AbstractExport diversification can be a driver of economic growth. Policies to diversify exports attempt to defy an existing comparative advantage in low-value-added goods to promote transformation. Many structural, general policy, and trade-related variables have been identified as important drivers of diversification. This article reviews the literature on the determinants of export diversification and applies Bayesian Model Averaging (BMA) to empirically tackle model uncertainty stemming from many possible determinants. It assesses the relevance and impact of up to 46 drivers of export diversification for up to 47 African countries and 123 trading partners from 1995 to 2018. It finds that African countries’ structural features are especially influential in determining export diversification, a potential explanation for Africa’s more concentrated export baskets compared to other world regions. Furthermore, trade policies, like tariffs and regional integration, are critical in promoting export diversification. These findings underscore the AfCFTA’s potential to foster export diversification across African economies. A diverse set of non-trade-related policy variables also significantly affect export diversification. Factors such as institutions, education, service sector, resource rents, financial development, FDI and exchange rate stability are relevant determinants in different settings. Moreover, the analysis shows that whom you trade with is also important for diversification, and that context-specific differences, such as specific regional integration contexts or type of commodity-dependence, matter. The results suggest that the goals to diversify exports at the extensive margin for structural change and the intensive margin to tackle volatility, seem often not achievable via the same policies.
Read full abstract