Detecting learning from importing: The effect of importing behaviour on R&D expenditures of Chinese firms

We argue that importing stimulates firms' learning-based R&D activities in the present and future periods. Our theory predicts that the strength of the learning effect depends on the type of imported materials, the industries, and the import destination. Using a dataset on Chinese manufacturing firms covering the episode of China's WTO accession, we employ the multiple-treatment propensity matching difference-in-differences method and find evidence supporting our hypotheses: (1) importing capital goods stimulates more R&D expenditures than importing intermediates; (2) R&D intensity increases more for capital goods importers in capital-intensive industries than those in labor-intensive industries; and (3) importers increase R&D intensity more by importing from high-income countries than low-income countries.