Is learning by exporting technology specific? Evidence from Chinese firms

ABSTRACT

Integrating the theory of heterogeneous firm and trade with the neo-Schumpeterian view, this paper examines whether learning-by-exporting processes in Chinese firms are technology specific. Using a combination of propensity score matching and difference-in-differences estimation, we find weak evidence that exports generate higher productivity and growth for Chinese firms. This learning effect is subject to the nature of technology across industries: learning-by-exporting processes favor sectors characterized by high levels of appropriability and technological opportunity, while they are hindered in sectors featuring a wider knowledge base and higher cumulativeness. This technology-specific nature of learning effects leads to discrepant post-export gains in productivity across different sectors as well as economies.

KEYWORDS:

• China
• heterogeneity
• learning by exporting
• technological regime

JEL CLASSIFICATION:

• F10
• O33
• O19

Acknowledgments

The authors are grateful to Stefan Bauernschuster, Uwe Cantner, Caterina Giannetti, Sarah Köster, Arianna Martinelli, Ljubica Nedelkoska, the anonymous reviewers, and the editor for their helpful comments and discussions. The authors also thank Djamila Koberstein and Jonathán Leupert for their research assistance. This work was supported by the German Research Foundation (grant number GK 1411/1), the National Natural Science Foundation of China (grant number 71503247), and the Institutes of Science and Development, Chinese Academy of Sciences (grant number Y9X3581H).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The dataset was updated to 2006, and the explanation of the dataset refers to the document generated by Hall, Jaffe, and Trajtenberg (2001)

2 We split the matched sample into two subsamples according to the indicators of technological regimes. Based on the notion proposed by Malerba and Orsenigo (1997) and Breschi, Malerba, and Orsenigo (2000), the subsample with the levels of appropriability, cumulativeness, knowledge base greater than their medians, and the level of technological opportunities less than the median is classified as Schumpeter Mark II, and the other subsample as Schumpeter Mark I. We conduct PSM-DiD estimations on productivity growth for two subsamples separately and the estimation results are shown in Table A3.

Additional information

Funding

This work was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft) [grant number GK 1411/1], the National Natural Science Foundation of China [grant number 71503247], and the Institutes of Science and Development, Chinese Academy of Sciences [grant number Y9X3581H].