New Revealed Comparative Advantage Index: Dataset and Empirical Distribution

Balassa Index (Balassa 1965) is widely used in the literature to measure country-sector Revealed Comparative Advantage (RCA). However, being computed on observed trade flows, it mixes up all the factors influencing trade flows. In particular, Balassa Index cannot isolate exporter-sector (ex ante) specific factors which are the source of comparative advantage in the spirit of the traditional trade model. Furthermore, Balassa Index suffers some empirical distribution weaknesses, mainly time instability and poor ordinal ranking property (Yeats 1985; Hinloopen and Van Marrewijk 2001). A recent paper by Costinot et al. (2012) provides a micro-founded version of the Ricardian model and suggests a new measure for comparative advantage. We build up on this paper, and present a dataset providing a new econometric based measure for Ricardian RCA.