Spatiotemporal heterogeneity and driving factors of PM2.5 reduction efficiency: An empirical analysis of three urban agglomerations in the Yangtze River Economic Belt, China

Abstract Understanding the spatiotemporal heterogeneities of PM2.5 reduction efficiency (PRE) and their driving factors are substantially critical for the atmospheric environmental governance. Using the balanced panel data covering 2003–2017 of the three urban agglomerations (UAs) in the Yangtze River Economic Belt (YREB), Yangtze River Delta (YRD), Middle-Reach Yangtze River (MRYR), and Cheng-Yu (CY), as the research sample, this paper quantified the PREs by non-separable input–output SBM-Undesirable model and then investigated the spatiotemporal heterogeneities through Dagum Gini Coefficient, Kernel Density Estimation and Markov Chain, as well as analyzed the driving factors of PRE using panel quantile regression. The results show that: (1) The overall PRE of the three UAs was relatively low, with an average of 0.630, the YRD showed the highest PRE (0.684), followed by the CY (0.615), while the MRYR suffered the lowest (0.588). (2) There existed significant differences in PRE among the three UAs, and the intensity of trans-variation was the main source. The Kernel Density Estimation and Markov Chain analysis showed that the gaps of PRE between cities were narrowing and gradually converged to the middle PRE level. (3) The impacts of seven selected driving factors including fixed asset investment, economic development level, investment in education, transport infrastructure, environmental regulation, information level and trade openness on PRE in the three UAs were significantly heterogenous across cities with different PRE scores. The findings of this paper are helpful to formulate differentiated policies in PM2.5 pollution reduction and promote the coordination between economic development and atmospheric environmental protection for the three UAs in YREB accordingly.