Purpose This study aims to address a critical gap in sustainability transitions literature by investigating the spatiotemporal evolution and, more importantly, the spatially heterogeneous driving mechanisms behind the green transition in China’s construction industry (GTCI). It moves beyond national averages to uncover how and why drivers vary across regions. Design/methodology/approach The authors use an integrated analytical framework to provincial panel data (2011–2021). GTCI level is assessed objectively using the criteria importance through intercriteria correlation–technique for order preference by similarity to ideal solution model. Its temporal dynamics and spatial distribution are analyzed through kernel density estimation and spatial autocorrelation (Moran’s I). The core of the methodology is the application of geographically weighted regression (GWR) to quantify the spatial non-stationarity of key determinants, including enterprise characteristics, environmental regulation and economic structure. Findings While the national GTCI shows improvement, regional disparities widen, forming a clear “east-high, west-low” gradient. The GWR results reveal that the effects of core drivers are not uniform but exhibit significant regional heterogeneity. For instance, the positive impact of enterprise scale and environmental regulation intensity is strongest in eastern coastal provinces and attenuates inland, while the role of ownership structure varies in direction across different geographical contexts. Originality/value This study contributes a novel, spatially explicit analytical framework to transition research. By empirically mapping the geographically varying effects of drivers, the authors provide a paradigm shift from one-size-fits-all understandings to a context-sensitive mechanism analysis. The findings offer a robust scientific basis for designing spatially tailored, evidence-based policies to achieve coordinated green development in construction and other geographically vast sectors.
Gan et al. (Fri,) studied this question.