China’s industrial green total-factor energy efficiency and its influencing factors: a spatial econometric analysis

The sustainable development of China’s economy is bottlenecked by resource shortage and environmental pollution. As the leading resource consumer and pollutant source, the industrial sector needs to improve its energy efficiency. This paper establishes a super epsilon-based measure (Super-EBM) model with bad outputs like environmental cost and evaluates the industrial green total-factor energy efficiencies (IGTFEEs) of 30 provinces in China during 2000–2017. Unlike previous research, the main contribution of this paper is to choose four environmental pollutants as bad outputs (industrial carbon dioxide, industrial sulfur dioxide, industrial chemical oxygen demand, industrial solid waste). By contrast, the previous studies mostly only take one environmental pollutant as bad output, i.e., the bad outputs are not fully measured. Then, the spatiotemporal dynamics and spatial correlations of the IGTFEEs were analyzed, and the influencing factors of IGTFEE were examined empirically with a spatial econometric model. Finally, this paper adopts generalized method of moments (GMM) to solve the endogenous problem, trying to assure the robustness of estimation results. The results show significant provincial differences in IGTFEE. Most eastern coastal provinces achieved satisfactory IGTFEEs, while most inland provinces had undesirable IGTFEEs. Eastern region achieved the highest IGTFEE, followed by central region; western region had the lowest IGTFEE. The IGTFEE improved over time in some provinces while worsened greatly in some provinces. The IGTFEE in most provinces need to be further improved. Global Moran’s I values indicate that the provincial IGTFEEs were clustered in space, rather than randomly distributed. Local indication of spatial association (LISA) map reflects significant local spatial clustering of provincial IGTFEEs. In addition, IGTFEE is significantly promoted by industrial structure, technological innovation, human capital, opening-up, and energy structure yet significantly suppressed by ownership structure and environmental regulation. Considering the endogeneity, GMM results show that the estimation results of the model were robust. Specific policy recommendations include vigorously developing high-tech industries, deepening state-owned enterprises reform, diverting more funds to research and development, cultivating versatile talents, introducing environmentally-friendly foreign capital, accelerating the implementation of clean energy development strategy, and widening the fund channels of pollution control investment.