Abstract

Balassa’s index of revealed comparative advantage is not a sufficient statistic for identifying the existence, magnitude, and source of Ricardian comparative advantage. We overcome this issue by developing an alternative structural approach to the measurement of Ricardian comparative advantage based on a quantitative trade model with firm and product selection, where both relative unit input requirements and relative factor endowments play a role. Through the lens of the model, we identify sets of sufficient statistics for the endogenous and exogenous components of a country’s productivity relative to the rest of the world. Among the exogenous components, we target the country’s relative state of technology as the exogenous determinant of its Ricardian comparative advantage.
Applying our approach to Chinese micro and macro data, we find that, while the state of technology has important implications for the micro behavior of individual firms, firms’ responses have, in turn, far-reaching macro implications for aggregate productivity and revealed comparative advantage.