Ecological efficiency, at its core, is a deceptively simple idea: how much economic and ecological benefit does a city squeeze out of every unit of land, capital, labor, and energy it consumes, while accounting for the pollution it dumps back into the environment? Unlike traditional land-use efficiency measures that reward raw growth, this metric deliberately penalizes cities for industrial wastewater, sulfur dioxide, and soot emissions. The concept, first formalized in 1990, has evolved over three decades into a cornerstone of sustainable development analysis, and China’s researchers have increasingly turned to it as the country grapples with the environmental bill of its unprecedented urbanization drive.

The research team, led by Tonghui Li and colleagues at Capital University of Economics and Business in Beijing, deployed a sophisticated measurement tool known as the super-efficiency EBM model. Unlike conventional data envelopment analysis, this approach incorporates slack variables and undesirable outputs directly into its objective function, allowing researchers to compare cities against a common efficiency frontier across the entire 2010 to 2023 study period. Inputs included built-up land area, fixed asset investment, year-end employment, and an energy consumption index derived from satellite nighttime light data, calibrated across two different satellite systems to ensure consistency. Outputs ranged from GDP and green coverage rates to industrial wastewater and air pollutant emissions.

The headline result is striking: average urban ecological efficiency climbed from 0.541 in 2010 to 0.748 by 2023, with growth accelerating markedly after 2016. Between 2010 and 2014, most cities scored below 0.5 on the efficiency scale, reflecting an era when economic slowdown, accumulated environmental pressures, and rapid urban land expansion constrained pollution abatement. By the 2020 to 2023 window, however, the number of high-efficiency regions had expanded significantly, with clusters emerging in Northeast China, the Yangtze River Delta, and the Yangtze River Basin. The researchers caution that the COVID-19 pandemic may have temporarily suppressed energy consumption and emissions in some regions, and they do not claim to have isolated the causal effect of any single policy.

Yet the national average conceals pockets of deep trouble. The study identifies what it calls efficiency collapse zones in provinces such as Shandong, Henan, and Shanxi, where ecological efficiency actually declined even as the rest of the country improved. These are not random outliers. In Shandong, the depression is concentrated in traditional industrial cities of the west; in Henan, in resource-based cities of the north; and in Shanxi, in heavy chemical industry bases of the south. All three regions share a structural signature: a heavy reliance on coal and energy-intensive industries that have proven difficult to decarbonize quickly.

Perhaps the most technically intriguing finding concerns convergence, the question of whether China’s cities are growing more alike or drifting further apart in ecological performance. The answer is nuanced. The coefficient of variation, a measure of relative dispersion, actually rose from 0.128 in 2010 to 0.214 in 2023, meaning there is no evidence of sigma-convergence, the statistical term for a narrowing gap between cities. But when the team applied a Spatial Durbin Model with two-way fixed effects, controlling for the fact that nearby cities influence one another, they found strong evidence of beta-convergence: cities that started with lower ecological efficiency grew faster than their higher-performing peers, a classic catch-up pattern rooted in neoclassical growth theory.

The convergence numbers carry real weight. Absolute beta-convergence proceeds at a speed of 0.0410 per year, implying a semi-convergence period of roughly 16.9 years, the time it takes for half of the existing gap to close. Conditional beta-convergence, which accounts for factors such as foreign openness, industrial structure, government economic influence, and environmental regulation, is slightly faster at 0.0419, with a semi-convergence period of 16.54 years. The fact that conditional convergence outpaces absolute convergence suggests that these structural factors actively shape how quickly laggard cities close the gap, a finding the authors say fills a gap in prior studies that identified convergence without explaining what drives it.

To dig into those drivers, the researchers ran two complementary analyses. An ordinary least squares regression, using one-period lagged explanatory variables and winsorized data to guard against outliers, revealed that industrial structure, economic development level, and technological level all show positive associations with ecological efficiency, with industrial structure adjustment producing the strongest positive effect. Conversely, foreign openness, local government economic influence, and environmental regulation level showed negative associations, with government economic intervention carrying the largest negative coefficient at -0.247. The authors interpret the environmental regulation result cautiously, noting that end-of-pipe treatment such as sewage processing requires substantial energy and resource inputs in the short term, which may temporarily depress measured efficiency.

But averages can mislead when dealing with a country as vast and varied as China, so the team turned to spatiotemporal geographically weighted regression, or GTWR, a technique that allows the relationship between each factor and ecological efficiency to vary by both location and time. The GTWR model dramatically outperformed the global OLS benchmark, achieving an R-squared of 0.435 versus 0.200, with lower residual error and a far better information criterion score. This confirms that the forces shaping ecological efficiency are not uniform across the country; they shift from city to city and year to year in ways a single national coefficient cannot capture.

The GTWR maps reveal a fascinating geographic evolution. In 2010, the negative association between foreign openness and ecological efficiency was widespread, consistent with the pollution haven hypothesis, in which pollution-intensive foreign investment relocates to less regulated regions. By 2023, positive associations had expanded across western China, suggesting that the composition of foreign investment and local technological conditions had shifted. The beneficial role of a service-oriented industrial structure, once concentrated in the Yangtze River Delta, had spread to provincial capitals and hub cities in the Chengdu-Chongqing and Wuhan metropolitan areas. Technological level’s positive influence also widened its footprint, while the areas where environmental regulation showed negative effects shrank notably, possibly reflecting a transition from narrow pollution-control measures toward more integrated environmental governance.

The authors are careful to frame their findings as statistical associations rather than causal proof, acknowledging that lagged variables cannot fully eliminate reverse causality or omitted-variable bias. Still, the policy implications are hard to ignore. For high-efficiency regions like the Yangtze River Delta, the combination of industrial upgrading and technological development offers a template. For struggling regions, the study points toward reducing dependence on energy-intensive industries, strengthening green technology commercialization, and rethinking fiscal mechanisms that may inadvertently reward resource-intensive growth. As China pursues its dual carbon goals of peaking emissions before 2030 and achieving neutrality before 2060, understanding which cities are converging, which are falling behind, and why could prove decisive for the next phase of the world’s largest urban sustainability experiment.

Subject of Research: Spatiotemporal evolution and convergence of urban ecological efficiency across Chinese cities

Article Title: Analysis of the spatiotemporal evolution, convergence trends, and influencing factors of ecological efficiency in Chinese cities

Article References: Li, T., Gao, W., Li, Q., & Li, X. (2026). Analysis of the spatiotemporal evolution, convergence trends, and influencing factors of ecological efficiency in Chinese cities. Frontiers in Environmental Science, 14, Article 1938089. https://doi.org/10.3389/fenvs.2026.1938089

Image Credits: AI Generated

DOI: 10.3389/fenvs.2026.1938089

Keywords: ecological efficiency, Chinese cities, super-efficiency EBM model, spatial convergence, GTWR, Spatial Durbin Model, urban sustainability, environmental regulation, industrial structure, beta convergence, dual carbon goals, nighttime light data