Eco-efficiency analysis can provide useful information about sustainability in the tourism industry, which has an important role in both global economy recovery and Sustainable Development Goals (SDGs), generating considerable indirect carbon emissions with respect to the supply chain due to its significant connections to other industries. This study, from the perspective of tourism sectors, including tourism hotels, travel agencies, and scenic spots, integrated the environmentally extended input-output analysis (EEIO) and data envelopment analysis (DEA) models to develop a research framework, analyzing the indirect carbon emissions of the tourism supply chain, evaluating eco-efficiency with respect to both direct carbon emissions and total carbon emissions (including direct and indirect parts), and exploring the driving factors of eco-efficiency of tourism sectors using Tobit regression models. This study took Gansu as a case, a province in China characterized by higher carbon intensity, an underdeveloped economy, and rapid tourism growth. The results demonstrate that (1) tourism hotels contribute the most carbon emissions in tourism sectors, especially indirectly due to the supply chain, with carbon emissions mainly resulting from the manufacturing of food and tobacco; (2) the eco-efficiency of tourism sectors in Gansu presents a U-shaped curve, which is consistent with Kuznets' theory; and (3) energy technology is key to improving the eco-efficiency of tourism sectors. The research results provide a clear path for the reduction of carbon emissions and the improvement of eco-efficiency in Gansu tourism sectors. Against the backdrop of global climate change and the post-COVID-19 era, our research framework and findings provide a reference for similar regions and countries who are in urgent need of rapid tourism development to effect economic recovery.
Keywords: carbon emissions; data envelopment analysis (DEA); eco-efficiency; environmentally extended input–output analysis (EEIO); supply chain; tourism sector.