the distribution. This implies high-growth countries are more likely to benefit from development assistance (in terms of general economic growth) than their low-growth counterparts. Second, the positive nexus between aid and per capita economic growth displays nonlinear patterns across distributions and specifications, with the correlations broadly higher in top quantiles than in bottom quantiles after controlling for the unobserved heterogeneity. Third, the aid-human development nexus is negative and almost similar in magnitude across distributions and specifications.

As a policy implication, there is need to improve management of aid funds destined for health and education projects in the sampled countries. Moreover, given the magnitude of the nexuses, while blanket aid initiatives could be applied for policies targeting the human development index (due to the absence of significant differences in the magnitude of estimated coefficients), such are unlikely to succeed for aid targeting economic prosperity at macro and micro levels. From the weight of the findings, given a policy of balancing the impact of aid, it could be inferred that low-growth countries would need more aid than their high-growth counterparts because of the less positive effects in the former countries.

This paper contributes to existing literature on the effectiveness of foreign aid by focussing on the distribution of the dependent variables (development dynamics). It is likely that high- and low-growth countries respond differently to development assistance.