The spatial-temporal complexity and dynamics of research collaboration: Evidence from 297 cities in China (1985-2016)

Integrating internal and external knowledge through research collaboration can be conducive to maintaining the innovative advantage. Existing studies have confirmed the important role of research collaboration in regional innovation systems, but a few of them focus on research collaboration at the city level in China. Taking advantage of a database containing 297 cities from 1985 to 2016, we explore the evolution of intra- and inter-city research collaboration and dynamics of inter-city network with a two-dimension quadrant analytical framework and social network analysis. The results are as follows: first, collaboration intensity has gradually increased and the polarization of collaboration between cities has been highlighted. Cities with weak innovation capability intend to collaborate with other cities, although such collaboration does not lead to improve their innovation performance directly. Second, inter-city collaboration is more than intra-city collaboration at the city level, and inter-city collaboration across provincial boundaries is more than inter-city collaboration within province. Third, collaboration between enterprises (EE) and collaboration between enterprises and academic institutions (AE) are the two most important patterns of research collaboration, but the dominant pattern has changed from AE to EE in China. Fourth, inter-city network presents the "core-periphery" structure and shows a decentralizing trend. Contributions of the provincial capital cities and municipalities in inter-city network are significantly greater than those of non-provincial capital cities. Besides, cities located in the eastern areas shown greater contributions than the central and western cities, while there is no obvious gap between the western and central cities.