This paper examines the learning-by-exporting effect in Chinese manufacturing firms from 2005 to 2007. The traditional view is that exporting can lead to increased productivity by facilitating access to the global market and, thus, information and cutting-edge technologies. This process has been explained by the learning-by-exporting theory, which is supported by mixed empirical evidence. A semiparametric estimation method was used to measure firm-level productivity and examine the contingent impact of exports on productivity. On the one hand, the exporting firms exhibited significantly higher productivity and faster growth than the non-exporting firms. On the other hand, the effect of exporting on productivity was dependent upon firms’ innovation behavior. Therefore, the learning-by-exporting effect is contingent on a firm’s innovation capabilities. Only sufficiently innovative firms could actually experience faster growth through exports and innovation. For non-innovative firms, exporting could even result in decreased productivity. This paper successfully reconciles the mixed findings from the existing literature and explains why both positive and negative evidence can simultaneously and reasonably exist under learning-by-exporting theory.
