A Structural Estimation of the Return to Infrastructure Investment in China

This paper develops a model of endogenous productivity to examine the impact of public infrastructure investment on output. It matches firm-level production data with province-level infrastructure data to address a set of wellknown identification issues in the literature, and employs a structural estimation to distinguish the long-run productivity effect from the short-run Keynesian demand effect of public infrastructure investment. The estimated rates of return of infrastructure are 9.2% and 2.5%, respectively, for revenue-based and quantity based total factor productivity. The returns almost triple once maximal spillover effects are taken into account. Firm-level evidences are consistent with a mechanism in which public infrastructure investment functions as a catalyst in facilitating resource reallocation from less productive firms to more productive firms.