Temporal-spatial evolution of green total factor productivity in China's coastal cities under carbon emission constraints

Coastal cities are the strategic core areas of China's economic growth. Measuring and understanding green total factor productivity (GTFP) growth in coastal cities is an essential prerequisite for achieving high-quality development (HQD). This paper aims to measure the accurate GTFP growth of China's coastal cities and grasp its temporal-spatial evolution. First, a three-stage dynamic analysis model is constructed to calculate the accurate GTFP growth with the effects of external environmental factors and random errors eliminated. In the context of the 3060 double carbon target, carbon reduction poses a challenge to the HQD of the coastal cities. To measure GTFP growth under the carbon emission constraints, carbon emissions are considered as one of the undesired outputs. On this basis, the temporal-spatial evolution of the GTFP index and its spatial effects are further explored using kernel density estimation and spatial Markov chains. The empirical results show that: (i) The adjusted GTFP index after eliminating the effects of environmental factors and random errors is significantly different from the original results. The GTFP index of coastal cities is underestimated without eliminating the effects of environmental factors and random errors. (ii) There are significant differences in the regional distri-bution of GTFP growth in coastal cities, with cities with faster GTFP growth gathering production factors from adjacent cities and causing a siphon effect. Finally, policy recommendations related to these findings are given.

Automated summary

This paper aims to measure the accurate GTFP growth in China's coastal cities and grasp its temporal-spatial evolution. A three-stage dynamic analysis model is used to calculate the accurate GTFP growth by eliminating the effects of external environmental factors and random errors. The study finds significant differences between the adjusted GTFP index and the original results, as well as significant regional distribution variations.