Co-evolution of regional integration and green innovation under two-layer network

With global low-carbon development, green innovation has becoming a key factor for countries to obtain and maintain core competitive advantages. However, the division of scientific research resources often fails to meet the requirements of economic development. Therefore, regional integration cooperation has become an inevitable choice for innovation and development. However, the effect of the impact, direction, and interaction of regional integration policies on green innovation decision-making remains uncertain. To study the synergistic development mechanism between the two, this study constructs a two-layer network model that includes the decision and influence layers and uses a graph neural network model to map the relationship between the two-layer networks. With the help of opinion dynamics model and public goods game model, we draw the following conclusions. (1) Limited communication between regions can encourage all parties to adopt active green innovation strategies. However, excessive communication can lead to disunity of opinions among all parties, which may not be conducive to green innovation and result in an increase in free-rider behavior. (2) Regional integration development can effectively prompt all parties to choose active green innovation strategies. (3) Increasing the breadth of regional integration cooperation is more effective in promoting the active participation of all parties in green innovation, compared to increasing the depth of regional integration cooperation. This study provides theoretical support for decision-making in regional integration and green innovation development.