Aid effectiveness in sustainable development: A multidimensional approach

With the knowledge gathered from our previous work in policy prioritisation, we were equipped to tackle one of the most important issues in development economics: the impact of international aid. For more than 60 years, this was a controversial topic because evidence on the effectiveness of aid at the macro level had been mixed. Hence, this paper conciliates many discussions around this topic and provides the first worldwide estimates on aid impact that are specific to each country and indicator.

What is the impact of international aid? We answer this question by linking disaggregated aid-flows data to a large set of indicators classified into the Sustainable Development Goals (SDGs). Since such linkage is not perfect (due to the nature of the data), we deploy an artificial intelligence model of the causal process through which changes in aid flows contribute to the dynamics of individual indicators. The model accounts for salient features of real-world development such as multidimensionality, complex interconnections between indicators, heterogeneous aid-to-expenditure ratios, rationally-bounded bureaucracies, fungibility, and the temporal structure of contemporary aid flows across development dimensions. The model does not require cross-country pooled data, so we calibrate its parameters for each of the 146 aid-recipient countries in our sample, preserving important contextual information of each nation. By producing counterfactual simulations where aid is removed, we obtain nuanced estimates of the impact of international assistance during the first decade of the 21st century, at the level of each country, SDG, and indicator. We validate our results using a sector-specific study with similar–but more aggregate–findings. Such a large and detailed picture of the multidimensional impact of aid has not been documented before.