Abstract: Network community is an important structural feature of social networks, and the changes in its members and structure have a significant impact on the access and acquisition of network resources. Based on social network theory and the knowledge-based view, we use Chinese A-share listed companies as samples to construct director networks at the corporate level and detect community structure. Then, we establish proxy variables for community evolution, divide different stages of community evolution, and empirically examine the impact of community evolution on green innovation. The results indicate that community evolution can significantly promote corporate green innovation, and the promoting effect of community evolution on green innovation is more significant during the decline stage than during the growth and stability stages,. The mechanism test finds that knowledge heterogeneity is a potential mediating mechanism between community evolution and green innovation. Further research shows that both the entry of new members and the exit of existing members have a positive impact on green innovation. Additionally, the promoting effect of community evolution on green innovation is more significant when the network community has a small size and a high density. This study not only enriches the research on social networks and green innovation from the meso-level network community perspective but also provides management insights for promoting knowledge sharing, collaborative innovation, and enhancing the quality and quantity of green innovation.

Key words: community evolution, green innovation, knowledge heterogeneity, director network