Export Performance in Africa and Asia's Manufacturing: Evidence from Firm-level Data

Abstract

Asia's share of manufacturing export to gross domestic product in the 1990s was more than five times that of Sub‐Saharan Africa. While explanations abound in the literature as to why Africa has failed, recent empirical work suggests that the reason for Africa's dismal export performance lies in a low skill‐to‐land ratio, which causes its comparative advantage to lie in primary exports. This result is derived from a basic 2 × 2 Hecksher–Ohlin model. However, the aggregation inherent in the industry or national level data sets that are used for testing this theory hides substantial heterogeneity. Input intensity heterogeneity violates the assumptions of the factor proportions framework and casts doubts on the reliability of estimation results. Moreover, measurement error on the skill variable at the aggregate level causes biases in Ordinary Least Squares estimation. This paper uses a combined micro‐ and macro‐level data set for investigating the determinants of export performance in Asia and Africa's manufacturing. Tobit estimations find no robust association between skill intensity and export performance in the textile and garment industries in Ghana, Kenya and India. Africa and Asia's significant performance gap might be better explained by there being poorer institutions in Africa as measured by the Knack and Keefer (1995) index and Kaufman et al.'s (1999) ‘rule of law’ composite index.