Accounting for Revealed Comparative Advantage: Economic Complexity Redux

Using the bipartite international trade networks to characterize the economic complexity has generated new insights into the patterns of economic growth across countries. The Economic Complexity Index (ECI) as charted by Hidalgo and Hausmann (2009) has been found particularly successful in explaining the difference in economic growth across countries. As staggering as its explanation power is, little has been known about the underneath factors influencing ECI. In this paper, we develop an analytically tractable framework that marries a full account of supply and demand factors to the global production network with a nested CES preference structure. The model provides formula offering a simple and intuitive decomposition splitting Balassa (1965)'s Revealed Comparative Advantage (RCA) into contributions of a number of different micro mechanisms. Based upon the decomposition, we study the relative importance of each margin embedded in ECI to explain the differences in income and inequality across countries. Our empirical results reveal that the positive correlation between complexity and income growth is mostly driven by the factors on the supply-side. By contrast, the negative association between economic complexity and inequality stems mainly from the demand factors underlying comparative advantage. We then propose a model-based instrumental variable to infer causal associations between economic complexity and per capita GDP growth. Our IV results show that a one standard deviation increase in ECI leads average income to grow faster by 0.65% to 0.72% annually.