Descriptive statistics.

Based on panel data collected from 2003 to 2020 across 30 provinces in China, the paper employs the spatial vector angle method and spatial Durbin model to investigate industrial agglomeration’s nonlinear and spatial spillover effects on the energy consumption structure’s low-carbon transition process (Lct). The results indicate the following: First, the influence of industrial agglomeration on Lct exhibits an inverted U-shaped pattern. As the degree of industrial agglomeration expands, its effect on Lct shifts from positive to negative. Second, industrial agglomeration demonstrates spatial spillover effects. It promotes the improvement of Lct in neighboring provinces through agglomeration effects. However, the continuous expansion of industrial agglomeration inhibits the improvement of Lct in neighboring provinces through congestion effects. Third, the heterogeneity test finds that industrial agglomeration has a significant role in promoting Lct in the samples of eastern region, but this effect is not significant in the samples of western and middle regions.