Drawing on the ‘buzz-and-pipeline’ framework, this paper examines the impact of intra- and interregional collaboration linkages on the innovation capacity of cities, as evidenced by the co-publication networks existing among 217 Chinese cities located in 20 city-regions. The results show that (1) intra- and interregional linkages have an inverted ‘U’-shaped and positive relationship with innovation capacity, respectively; (2) the positive effects of intra- and interregional linkages on urban innovation can be enhanced when combined with triadic closure and structural holes, respectively; and (3) there exists a complementary relation between intra- and interregional linkages in facilitating cities’ innovation capacity.

Drawing on data on scientific co-publications derived from the Web of Science for the periods 2002–2006 and 2012–2016, we construct and analyse a key element of China's intercity knowledge networks (CIKNs): scientific collaboration networks. Employing network-analytical and exponential random graph modelling techniques, we examine the evolving structures and driving mechanisms underlying these CIKNs. Our results show that the density of the CIKNs has significantly increased over time. CIKN flows are dense in the Southeastern but sparse in the Northwestern part of China, with the Hu Line acting as a clearly visible border. As the dominant knowledge centre, Beijing is involved in scientific collaboration networks throughout the country, with the diamond-shaped structure anchored by Beijing-Shanghai-Guangzhou-Chengdu becoming evident. We find that preferential attachment and transitivity are significant endogenous processes driving scientific collaboration, while a city's administrative level and R&D investment are the strongest exogenous factors. The impact of GDP and geographical proximity is limited, with institutional proximity being the most sizable of the well-known suite of proximity effects. Highlights

This paper examines the emergence of China – now the world’s largest source of scientific publications – in global science from the perspective of the connectivity of its major cities in interurban scientific collaboration networks. We construct collaboration networks between 526 major cities (including 44 Chinese cities) for 2002–2006 and 2014–2018 based on co-publication data drawn from the Web of Science. Both datasets are analyzed using a combination of different centrality measures, which in turn allows assessing the shifting geographies of global science in general and the shifting position of Chinese cities therein in particular. The results show that: (1) on a global scale, the bipolar dominance of Europe and North America has waned in light of the rise of Asia-Pacific and especially China. Most Chinese cities have made significant gains in different centrality measures, albeit that only a handful of cities qualify as world-leading scientific centers. (2) The rise in connectivity of Chinese cities is therefore geographically uneven, as cities along the East Coast and the Yangtze River corridor have become markedly more prominent than cities in other parts of China. The uneven trajectories of Chinese cities can be traced back to changing institutional, economic, and geopolitical contexts. (3) Evolution in the global scientific collaboration network exhibits strong ‘Matthew Effects’, which can be attributed to the path-dependent nature of knowledge production and preferential attachment processes in scientific collaboration.

Based on the role of innovation in domestic economic development circulation (DEDC) and the impact of international economic development circulation (IEDC) on innovation, we construct an analytical framework for the mediation effect of innovation. Drawing on China’s provincial panel data, we employ the causal steps approach to explore the impacts of IEDC on each link in the DEDC through the mediation effect of innovation. The results show that import have a direct effect on each link and export have a negative influence on distribution and consumption. Outward foreign direct investment (OFDI) has a negative impact on production and circulation through innovation. Foreign direct investment (FDI) and invisible ways have different indirect effects on each process. In addition, the influence of IEDC decreases from east to west in China, whereas import and export better promote DEDC in the central and western regions.