Title:Sino-US S and T Frictions and Transnational Knowledge Flows: Evidence from machine learning and cross-national patent data

Abstract:This paper identifies the impact of China-U.S. science and technology (S&T) friction on knowledge flows in different fields, using data on invention patent applications from China, the U.S., Europe, and the World Patent Office (WPO) along with machine-learning-based econometric methods. The empirical results find that the negative impacts of China-U.S. S&T frictions on cross-border knowledge flows are confined to a limited number of technology areas during the period of our observation. This paper further explores the characteristics of the negatively impacted technology areas, and the empirical results show that technology areas that rely more on basic scientific research, where the distribution of U.S. scientific and technological strength is more concentrated, and where the gap between U.S. and Chinese science and technology is narrower, are more likely to be the victims of the Sino-U.S. S&T friction.