Characteristics and formation mechanism of carbon emission efficiency spatial correlation network: Perspective from Shandong Province

• The CEE spatial network of Shandong Province was examined. • The network structure demonstrates good connectivity and robustness. • Jinan, Qingdao, Zibo, and Dongying occupy central positions within the network. • Technology effect significantly reduces carbon emissions. • Geographic adjacency and economic differences significantly affect the CEE spatial network. Achieving the dual carbon goals requires profound systemic transformations across the economy and society. Accurately characterizing the spatial correlation network features and mechanisms of carbon emission efficiency (CEE) is critical for promoting regionally coordinated development. This study analyzes 16 prefecture-level cities in Shandong Province, China, from 2006 to 2021, using the Super-SBM model to measure urban CEE. A modified gravity model and social network analysis were combined to explore the spatial network structure, while the Logarithmic Mean Divisia Index and Quadratic Assignment Procedure models investigated its formation mechanisms. The results show that Shandong’s overall CEE has improved but remains spatially imbalanced, characterized by a persistent “higher in the east, lower in the west” pattern. Key cities, such as Jinan and Qingdao, emerge as central nodes in an evolving multi-polarized spatial network, exerting strong influence. Network density has increased over time, enhancing inter-city collaboration in emission reduction. Among driving factors, technological innovation significantly reduces emissions, though economic growth plays a stronger role in shaping network connections. Geographic adjacency further reinforces spatial correlations. While Shandong’s CEE has shown steady progress, targeted efforts in carbon management and technological advancements for less efficient cities are crucial for achieving collaborative emission reduction across the province.