Study on the evolutionary characteristics of the spatial association network of tourism eco-efficiency and its formation effect in China.

Against the backdrop of global sustainable development and ecological civilization construction, new quality productivity and high-quality development have become key drivers of industrial upgrading. Tourism eco-efficiency (TEE) serves as both a core indicator of the green and sustainable development of the tourism industry and a crucial lever for promoting high-quality development. Emphasizing technological innovation and resource optimization, new quality productivity enhances TEE levels and facilitates the green transformation of the industry. Therefore, accurately measuring TEE and exploring its spatial correlation network evolution are essential for advancing high-quality development in the tourism sector. In this work, the TEE of China's provincial regions was measured from 2011 to 2022 using a three-stage super-efficiency slack-based measure data envelopment analysis model (three-stage SE-SBM-DEA model), which was built as a TEE assessment index system. The spatial correlation network evolution characteristics and correlation network effects of TEE were analyzed using the modified gravity model and social network analysis, and the formation mechanism was explored with the help of the exponential random graph model (ERGM). The findings demonstrate that: (1) The overall mean value of TEE in China is 0.6657, with an overall fluctuating upward trend and significant regional differences. It mainly exhibits a spatial pattern of "higher in the east and lower in the west, with coastal regions outperforming inland regions." The eastern coastal regions (e.g., Shanghai, Jiangsu, Guangdong) have higher efficiency, while the western regions (e.g., Qinghai and Xinjiang) and northeastern regions (e.g., Jilin and Heilongjiang) have lower efficiency. (2) The TEE network shows an evolutionary feature from "dual-core" to "dual-core dominated and multi-polar development." The network density and efficiency are growing, and the inter-provincial spatial correlation is increasing, but the stability of the network structure needs to be further strengthened. The network relevance is always 1.0000, and the accessibility is remarkable. (3) Beijing, Shanghai and Jiangsu are the key nodes in the network, with stronger control and dominance over the elements needed to improve TEE. And regions like Xinjiang, Ningxia, Qinghai, Yunnan, Heilongjiang, Liaoning and Jilin are in a marginal position in the network, making it difficult to influence and control other regions. (4) Reciprocity, transmission, convergence, and extension in the endogenous network structure are significant factors driving the establishment of TEE networks in China. The driving relationship of heterogeneity, receiving and sending out effects of economic development level and accessibility in node attributes is obvious. The effect of external networks varies by geographical distance, and moderate distance contributes significantly to the establishment of TEE networks. This research may give important scientific support for the effective growth of TEE networks.