Innovation Network Embedding and Firm Innovation Performance: Evidence from the Yangtze River Delta Region

Abstract

This study examines the relationship between firms’ network embedding and innovation performance, with particular attention to the distinct roles of relational and structural embedding. Relational embedding is operationalized as the breadth of firms’ direct collaborative ties, while structural embedding is measured using network constraint. Using annual patent co-application networks involving 190 A-share listed firms in China’s Yangtze River Delta from 2017 to 2023, the study employs panel regression models with firm and year fixed effects and firm-clustered robust standard errors. The findings indicate that relational embedding is positively and significantly associated with firms’ innovation performance. This relationship remains significant after controlling for network constraint, suggesting that broader direct collaborative connectivity provides information that is not captured by structural network position. Network constraint is negatively and significantly associated with innovation performance in the linear specification. Although the quadratic model produces a positive and statistically significant squared term, the estimated turning point falls outside the observed range of network constraint. Moreover, the squared term becomes statistically insignificant when relational and structural embedding are examined jointly. Therefore, the evidence does not provide robust support for a U-shaped relationship between network constraint and innovation performance; within the observed range, greater network constraint is primarily associated with lower innovation performance. By jointly examining direct connectivity and network constraint, this study contributes to the network and innovation literature by demonstrating that relational and structural measures capture distinct dimensions of firms’ network positions. The findings provide practical implications for firms’ collaboration strategies and regional innovation policy, while offering a relevant teaching case for patent-network analysis, innovation management, and digital business education.