Source: Andres Rodriguez-Pose and Zhuoying You, “Bridging the innovation gap. AI and robotics as drivers of China’s urban innovation,” working paper from Utrecht University, 2024.
Commentary: In a recent working paper, Andres Rodriguez-Pose and Zhuoying You analyzed patent data from China’s State Intellectual Property Office, robot installation density data from the Chinese Census Database and International Federation of Robotics, and the number of keywords related to artificial intelligence (AI) in the annual work reports of 270 Chinese cities to examine the role of AI and robotics adoption in promoting local technological innovation in China. Based on these datasets, the authors found that AI and robotics directly promote technological innovation while also bolstering the impact of science and technology (S&T) investments on technological innovation. Indeed, AI and robotics were both found to be positively associated with technological innovation using the OLS and IV-2SLS econometric methods. Using the OLS method, the authors concluded that when municipal governments incorporated an additional 1 percent of AI keywords in their work reports (a measure of AI adoption), patent applications per 10,000 inhabitants increased by 0.219. Meanwhile, when one additional industrial robot per 10,000 employees was incorporated in a city, patent applications per 10,000 inhabitants increased by 0.558. Moreover, the study also concluded that AI had a statistically significant positive effect on S&T expenditures on local technological innovation.
Given that AI and robotics are found to have a positive impact on local technological innovation, the authors also examined the role of AI and robotics in closing the innovation gap between more and less innovative cities in China. Compared to S&T investments, AI and robotics were found to contribute less to a widening innovation gap between Chinese cities. Indeed, AI was found to result in about the same innovation returns for the bottom (less innovative) and top (more innovative) cities. Similarly, robot adoption was found to have a more substantial impact on technological innovation for less innovative cities than more innovative ones. This is compared to S&T investments that led to more returns for the top deciles compared to those in the bottom.
Based on their findings, the authors conclude by asserting that “AI and robotics are not only important drivers of innovation on their own, but that they also contribute to enhance the returns of S&T, the traditional bulwark of innovation policies…[Moreover, they] can become very useful tools to bridge the innovation gulf between the more and less innovative regions.”
