Africa has had slow growth and a massive exodus of capital. In many respects it has been the most capital-hostile region. We review and interpret the aggregate-level and microeconomic literatures to identify the key explanations for this performance. There is a reasonable correspondence of the two sets of evidence, pointing to four factors as being important. These are a lack of openness to international trade; a high-risk environment; a low level of social capital; and poor infrastructure. These problems are to a substantial extent attributable to government behavior, and the paper includes a review of the political economy literature addressing that behavior.

We distinguish between policy and "destiny" explanations of Africa's slow growth during the past three decades. Policies were poor: high export taxation and inefficient public service delivery, and "destiny" was adverse: landlocked, tropical locations, and terms of trade deterioration. During the 1990s, Africa's economic policies improved, although with considerable variation both between countries and between policies: trade and exchange rate policies improved much more than service delivery. Thus, the differing explanations of past slow growth imply different predictions for growth in the coming decade. We argue that poor public economic services are likely to be the binding constraint.

We use firm-level panel data for the manufacturing sector in four African countries to investigate whether exporting impacts on efficiency, and whether efficient firms self-select into the export market. Based on simultaneous estimation of a production function and an export regression, our preferred results indicate significant efficiency gains from exporting, which can be interpreted as learning by exporting. We show that modelling unobserved heterogeneity by a flexible approach is important for deriving this conclusion. A policy implication of our results is that Africa would gain from orientating its manufacturing sector towards exporting.