Import Variety And Technological Progress: Evidence From China Industry

Over the past decade，research about the new products in the international tradehas attracted much attention，both endogenous growth theory and heterogeneous-firmtrade theory have demonstrated that product variety can promote economic growth，and a large number of scholars has carried on the empirical research to this conclusion.Therefore, at the background of continual trade surplus, studying China’s import fromthe angle of product variety can provide theoretical basis for making a rational importpolicy to solve the problem of trade imbalances, which shows important practicalsignificance.Utilizing disaggregated UNCOMTRADE trade data at the five-digit level（SITCrevision3）for China from2002-2011, this paper respectively uses Feenstra method（1994,2007） and entropy index（2008） to measure the import variety for31industries. With the results from these two methods, both of them show that thecapital-intensive industries’ import variety is the largest, followed by thelabor-intensive industry and the resource-intensive industry. But as for the specificindustry, the ranking of the import variety is different. And the main reason may bethat Feenstra method is from the perspective of the generation of new products anddisappearance of existing products while Entropy index is from the perspective of thespatial distribution of the product. On the basis of the above research, this paperextends the empirical strategy to investigate the effect of the change in import varietyon TFP by using GLS method and system GMM method. By the method of Feenstrawe derived, this paper has demonstrated that import variety has a significant positiveimpact on their industries’ TFP. On the basis of entropy index, this paper has foundthat different related varieties have different effect on TFP. The unrelated importvariety can restrain the increase in TFP, when the related import variety has a positiveimpact on the increase in TFP, and both have passed the significance test. Thefindings vary across industries owing to their factor intensity. Finally, the paperprovides policy recommendations according to the conclusion we derived.