Measuring, dynamically evolving, and converging trends of green innovation efficiency in Chinese cities

Green innovation plays a crucial role in promoting high-quality regional development. Enhancing green innovation efficiency is of great significance for implementing the innovation-driven development strategy and improving urban green development. Based on panel data from 277 prefecture-level and above cities in China from 2011 to 2021, this paper uses a super-efficiency SBM model to measure green innovation efficiency. The Kernel density estimation method and Dagum Gini coefficient are adopted to analyze the dynamic evolution trend and regional disparities of green innovation efficiency. Furthermore, spatial convergence models are constructed to explore its convergence characteristics. The findings are as follows: (1) Overall, green innovation efficiency in China shows a fluctuating upward trend, entering a stable growth phase after 2017. (2) In terms of dynamic evolution, from 2011 to 2021, urban green innovation efficiency is mainly concentrated at a medium level, showing a rightward trend and a certain degree of convergence. The eastern region shows a transition from a bimodal to a unimodal distribution, indicating a weakening polarization; the central region has the weakest agglomeration and the most severe polarization; while the western region shows some agglomeration but with a pronounced left tail. (3) In terms of regional differences, the regional disparities in green innovation efficiency among Chinese cities exhibit a fluctuating downward trend, reflecting spatial imbalance and relatively slow overall development. The eastern region has the largest internal disparities, while the central and western regions differ most from each other. Hypervariable density is the main source of regional disparities. (4) Regarding convergence, green innovation efficiency across the country and in the three major regions exhibits significant convergence, absolute spatial convergence, and conditional spatial convergence. Influencing factors vary across regions.