Estimating the Productivity Selection and Technology Spillover Effects of Imports

In the wake of falling trade costs, two central consequences in the importing economy are, first, that stronger competition through increased imports can lead to market share reallocations among domestic firms with different productivity levels (selection). Second, the increase in imports might improve domestic technologies through learning externalities (spillovers). Each of these channels may have a major impact on aggregate productivity. This paper presents comparative evidence from a sample of OECD countries. We find that the average long run effect of an increase in imports on domestic productivity is close to zero. If the scope for technological learning is limited, the selection effect dominates and imports lead to lower productivity. If, however, imports are relatively technology-intensive, imports also generate learning that can on net raise domestic productivity. Moreover, there is somewhat less selection when the typical domestic firm is large. The results support models in which trade triggers both substantial selection and technological learning.