城际科研合作、网络结构嵌入与知识创新收敛关系研究

Research on the Relationship between Intercity Scientific Research Cooperation, Network Structure Embedding and Knowledge Innovation Convergence

摘要:

本研究基于经济收敛和社会网络理论,分析了科研合作网络影响知识创新收敛的潜在路径,提出网络构建、嵌入过程中存在的合作累积效应、知识扩散效应和网络结构效应是加速知识收敛的潜在原因。进一步,利用2000-2020年中国283个地级市组成的面板数据,以及手工整理的各城市科技论文发表、合作数据,在明确中国城际知识创新收敛特征的基础上,考察合作网络构建、嵌入对知识创新收敛的影响。具体研究发现：(1)无论是规模还是质量,中国知识创新皆存在σ收敛、绝对β收敛和条件β收敛等典型化事实。(2)当考虑中心性、结构洞等网络结构特征后,整体知识创新收敛速度得到进一步提升。(3)对大多数城市而言,中心性提升和结构洞丰富,能够促使地区有着更快的收敛速度。研究结论表明,科学边缘城市需积极融入全国科研合作网络,缩小与前沿知识的差异,缓解区域科技发展不平衡；科学中心城市可利用自身的网络位置优势,搭建信息资源平台,为网络中没有直接联系的地区提供合作机会,从而实现区域科学知识高水平收敛。

Abstract:

This study, rooted in economic convergence and social network theory, examines the impact of research cooperation networks on knowledge innovation convergence. It proposes that during network construction and embedding, cumulative cooperation effects, knowledge diffusion, and network structure effects are potential factors speeding up knowledge convergence. Using panel data covering 283 Chinese prefecture-level cities from 2000 to 2020 and manually collected data on scientific paper publishing and cooperation, we investigate the influence of cooperation network establishment and integration on knowledge innovation convergence. Key findings include: (1) China exhibits both scale and quality convergence in knowledge innovation, including σ Convergence, absolute β Convergence, and conditional β Convergence. (2) When considering network structural features like centrality and structural holes, overall knowledge innovation convergence accelerates. (3) For most cities, enhancing centrality and fostering structural holes lead to faster convergence. In summary, this study recommends that science-peripheral cities actively join the national research cooperation network to bridge knowledge gaps and address regional technology disparities. Science-central cities can use their network advantages to create information hubs and extend cooperation opportunities to unconnected regions, fostering high-level regional knowledge convergence.