Research on the characteristics of spatial and temporal differentiation and driving factors of green development — from the perspective of the counties in China

As an important level in the administrative system, counties have abundant natural and agricultural resources and play a key role in promoting green development strategies. This article uses the entropy method to construct a comprehensive evaluation index system for green and low-carbon development from five dimensions: green innovation, green coordination, green efficiency, green openness, and green sharing. It uses econometric methods such as Moran’s index and kernel density estimation to empirically analyze the level of green and low-carbon development in Chinese counties, differentiation characteristics, and spatial effects. Furthermore, geographically weighted regression model is introduced to explore and analyze the driving factors of various variables on the differences of green development in Chinese counties. The study finds the following: From a temporal perspective, the overall level of green development in Chinese counties shows a fluctuating upward trend; From the perspective of spatial evolution, the green development of Chinese counties presents a regional distribution characteristic of “moderate distribution in west, high in the east and low in west”; The level of green development in counties has a significant spatial positive correlation. The driving factors of economic development level, green technology innovation, and government support mainly have a positive impact on the green development of Chinese counties, while the quality of human resources mainly has a negative impact. The research results emphasize that there is obvious spatial heterogeneity in the impact of each driving factor. By implementing differentiated green development strategies to alleviate resource and environmental problems between counties, deepening the concept of green and low-carbon development, optimizing the industrial layout structure of counties, and adhering to the strategy of innovation driven development, suggestions can help counties achieve sustainable development goals.

1 Introduction

The traditional model of high energy consumption and pollution is no longer sufficient for sustainable development. The acceleration of industrialization has led to the continuous change and optimization of traditional technologies and industries, which has created enormous social wealth and driven the economic growth of countries around the world. However, natural resources and fossil fuels have been overused and overexploited giving rise to a series of serious climate and environmental problems in this development process dominated by economic growth. Excessive emission of solid pollutants and indiscriminate deforestation of virgin forests have led to increased salt crystal content in lakes and frequent occurrences of extreme weather. In order to effectively deal with global warming and the adverse effects of industrialization, the international community has begun to advocate a new development model, namely, green development.

Many scholars have published extensive studies and insights on the theoretical foundations and practical applications of the concept of green development. In a comprehensive view, existing studies mainly examine this issue from three dimensions: research perspectives, analytical methods, and driving factors. First of all, there are obvious difference of research perspectives between domestic and foreign countries. Western countries mainly focus on industry analysis, green technological innovation is regarded as an important means to realize the concept of sustainable development, and in order to effectively implement green technological innovation policies, countries around the world mainly focus on environmental regulation and financing constraints (Min et al., 2021), and confirms that green finance can promote the development of green micro-enterprises as well as realize green technological innovation (Ronaldo and Suryanto, 2022). Green bonds are one of the financial instruments for financing environmentally friendly projects that can efficiently meet ESG (Environmental, Social, Governance) objectives, while a favorable regulatory environment and high-quality disclosure are necessary conditions for the growth of green bonds (Saeed et al., 2022). China scholars, on the other hand, is mainly inclined to macro-regional studies, and the State Council issued the document on comprehensive green transformation (The State Council the People’s Republic of China, 2024), which points out that it is necessary to create a green development highland. At present, the researches object of regional green development are also mostly inclined to the areas involved in the document, and some scholars are involved in the analysis of micro-industry trends (Nie and Yao, 2023). Green development of agriculture is a key task of industrial transformation and upgrading in the new era (Zhang et al., 2024). Second, in terms of analytical methods, most scholars use a research method that combines static evaluation and dynamic analysis to analyze the spatial and temporal differences in regional development. Xiong et al. (2024) constructed an indicator system for the level of agricultural green development by combining the panel data within the Yangtze River Basin from 2006 to 2020, and utilized the entropy weight method of static evaluation to assign weights to the corresponding indicators and measure the level of agricultural green development in each province. Yang et al. (2024) used the entropy weight method, coupling the coordination degree model and obstacle degree model to analyze the spatio-temporal evolution characteristics and obstacle factors of the coupling and coordination of digital economy and green development, and put forward different types of coupling and coordinated development promotion paths. Finally, in terms of driving factors, existing studies show that various factors affect green development to different degrees. Li et al. (2023) tested the mediating effect of the intermediate mechanism of market integration affecting green development through industrial restructuring. Song et al. (2018) and other scholars pointed that an increase in energy consumption exerts a negative effect on economic growth in the long run, economic growth and industrial structure upgrading can effectively reduce energy consumption.

There are many studies already conducted on regional green development but no consensus has yet been reached on its definition (Wang and Zhang, 2020). First, from the standpoint of the research object, very few studies have been conducted on the green development of Chinese counties, most researchers concentrate on the national scope, provinces, cities, the Yangtze River Economic Belt, the Yellow River Basin or other pilot demonstration zones. Even if part of the literature is based on the county perspective, it is mostly analyzed from the micro level of the industry. Secondly, from the perspective of research methodology, the current assessment method of green development efficiency mostly adopts qualitative analysis, constructs green development level index system from the dimensions of economic level, social production, ecological environment, and researches green transformation from the perspective of “green ecology-green production-green life” three-life space, and the current construction of indicators of green development level seldom takes into account the utilization of resources and trade. Third, in terms of driving factors, the synergistic relationship between the digital economy and green development is mostly analyzed by integrating panel data within a limited region and using green total factor productivity, while the impact on urban green development is seldom analyzed by considering time and space differences. But counties as important subdivisions within China’s administrative system play a vital role, the systematic analysis of green development at the county level not only helps promote the sustainable development of local economies and societies, but also contributes to the achievement of national green development goals, thereby supporting the construction of a Beautiful China.

The study makes three key contributions to the existing literature: First, this paper clarifies the concept of green development and identifies the core dimensions of green development and constructs a comprehensive indicator system from five aspects—green innovation, green coordination, green efficiency, green openness, and green sharing—to systematically evaluate regional green development levels. Second, by integrating Moran’s I and kernel density estimation methods, this study empirically examines the spatial heterogeneity and dynamic evolution of the geographical patterns of green development across Chinese counties. Third, this paper employs spatiotemporal geographically weighted regression (GTWR) model to systematically investigate the heterogeneous effects of influencing factors and the underlying driving mechanisms of county-level green development in China.

2 Construction of comprehensive evaluation indicator system for green and low-carbon development in Chinese counties

2.1 The basic concept of county green low-carbon development

Counties should be able to optimize the structure of energy use, build new forms of high-energy-consuming industries, improve the living environment of residents in the counties to build a low carbon modern industrial system and realize green low carbon development (Zhang and Zhong, 2022). The connotation of green development can be elaborated from five dimensions. (1) Green innovation. The innovation dimension exhibits a direct theoretical alignment with ecological modernization theory. Ecological modernization theory emphasizes achieving a relative decoupling between economic growth and environmental improvement through technological progress, institutional innovation, and transformations in production modes (Jing and Gao, 2021). Green innovation constitutes the core driving force of the ecological modernization process. Therefore, incorporating innovation as a key dimension of green development reflects the theoretical logic that environmental performance can be enhanced through technological advancement and knowledge accumulation. (2) Green coordination. The coordination dimension primarily corresponds to sustainable regional development theory. This theory underscores the coordinated evolution of economic, social, and ecological systems across spatial and temporal dimensions, with particular attention to the allocation of internal regional factors and the coordination between urban and rural areas as well as among industries (Liao and Jiao, 2005). Green development requires the formation of dynamic balance among various development elements so as to avoid resource misallocation and the spatial transfer of environmental pressures. Accordingly, green coordination embodies the theoretical requirement for achieving systemic optimization and structural balance at the regional level. Counties are expected to play a bridging role by exploring diversified, differentiated, and high-quality pathways for urban and rural integrated development, thereby ensuring a transition toward a higher-quality development stage (Zhai and Hou, 2020). (3) Green efficiency. The efficiency dimension is closely related to theories of resource allocation efficiency and the internalization of externalities in environmental economics. Environmental economics argues that environmental problems fundamentally stem from inefficient resource allocation and the existence of environmental externalities. Enhancing the efficiency of energy, land, and capital use can reduce pollutant emissions under given resource constraints and achieve an optimized balance between economic output and environmental costs. High-quality regional economic development is fundamentally driven by the promotion of green development. Existing studies on green development efficiency indicate that disparities and complexities in efficiency levels arise from the combined effects of uneven natural endowments and socio-economic conditions (Guo et al., 2022). (4) Green openness. The openness dimension can be interpreted from the perspectives of ecological modernization and international division of labor in environmental economics. Openness not only entails the liberalization of trade and investment but also reflects mechanisms of technology diffusion, institutional learning, and the transmission of environmental standards. Through external openness, regions can introduce more advanced green technologies and management experience, thereby enhancing local environmental governance capacity and overall green development performance. (5) Green sharing. The sharing dimension is highly consistent with the principles of social equity and intergenerational justice embedded in sustainable development theory. Green development places emphasis not only on efficiency and growth but also on the equitable distribution of development outcomes across different social groups and regions. By improving the provision of public services and ecological welfare, the sharing dimension helps strengthen the social foundation and long-term sustainability of green transformation. In 2018, the State Council issued the Opinions on Establishing a More Effective Mechanism for Coordinated Regional Development which explicitly stated that building a new framework for coordinated regional development characterized by efficiency, fair competition, environmental harmony, and shared outcomes requires resolutely breaking down interregional interest barriers and policy fragmentation, and accelerating coordinated development across provincial boundary areas (Zhang and Xiong, 2024).

In summary, the five dimensions of innovation, coordination, efficiency, openness, and sharing respectively correspond to technological progress, structural coordination, resource allocation, external connectivity, and social equity, and collectively resonate with ecological modernization theory, sustainable regional development theory, and environmental economics. Together, they provide a relatively solid theoretical foundation for constructing a multidimensional evaluation framework for green development.

2.2 Construction of comprehensive evaluation index system

This study combined the geographical characteristics and development mode of county economic entities as well as the existing research choose 19 county level indicators from five parts: green innovation, green coordination, green efficiency, green openness, and green sharing by adhering to scientific, comprehensive, and data availability. The fixed-line telephone Subscribers to measure the level of telecommunications infrastructure development consistent, although this indicator may indeed appear somewhat outdated in the context of the rapid development of the digital economy, at the county level, more representative measures such as broadband penetration, mobile data traffic, or the use of digital platforms, often suffer from substantial data gaps or inconsistencies in statistical definitions over long time periods. By contrast, the number of fixed-line telephone subscribers exhibits greater statistical stability and temporal continuity, and can reflect the historical accumulation of basic communication infrastructure and the early coverage of information networks (Zheng et al., 2014). Following existing studies (Fan et al., 2003), financial support is measured by the number of financial institution outlets which reflects the accessibility and coverage of financial services within a region. Referring existing studies (King and Levine, 1993), innovation input is proxied by financial and fiscal support indicators. Specifically, the ratio of Year-end loan balance of financial institutions and Per capita general budget expenditure to GDP (Gross Domestic Product) is used to capture support for innovation, innovation output is measured by Invention patent applications and utility model patent applications Griliches (1990). In the urban and rural coordination dimension, this study measures coordination from two aspects: urban-rural financial coordination and urban-rural commercial circulation. Following Fan et al. (2014), the ratio of percapita savings deposit balance of urban and rural residents to GDP is used to reflect the coordination of financial resource allocation and drawing on Tang (2025). The use of the Belt and Road Initiative (BRI) dummy variable to measure the degree of openness is primarily grounded in the analytical framework of institutional openness and spatial locational advantages derived from theories of regional development and external openness. Through policy support, infrastructure connectivity, and trade facilitation, the BRI has exerted significant effects on external economic linkages and factor mobility in regions along its routes. Although the initiative exhibits a certain degree of exogeneity, its influence on county-level openness tends to materialize gradually through improvements in transportation conditions, adjustments in industrial layout, and the intensification of economic activities (Liu and Ran, 2023). Industrial structure coordination is measured by the ratio of value added in the tertiary industry to that in the secondary industry which captures the relative advancement and coordination of the industrial structure. Following the approach of Guo et al. (2023), this study uses the share of land transactions as a proxy to measure the level of resource utilization. Economic growth quality is captured from four dimensions: Per capita fiscal general budget revenue to GDP (Chen et al., 2025), GDP per capita level (Chen et al., 2025), Satellite Lighting Data (Xu et al., 2015), and the number of newly registered Business Data (Chen et al., 2025). The degree of foreign trade openness is measured using counties located in regions along the Belt and Road Initiative, including counties within the capital cities of related provinces as well as selected key port counties (Li and Li, 2025). Domestic trade is captured from both the demand and supply sides, the ratio of total retail sales of consumer goods to GDP is used to measure domestic demand and internal trade activity (Gong and Liu, 2025), while the number of industrial enterprises above-scale reflects the industrial supply capacity supporting domestic trade(Xu and Li, 2025). Green production is measured by Carbon dioxide emissions (Tong, 2020) and PM2.5 annual average emission (Tong and Li, 2019). Coverage of Forests, Grasslands and Waters is used to measure the green Life (Wu and Liu, 2023).

2.3 Research area and data source

Considering the accuracy and operability of data acquisition, the research object determined in this paper is China’s county-level administrative regions from 2000 to 2021. Most of the research data come from China Statistical Yearbook, China Urban Statistical Yearbook, China County Statistical Yearbook and official government websites, and a small part of the data come from the statistical bulletins of cities and counties, the patent data involved come from the State Intellectual Property Office, and the energy data come from China Energy Statistical Yearbook. The 916 counties have been obtained by eliminating and reducing the missing values of the data. In this study, missing values are imputed using interpolation. Interpolation inherently relies on observations from adjacent time points, and the imputed values typically exhibit linear or smoothed patterns which may underestimate the actual volatility of the indicators in real-world conditions. Given that the interpolated data primarily concentrate on earlier years, the expected disturbance to the relatively discrete spatial structure across counties is limited. Theoretically, the direction and magnitude of bias in the Moran’s I index after interpolation depend on the spatial distribution pattern of missing data rather than the missing rate itself. In this study, missing data exhibit a random scattered distribution (non-clustered missingness), allowing for controllable effects on both global and local spatial autocorrelation patterns. Moreover, the locally weighted estimation characteristic of the GTWR model inherently exhibits robustness to interpolation noise. Even if individual interpolated values are biased, their impact remains confined within local spatiotemporal windows. Consequently, this study has successfully contained potential influences within an acceptable range.

2.4 Research methods

2.4.1 Entropy value method

In this paper, we first adopt the equal weight to construct the indicator weights, which is a simple and direct weight allocation method that assigns the same weight to all evaluation indicators, usually 1 divided by the total number of indicators. Before conducting the data-driven adjustment of indicator weights in the entropy method, adopting equal-weight initialization as the baseline setting has clear methodological justification and statistical rationality. The core principle of the entropy method lies in using the information entropy of each indicator to capture its degree of variation, thereby enabling an objective correction of weights. Therefore, assigning differentiated weights based on subjective judgment prior to the entropy-based computation would interfere with the subsequent process in which weights are “self-organized” by the data. Equal-weight initialization, serving as a zero-bias starting point, eliminates the influence of subjective preferences on the prior weight structure and ensures that the final weights are entirely determined by the empirical distribution of the indicators, thus maintaining methodological objectivity and neutrality. Meanwhile, from an operational perspective, equal-weight initialization guarantees that all standardized indicators share a consistent scale prior to entropy calculation which is essential for ensuring the comparability and validity of the entropy-derived weights (Xiong et al., 2024). The following are the precise stages in the calculation:

i. Data standardization: different types of indicators are standardized in different ways. The calculation process is shown in Equations 1, 2.

Positive indicators:

${\mathrm{X}}_{\text{ij}}^{\prime }=\frac{{\mathrm{x}}_{\text{ij}}-\min {\mathrm{x}}_{\mathrm{j}}}{\max {\mathrm{x}}_{\mathrm{j}}-\min {\mathrm{x}}_{\mathrm{j}}}(1)$

Negative indicators:

${\mathrm{X}}_{\text{ij}}^{\prime }=\frac{\max {\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\text{ij}}}{\max {\mathrm{x}}_{\mathrm{j}}-\min {\mathrm{x}}_{\mathrm{j}}}(2)$

Among them, $i$ and $j$ denote the county and the measure respectively and $X_{ij}^{\prime }$, $x_{ij}$, $\max x_{ij}$ and $\min x_{ij}$ denote the positive and negative standardized values, actual observed values, maximum values and minimum values of the corresponding measure $x_{ij}$ respectively.

ii. Calculate the weight of the indicator p: Calculate the weight of the first sample of $\mathrm{i}$ under the indicator $x_{ij}$, and $p_{ij}$ is the weight of the first sample of $\mathrm{i}$ in the indicator j. The calculation process is shown in Equations 3.

$p_{ij}=\frac{X_{ij}^{\prime }}{{\sum }_{i=1}^n{X}_{ij}^{\prime }}(3)$

iii. Calculate the information entropy e with the Equations 4:

$e_j=-k{\displaystyle \sum _{n=1}^n}\left(p_{ij}*\ln p_{ij}\right)(4)$

iv. Calculate the information redundancy of $d_j$ Equations 5:

$d_j=1-e_j(5)$

v. Determine the weight of each indicator w Equations 6:

$w_j=\frac{d_j}{{\sum }_{j=1}^n{d}_j}(6)$

The results of calculating the weights of each indicator to measure the level of green and low-carbon development of the county are shown in Table 1.

vi. Calculate the comprehensive score Equations 7:

$Score={\displaystyle \sum _{j=1}^n}\left(w_j*X_{ij}^{\prime }\right)(7)$

Table 1

Table 1. Comprehensive evaluation indicator system for county-level green and low-carbon development.

Score is the comprehensive index of green low-carbon development of the ith county. The comprehensive score is the decisive index used by the entropy value method for the final comprehensive evaluation. The comprehensive score of each evaluation index is obtained by multiplying the data of each index with the corresponding weights and then accumulating them to measure its comprehensive level.

To validate the impact of index construction methods on research conclusions, this study re-constructed the green development index using an equal-weighting approach. The global correlation coefficient test revealed that the equal-weight index correlated with the benchmark index at 0.841. Annual analysis reveals the correlation coefficient remained stable between 0.765 and 0.925 from 2000 to 2021 with a standard error of only 0.018. All indices revealed highly consistent spatiotemporal patterns despite structural differences in weighting schemes. In summary, the equal-weighting validation confirms the methodological robustness of this study’s conclusions on spatiotemporal differentiation in county level green development, demonstrating that core methodological findings remain unaffected by weighting scheme choices.

2.4.2 Kernel density estimation

The smoothed probability density of random variables is mostly estimated using the nonparametric technique known as kernel density estimation (KDE). The overall probability density image is constructed by applying a smoothing function, the kernel function, to each data point and superimposing these functions. By not presetting the form of a particular distribution, the KDE is able to flexibly adapt to data distributions of different shapes. The fundamental principle of this approach is using kernel functions to calculate a weighted average of data points based on time series changes. This paper utilizes the kernel density estimation curve position, shape, kurtosis and other characteristics to demonstrate the efficiency evolution characteristics, location and extensibility of the county’s green and low-carbon development index in different phases, with the specific formulas with Equations 8, 9:

$f\left(x\right)=\frac{1}{N\times h}{\displaystyle \sum _{i=1}^N}K\left(\frac{X_i-x}{h}\right)(8)$

$K\left(x\right)=\frac{1}{\sqrt{2\pi }}{e}^{-\frac{1}{2}{x}^2}(9)$

Where: $f\left(x\right)$ is the density function measuring the level of green low-carbon development of counties; N is the number of observations; $X_i$ is the number of independently and identically distributed observations; h is the optimal bandwidth determined by the “thumb rule”; too large a bandwidth will result in a smoother curve and lower estimation accuracy; too small a bandwidth may result in overfitting, and interference by the noise term.

2.4.3 Center of gravity - standard deviation ellipse analysis method

The center point of the ellipse indicates the center of gravity of the data which can respond to the overall location of regional economic activities or resource distribution, and is crucial to the investigation of the spatial change law of unequally distributed attributes. This analysis method is widely used in Geographic Information System which can visualize the trend and pattern of data distribution in space (Liang et al., 2022). The specific calculation formulas are as shown in Equations 10–13.

$\text{Mean\hspace{0.17em}center}:\overline{X_w}=\frac{{\displaystyle {\sum }_{i=1}^n}{w}_i{x}_i}{{\displaystyle {\sum }_{i=1}^n}{w}_i};\overline{Y_w}=\frac{{\displaystyle {\sum }_{i=1}^n}{w}_i{y}_i}{{\displaystyle {\sum }_{i=1}^n}{w}_i}(10)$

$\mathrm{X}-\text{axis\hspace{0.17em}standard\hspace{0.17em}deviation}:{\sigma }_x=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^n}{\left(w_i{\stackrel{\sim }{x}}_i\cos \theta -w_i{\stackrel{\sim }{y}}_i\sin \theta \right)}^2}{{\displaystyle {\sum }_{i=1}^n}{w}_i^2}}(11)$

$\mathrm{Y}-\text{axis\hspace{0.17em}standard\hspace{0.17em}deviation}:{\sigma }_y=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^n}{\left(w_i{\stackrel{\sim }{x}}_i\cos \theta -w_i{\stackrel{\sim }{y}}_i\sin \theta \right)}^2}{{\displaystyle {\sum }_{i=1}^n}{w}_i^2}}(12)$

$\text{Azimuth\hspace{0.17em}angle}:\tan \theta =\frac{\left({\displaystyle {\sum }_{i=1}^n}{w}_i^2{{\stackrel{\sim }{x}}_i}^2-{\displaystyle {\sum }_{i=1}^n}{w}_i^2{{\stackrel{\sim }{y}}_i}^2\right)+\sqrt{{\left({\displaystyle {\sum }_{i=1}^n}{w}_i^2\times {\stackrel{\sim }{x}}_i^2-{\displaystyle {\sum }_{i=1}^n}{w}_i^2\times {\stackrel{\sim }{y}}_i^2\right)}^2+4{\displaystyle {\sum }_{i=1}^n}{w}_i^2\times {\stackrel{\sim }{x}}_i^2\times {\stackrel{\sim }{y}}_i^2}}{2{\displaystyle {\sum }_{i=1}^n}{w}_i^2{\stackrel{\sim }{x}}_i{\stackrel{\sim }{y}}_i}(13)$

Among them, $\left(x_{i,}{y}_i\right)$ denotes the spatial location of the county’s green low-carbon development level, $\left(\overline{X_{w,}}\overline{Y_w}\right)$ denotes the weighted mean center, ${\stackrel{\sim }{x}}_i$ and ${\stackrel{\sim }{y}}_i$ denote the coordinate deviation from the location of each research object to the mean center, respectively; $w_i$ denotes the weight, n denotes the total number of elements under study, and $\theta$ is the ellipse azimuthal angle, which denotes the angle formed by rotating clockwise to the long axis of the ellipse from the due north direction.

2.4.4 Spatial autocorrelation analysis

Spatial autocorrelation analysis mainly assesses the degree of aggregation or disaggregation of attribute values of spatial elements through global and local autocorrelation indices (Han et al., 2024). In economics, this analysis can be used to study the spatial distribution characteristics of regional economic development levels. That is, whether there is a spatial interaction of an economic indicator among regions within a country (The calculation results are shown in Table 2), and if so, whether it is mutually reinforcing or inhibiting is analyzed. And the specific calculation steps are shown as Equations 14, 15.

$I=\frac{n}{S_0}\times \frac{{\sum }_{i=1}^n{\sum }_{j=1}^n{w}_{ij}\left(y_i-\overline{y}\right)\left(y_j-\overline{y}\right)}{{\sum }_{i=1}^n{\left(y_i-\overline{y}\right)}^2}(14)$

$I_i=\frac{Z_i}{S^2}{\displaystyle \sum _{j\ne i}^n}{w}_{ij}{Z}_j=\frac{y_i-\overline{y}}{S^2}{\displaystyle \sum _{j\ne i}^n}{w}_{ij}\left(y_i-\overline{y}\right)(15)$

Table 2

Table 2. Correlation coefficient test.

The sum of all the components in the spatial weight matrix is S₀, and n is the total number of spatial units among them. $w_{ij}$ is the spatial weight between spatial units i and j. $y_{i,}{y}_j$ is the attribute value. $Z_i$ is the normalized attribute value, and $S^2$ is the variance of all the attribute values. The geographic autocorrelation throughout the entire study region is reflected in $I$; the higher the number, the more pronounced the spatial correlation. $I_i$ denotes the local Moran’s I index, which is used to identify local aggregation phenomena or outliers.

2.4.5 Spatio-temporal geographically weighted regression model

The spatio-temporal geographically weighted regression (GTWR) model can simultaneously capture spatial heterogeneity and temporal non-stationarity, allowing regression coefficients to vary across both space and time. In contrast, GWR and Multiscale Geographically Weighted Regression (MGWR) only account for spatial variations, while spatial panel models typically assume globally homogeneous parameters, making them inadequate for identifying localized relationships that evolve over time. Therefore, adopting GTWR enables a more comprehensive representation of spatial positional effects and effectively addresses both temporal and spatial non-stationarity, allowing for a more accurate examination of the spatial heterogeneity of driving factors (Weng et al., 2024). The expression is shown as Formula 16:

$y_i={\beta }_0\left(u_i,v_i,t_k\right)+{\displaystyle \sum _{j=1}^p}{\beta }_j\left(u_i,v_i,t_k\right)x_{ij}+{\epsilon }_i(16)$

Where: i = 1, 2, …, n; j = 1, 2, …, m; k = 1, 2, …, T; $x_{ij}$ represents the value of the $j$ variable at point $i$, $y_i$ indicating the green development index of the ith research object at time $t_k$; $u_i,v_i,t_k$ represents the latitude and longitude coordinates and time variables respectively, ${\epsilon }_i$ is the random perturbation term, ${\epsilon }_i\sim N\left(0,{\sigma }^2\right)$; ${\beta }_j\left(u_{i,}{v}_{i,}{t}_k\right)$ is the regression coefficient of the jth variable of the ith research object at time $t_k$, and ${\beta }_0\left(u_{i,}{v}_{i,}{t}_k\right)$ is the intercept.

3 Spatial and temporal evolution of green development level in Chinese counties

3.1 Characteristics of temporal changes in the level of green development in Chinese counties

This paper takes 3 years as the interval point, selects 2000, 2003, 2006, 2009, 2012, 2015, 2018, and 2021 as the observation time point, and draw the Gaussian kernel density two-dimensional distribution map of these 8 years, as shown in Figure 1.

Figure 1

Figure 1. Distribution dynamics of green development level in Chinese counties.

It can be concluded that the green development level of Chinese counties shows an overall upward trend from the trend of each curve. This may be attributed to national policy promotion, emerging technology progress, changes in market demand between 2000 and 2009, the green and low-carbon development index of Chinese counties was mostly concentrated in the range of 0.0 and 0.2, and the kernel density in each year was generally low and fluctuated little. There is gradual decrease in the height of the peak, and the density curve was shifted to the right as a whole showing the phenomenon of right-tailed trailing and the trend of convergence in 2012. The level of green development is increasing and regional differences are shrinking. Starting from 2015, there is an obvious right trailing, and the peak position of kernel density gradually shifts to the right, between 0.3 and 0.35, showing a multi-peak distribution and a rapid decrease in the peak height, showing a dynamic convergence trend. During the period from 2018 to 2021, the height of the main peak is increasing and the right trailing phenomenon is weakening, indicating that the difference in green development level between Chinese counties is gradually narrowing in this period, or the growth rate of green development level of some leading county administrative units is slowing down. This change may be related to the balanced development of national policies, the improvement and optimization of resource allocation, the updating of science and technology, and the increased attention to green development in each region. The construction and development of these new industries often require a high level of green development and potential, making the green development level index shift; consumer demand for green products tends to be diversified and high-end, not only focusing on the environmental protection of the product itself, but also requiring manufacturers to focus on the green standards of the product production process and supply chain.

3.2 Features of the spatial distribution of China’s green development levels

The evolution of China’s county-level green development levels at three different time points: 2001, 2011, and 2021. The natural breakpoint method is used to classify the green development level of Chinese counties into four levels, which are low level region, lower level region, higher level region and high level region base on the combined data (see Table 3).

Table 3

Table 3. Division criteria for comprehensive evaluation of green development level of Chinese counties.

According to the results presented in Figure 2, it can be seen that the green development level of Chinese counties has obvious spatial heterogeneity. The green development level of county-level administrative units in each region shows large fluctuations and differences in the three selected study periods. Counties with higher and higher levels of green development nationwide show an expanding trend from 2001 to 2021. In terms of spatial distribution, the green development level of Chinese counties can be roughly summarized as the distribution characteristic of “high in the east, low in the middle, and moderate in the west” (Zhirong et al., 2022).

Figure 2

Figure 2. Evolution of spatial pattern of green development in Chinese counties in 2001, 2011, and 2021. Note: Produced according to the standard map of the Ministry of Natural Resources Review No. GS(2019)1822, with no modification of the base map. (a) 2001 year. (b) 2011 year. (c) 2021 year.

Most of China’s counties were at low and lower levels of green development in 2001. Counties with high levels of green development included only Longquan City and Anxi County in Zhejiang Province, Zhengxiangbai Banner in the Inner Mongolia Autonomous Region; and Jinghe, Luntai and Qinghe counties in the Xinjiang Uygur Autonomous Region, among several others. This phenomenon may be constrained by factors such as economic development, industrial structure and resource endowment. First, from the perspective of economic development. China began a new phase of development in 2001. Significant amounts of pollutants and dangerous gasses were created and released in an effort to support the quick economic growth. Secondly, from the perspective of industrial structure, the traditional energy pattern of “more coal, less oil, less gas” makes heavy industry and high energy-consuming industries occupy the “leading” position in the construction and development of various industries, which bring high income and high returns, but also lead to high resource consumption and high pollution emissions (Zhong et al., 2024). In addition, there are significant differences between the counties in China from the point of view of resource endowment. The counties with abundant material resources may rely more on resource development such as in the central region, the resource advantage formed by the coal resource endowment determines the rapid expansion of energy infrastructure and the construction of crude economic development so its green development level is at a lower level. However, green development is given priority in areas with comparatively less resources and slower development such as the Xinjiang Uygur Autonomous Region, and southwestern counties like Gansu as well as in the central-western border regions. The lagging level of green development in China’s counties in 2001 was caused by a development model that sacrificed the environment.

In 2011, the level of green development of Chinese counties has increased considerably compared with that of a decade ago. With the exception of counties in central China, such as Anhui and Henan, and individual counties of Jinta County and Dunhuang City in Gansu Province, and Shaya County in Xinjiang Uygur Autonomous Region, which are still at the low level of the first grade, the green development level of counties in other regions has generally increased to the second and third grades. In particular, some counties in the southeastern regions of Jiangsu, Zhejiang and Guangdong have moved to a higher level of green development. This may be attributed to national policy support and regional technological innovation. Since the United Nations put forward the “Green Economy Initiative” in 2008 and the release of the “Global Green New Deal Policy Outline” in 2009, it has triggered global attention to green technological innovation. It is gradually recognized that the environmental problems caused by the excessive emission of pollutants and harmful gases are very serious and bad (Jiang and Liu, 2024). The national government publicly released a notice on the development of carbon emission trading, prompting the government to provide the international community with information on carbon emission trading (Guo and Yong, 2023), prompting enterprises to transform to low energy consumption and high efficiency. In accordance with national policies, the level of green development in some western counties, represented by Fuyun County, Heshuo County, and Yuli County in Xinjiang Uygur Autonomous Region, has gradually improved, and spatially presents the characteristics of continuous cluster development, further narrowing the gap with counties in other regions. Most of these counties have rich natural resources and extremely high forest coverage, and they are able to fully utilize their unique resources to improve economic efficiency, and also maintain a leading position in green transformation. In addition to national policy support, the innovation and promotion of energy-saving technologies not only realizes industrial restructuring, but also greatly improves the level of green development. Some counties in the eastern coastal region, such as Dongtai City in Jiangsu Province, Guangrao County in Shandong Province, and Yuyao City in Zhejiang Province, have taken advantage of the government’s policy support, combined with well-developed industrial infrastructures and technological innovation capabilities, to show clear advantages in green development.

The level of green development in each county will be significantly improved except for some counties in central China in 2021, such as Henan, Anhui, and Shanxi, where green development remains at a relatively low level, nearly two-thirds of counties are at the third or fourth level of green development. Counties in eastern China have seen the most significant improvement in green development, with most reaching the fourth level of high-level development. 2021 is the first year after China put forward the “double carbon” goal, and also the first year of the national carbon emissions trading market, the state will be the first year of the carbon emissions trading system and the registration and settlement system. The national carbon emissions trading system and registration and settlement system are set up in the Yangtze River Economic Belt which not only establishes the Yangtze River Economic Belt as a pillar in the national carbon trading market but also promotes the construction of green development of counties in the region (Xie et al., 2024). For example, Nanchang County in Jiangxi Province, Tianmen City in Hubei Province, and Xiangyin County in Hunan Province have all improved their green development level from a low level stage to a high level stage in 10 years, and the construction of regional green development has achieved remarkable results. Furthermore, a number of counties in eastern China are situated close to the sea, including Leting County in Hebei Province, Haiyan County in Zhejiang Province, and Junan County in Shandong Province. The superior geographic location has given them many excellent harbors and a developed marine economy and tertiary industry which is why the green development is more efficient. While the development level of some counties in the northeast region is still at a lower level probably due to the concentration of heavy industry in the north, high energy demand, and the main reliance on coal-fired heating in the winter, which leads to the limitation of green development (Li and Du, 2023).

3.3 Center of gravity migration analysis of green development level in Chinese counties

Different morphological changes of the standard deviation ellipse indicate different spatial displacements when the standard deviation ellipse is in contraction state (As is shown in Table 4 and Figure 3). It indicates that the distribution of green development location in Chinese counties shows regional aggregation effect. When the standard deviation ellipse is in expansion state, it indicates that the distribution of green development location in Chinese counties shows regional diffusion effect, and when the shape of the standard deviation ellipse is unchanged, it indicates that the distribution of green development location in Chinese counties is kept stable (Yang et al., 2022). The overall “northwest-southeast” spatial distribution pattern was presented by the standard deviation ellipse of China’s county-level green development spatial distribution from 2000 to 2021 which nearly encompassed the middle and eastern regions.

Table 4

Table 4. Elliptic eigenvalues of standard deviation of green development level in Chinese counties.

Figure 3

Figure 3. Standard deviation ellipse analysis of green development in Chinese counties in 2001, 2011, and 2021. Note: Produced according to the standard map of the Ministry of Natural Resources Review No. GS(2019)1822, with no modification of the base map.

From the position of the center of gravity of the standard deviation ellipse, the center of gravity of the distribution of the green development level of Chinese counties in the sample period from 2000 to 2021 ranges from 32.64°N to 32.89°N, 111.13°E to 112.11°E. From the trajectory of the center of gravity of the standard deviation ellipse, the center of gravity migrates to the southeast during the sample period, and it is initially shifted from Xichuan County, Nanyang City, Henan Province to the northeast to Zhenyang City, Nanyang City, Henan Province, and then to the northeast. The center of gravity shifted from Xichuan County in Nanyang City, Henan Province to the northeast to the border between Zhenping County and Dengzhou City in Nanyang City, but the whole center of gravity was attributed to Zhenping County. Afterwards, it moved in a southwesterly direction, centering on Zhenping County, to Danjiangkou City, Shiyan City, Hubei Province. From the standard deviation ellipse center of gravity distance, from 50.42 to 93.21 km, the center of gravity distance shows an obvious increasing trend, and the spatial gap of green development in counties shows a gradually expanding trend.

In terms of the change of the long and short axes, the standard deviation ellipse’s long half-axis gradually shrinks, indicating a gradual reduction in the gap between the east and west China counties’ levels of green development. China was going through a period of fast economic development in 2001. And the main economic benefits came from the heavy industry and high-energy-consuming enterprises which led to the problems of pollution and hazardous gas emissions. And the western region by virtue of its natural geographic location has a large amount of green vegetation and there are fewer sources of pollution. The western region of the county green development is more efficient, and the distribution of the county green development and construction of the various regions is more decentralized. However, as China’s awareness of environmental protection has been raised, while the traditional economic structure has been adjusted, the traditional crude economic growth has been gradually outlawed by the quality and efficiency intensive growth. The eastern region has emerged in the new round of green development construction, leading the way in China’s county green development construction through research and development, investment in new energy sources, promotion of green industry construction, use of green credit, green bonds and other financial instruments. It reflects that the imbalance of green development among regions has been improved to some extent. The short semi-axis decreases and then increases, indicating that in the north-south direction the green development of counties initially shows a clustering and synergistic development, and then shows a diffusion trend, i.e., there are differences in the mode and level of green development among counties. The Chinese government put forward different strategic plans in the 13th Five-Year Plan regarding the green development construction in different regions (Zhong and Shao, 2022). This may have promoted the synergistic and differentiated development of counties in different green industry fields or ecological protection modes, such as eco-tourism, organic agriculture, clean energy development, etc., which made the distribution of the short semi-axis expand again.

In terms of azimuth change, the azimuth angle increases from 150.39 ° to 151.52 °, with a small range of change, showing a small clockwise rotation of the ellipse, which indicates that counties located in the northwestern or southeastern direction have a relatively faster level of green development. Afterwards, the azimuth angle shrinks from 151.52 ° to 21.10 ° with a large counterclockwise rotation of the ellipse, which indicates that the center of gravity of China’s green development in counties has shifted from the southwest direction to the northeast direction more obviously, the level of innovation and development of the cities located in the southwest or northeast direction is relatively fast. From the standard deviation ellipse area, it displays a trend of going down and then up, and the standard dispersion oval shape progressively trends toward a positive circle, indicating that the green development of counties inside the ellipse was faster than that of counties outside the ellipse from 2001 to 2011. But in the year of 2011 and 2021, counties outside the ellipse had a higher degree of green development than counties inside. This shows that the spatial distribution pattern of green development level in Chinese counties is characterized by “agglomeration first and expansion later”. From the perspective of the flat rate index, the flat rate shows a decreasing trend from 2001 to 2021, indicating that the spatial distribution of green development in counties is unbalanced.

3.4 Features of the spatial clustering of China’s green development levels

3.4.1 Global spatial auto-correlation

According to the measurement results listed in the Table 5, it can be found that the global Moran’s I index of green development of Chinese counties always maintains a positive value during the study period, and all of them pass the Z test at 1% significance level, rejecting the original hypothesis, that is to say, there is a 99% certainty that this data is non-randomly distributed, and the possibility of random distribution is less than 1%. It suggests that there is a significant positive aggregation effect in the spatial distribution and that the green development of Chinese counties during this period exhibits significant spatial positive correlation characteristics. Further examination of the dynamic evolution trend of Moran’s I index, as shown in Figure 4, reveals that the index generally exhibits a declining trend, suggesting that the spatial agglomeration effect of China’s green development is progressively waning, though there are minor oscillations in the decline process. Specifically, from 2000 to 2010, the Moran’s I index of green development in Chinese counties showed a small decline, thus indicating that the spatial agglomeration effect of green development in counties weakened during this period. Then from 2010 to 2017, the global Moran’s I index dropped sharply from 0.5740 to 0.2690, with the positive spatial correlation fluctuation increasing and the differences among counties increasing. The global Moran’s I index showed a small increase from 2017 to 2019, and then the index slowly dropped to 0.2779, reflecting the spatial green development of Chinese counties and regions synergy weakened.

Table 5

Table 5. Global Moran’s I test of green development index of Chinese counties from 2000 to 2021.

Figure 4

Figure 4. Global Moran’s I value of green development in Chinese counties from 2000 to 2021.

3.4.2 Local spatial auto-correlation

The global Moran’s I index test aims to determine the overall correlation of green development of Chinese counties. To examine the clustering features of counties’ green development levels in further detail, this study selected the data in 2005, 2010, 2015, and 2020, and visualized them with reference to Moran’s scatter plot and combined with ArcGIS10.8 software, as shown in Figure 5. The results show that the green development of Chinese counties presents four types. Among them, the H-H and L-L patterns represent the homogenization development tendency among counties, i.e., counties with higher (lower) levels of green development are surrounded by neighboring counties with similarly high (low) levels of green development, and a certain positive correlation is shown between them; the H-L and L-H patterns show the heterogeneous development trend among counties, i.e., counties with higher (lower) levels of green development are surrounded by counties with lower (higher) levels of green development, and the difference in green development among counties is more significant.

Figure 5

Figure 5. Spatial distribution of Moran’s I scatter points for green development in Chinese counties in 2005, 2010, 2015, and 2020. Note: Produced according to the standard map of the Ministry of Natural Resources Review No. GS(2019)1822, with no modification of the base map. (a) 2005 year. (b) 2010 year. (c) 2015 year. (d) 2020 year.

As can be seen from Figure 5, counties in China that showed obvious spatial positive correlation in green development from 2005 to 2015 accounted for 57.42%, 50.87% and 50.00% respectively, indicating that the local spatial agglomeration characteristics of green development in Chinese counties are relatively stable, showing an obvious high-high agglomeration and low-low agglomeration polarization pattern. However, it can also be learned that over time, there is a small decline in the “neighborhood convergence” characteristic of green development in Chinese counties, which is in line with the global autocorrelation test’s findings. At the same time, the number of counties with low-low agglomeration declines from 305 to 194, indicating that more and more counties with low level of green development are gradually improving their own green development level, and voluntarily withdrawing from the low-low block. Specifically, counties belonging to high-high agglomeration are mainly distributed in the southern region and a few northeastern regions of China, including counties in Fujian Province, Yunnan Province, Guangdong Province and Jilin Province. On the one hand, most of these regions have pristine and abundant natural resources, which not only help to maintain soil and water and regulate the climate, but also create favorable conditions for Research and Development investment in clean energy and promote green energy transformation; On the other hand, certain southern counties have gotten ahead of the curve and advanced quickly in terms of economic development, and with their high level of Research and Development and emerging technology and equipment, they have obvious advantages in the fields of environmental monitoring, pollution control, energy saving and emission reduction. On the other hand, some counties located in the southern region started to build their economies early and developed quickly, and with their high level of Research and Development and emerging technology and equipment, they have obvious advantages in environmental monitoring, pollution control, energy saving and emission reduction. As a result, the green development of counties in these regions shows a high level. Meanwhile, based on China’s regional development strategy, some of the counties have positively influenced the green development level of the neighboring counties while improving their own green development level, and the radiation effect is obvious. Low-low agglomeration counties are mostly found in the center areas of the provinces of Shanxi, Hebei, Hubei, and Henan. These regions are geographically connected and integrated, rich in coal and petrochemical energy, and highly dependent on carbon-intensive industries for urban construction, which leads to a lower level of green development in the counties. The distribution pattern of counties in the L-H and H-L agglomerations is fragmented. Due to the differences in industrial construction, geographic environment and policy orientation, these regions did not have a diffusion effect on the neighboring counties while realizing their own green development construction. However, with the passage of time, the number of counties included in this characteristic region has declined significantly, and only 21.07% of the counties will be located in this region by 2020, proving once again that the homogeneity of green development in China’s counties is gradually increasing.

4 Analysis of driving factors for green development in Chinese counties

4.1 Choosing the motivating elements for green development

It is known that there are significant non-equilibrium and non-smooth spatiotemporal divergence characteristics of green development in Chinese counties. In this paper, using the spatiotemporal geographically weighted regression model and referring to the existing research results (Lina et al., 2023; Guo et al., 2020; Wu et al., 2024), we quantitatively analyze the reasons for the regional differences and spatial pattern evolution of green development in Chinese counties from multiple perspectives to provide experience and program guidance for counties lagging in the construction of green development (see Table 6).

1. Level of economic development. The relationship between the level of economic development and green development is multi-dimensional, not showing a single trend of change, there is “U” shaped Environmental Kuznets Curve (EKC) between the two, i.e., with the increase of regional per capita income, environmental pollution will first increase and then decrease, the level of green development will first decline and then rise, showing an inverted U-shaped curve (Asumadu et al., 2023). Therefore, this paper chooses the logarithmic GDP per capita (PGDP) as the specific measurement index of economic development level.

2. Industrial structure upgrading (IST). Industrial structure is the core element to measure the efficiency of a region’s economic development and the way of resource utilization. Modernizing the industrial structure can help the county’s conventional enterprises with high energy use and pollutant levels shift to a low-carbon, greener direction and encourage the development of green infrastructure. The tertiary industry’s added value as a percentage of GDP is used in this paper to describe it.

3. Green technology innovation (GTA). Green technology innovation is the most intuitive indicator that reflects the degree of regional green development. It not only reflects the region’s efforts and achievements in environmental governance and green construction, but also directly affects industrial structure, production efficiency and economic resilience. Vigorous Research and Development investment in green emerging technologies can reduce the environmental pressure in the region, attract more green investment, promote the formation of green industry clusters in the region, and provide new impetus and growth points for economic development (He et al., 2024). In this paper, it is expressed by adding one to the number of patent applications after taking the logarithm.

4. Quality of human resources (HUM). It is intuitively reflected in the degree of education level of workers. General secondary schools education constitutes a crucial foundational stage in the formation of human capital, and the scale of secondary school enrollment can, to some extent, reflect the educational attainment of the local population as well as the potential supply of future labor quality. This measure is particularly meaningful in less developed regions or counties dominated by labor intensive industries where basic educational attainment plays a significant role in shaping the local workforce (Ma et al., 2025). The percentage of the population at the end of the year that is enrolled in general secondary schools is used in this paper.

5. Strength of government support (GOV). As a key force in providing policy support and market demand, the government is also crucial to the growth of regional green development. This paper characterizes the general budget expenditure of local finance as a proportion of regional Gross Domestic Product.

Table 6

Table 6. Indicators of drivers of green development in Chinese counties.

4.2 Examination of the spatiotemporal, regionally weighted regression model’s output

This study employs the GTWR model to characterize the spatiotemporal heterogeneity of drivers for green development at the county level. Compared to MGWR, GTWR employs a single bandwidth across the three-dimensional spatiotemporal space, making it more suitable for identifying the overall spatiotemporal structure of multi-factor coupled systems like county-level green development. This approach avoids the overfitting risk caused by variable-specific bandwidths in MGWR models while significantly improving computational efficiency and result interpretability for large-sample panel data (N = 20,152). Compared to spatial panel models (SEM/SAR), this study focuses on describing spatiotemporal differentiation rather than inferring causal mechanisms. The GTWR model’s variable coefficient property perfectly aligns with the core objective of capturing structural evolution. In contrast, panel models’ implicit assumption of constant parameters leads to significant model specification errors. The comprehensive score of each Chinese county’s level of green development is used as the explained variable in this paper to construct the GTWR model. The descriptive statistics of the regression coefficients of the standardized drivers are shown in Table 7. The results show that the variance inflation factor (VIF) of each explanatory variable is less than 10, there is no obvious multicollinearity among the drivers, which meets the requirements of the model. From the OLS and GTWR models to test the global regression results, R² was 0.41 and 0.63, and the regression coefficients of each driver were significant (P < 0.1) which indicated that the GTWR model with the integrated temporal and spatial dimensions was significantly superior to OLS, and that the GTWR model could be utilized to explore the driving factors of green development level in Chinese counties.

Table 7

Table 7. Descriptive statistics of regression coefficients of drivers in GTWR model.

The parameters of the driving elements of each Chinese county’s level of green development from 2000 to 2021 were estimated using an ordinary least squares estimation model and a spatiotemporal spatially weighted regression model, respectively. The optimal spatiotemporal bandwidth was determined by minimizing the corrected Akaike information criterion (AICc). Model diagnostics (Table 8) show that the AICc for the GTWR model is −4879.76, which is 681.46 lower than that of the OLS model (AICc = −4198.30). The goodness-of-fit R² increased from 0.41 to 0.63, the residual sum of squares (RSS) decreased from 10.77 to 6.85, and the prediction error was reduced by 36.4%. From the OLS and GTWR models to test the global regression results, the regression coefficients of each driver were significant (P < 0.1), which indicated that the GTWR model with the integrated temporal and spatial dimensions was significantly superior to OLS, and that the GTWR model could be utilized to explore the driving factors of green development level in Chinese counties.

Table 8

Table 8. GTWR and OLS model selection.

This study employs the Geographically and Temporally Weighted Regression (GTWR) model to characterize the spatiotemporal heterogeneity of influencing factors for green development at the county level. Unlike traditional Ordinary Least Squares (OLS), which assumes globally constant parameters, GTWR constructs a three-dimensional kernel function by incorporating spatiotemporal coordinates (longitude, latitude, time), allowing the regression coefficients β to vary dynamically with geographical location and time.

The spatial difference of different driving factors on the level of green development is obvious, and the degree of influence on each county is not the same. Under the GTWR framework, the fitting coefficients of each county form independent spatio-temporal sample points. Through color gradients and spatial superposition, statistical significance and geographical patterns can be presented simultaneously. The specific results are shown in Figure 6.

1. Spatial difference analysis of the role of economic development level. Economic development level had a positive effect on the green development of counties nationwide in the year of 2000 and 2021, with the positive impact concentrated on counties in central regions like Beijing-Tianjin-Hebei, Shanxi Province, and Shandong Province, showing a gradual trend of growth. Based on the regional resource endowment and industrial foundation, the counties in these regions have constructed differentiated development paths to realize the synergistic coupling of economic growth and ecological protection. Take Shanxi Province as an example, Ruicheng County based on the advantage of barren slopes of Zhongtiao Mountain and sufficient sunshine, implements the concept of ecological priority, actively builds a diversified composite clean energy system, reduces the consumption of traditional coal energy, and helps realize the dual-carbon goal. Bali County has established a whole industrial chain including deep processing, scientific research and planting, intelligent storage and scientific and technological research and development through the sea buckthorn forests of 490,000 acres (Guo and Zhang, 2025). However, some regions exhibited negative effects in 2021. The increase of economic development level is instead unfavorable to the local green and low-carbon development. This fully indicates that there is significant spatial heterogeneity in county economic growth and eco-efficiency which coincides with the findings of the China Green Economy Development Report 2018. It is mainly concentrated in counties in South China and regions such as Yunnan, Guizhou and Sichuan, including Rongshui Miao Autonomous County in Liuzhou City; Yangshuo County and Pingle County in Guilin City, and Wufeng Tujia Autonomous County, Yidu City, and Public Security County in Hubei Province; and Kaiyang County, Fengdu County, and Fumin County. These counties are rich in natural resources and have a favorable ecological environment. However, the reason for this result may be that in the process of development, too much attention is paid to the speed of economic growth and the realization of short-term economic benefits, and there are problems of irrational utilization and over-exploitation, ignoring the negative impacts of project construction on the regional ecology, resulting in the wastage of resources and environmental pollution, thus restricting green development and economic sustainability. In addition, there are also some resource-poor counties, these late-developing areas face difficulties in attracting investment and expanding industries, and the level of economic development is low. At the same time, in order to realize ecological environmental protection, it is necessary to invest more funds in ecological construction and restoration, which results in financial constraints in the counties, and further leads to difficulties in landing environmental projects such as environmental governance, ecological restoration and green infrastructure construction, as well as difficulties in solving the problem of pollution, thus resulting in the formation of Ecological governance input and financial sustainability of the double pressure.

2. Spatial difference analysis of the role of industrial structure upgrading. There are both advantages and disadvantages to improving an industrial structure for a county’s green growth. Overall, it displays a tendency of diminishing ladder distribution from east to west, and the regression coefficient’s sign shifts from positive to negative. Interannual variations cause the detrimental effect to move from the periphery to the core. The detrimental impact on the region’s green development is especially evident in the Northeast and East China as a result of the ongoing modernization of the industrial structure. 2021, the positive high value area is mainly located in the South Xinjiang region and individual counties and cities in the middle of the Ili River Valley, such as: Luntai and Heshuo counties in Bayin’guoleng Mongol Autonomous Prefecture of the South Xinjiang region; Wensu and Aksu counties in Aksu region, and also in Yining and Huocheng counties, etc. The positive high value area is mainly located in the south of the border and the central Yili River Valley. The southern border region has many deserts and Gobi, and the distribution of oases is more concentrated. These counties mainly rely on natural resources such as land, water and grassland, and the industrial structure is relatively homogeneous, so it plays a more important role in encouraging the counties’ green growth. The Inner Mongolia Autonomous Region’s eastern counties and cities are home to the majority of the negative high-value areas. Although the state in the “14th Five-Year Plan” and other documents clearly optimize the industrial structure and the goal of green development, but industrial structure upgrading to the green development of the county is mostly negative impact, this is due to the vastness of China, the counties of the resource endowment, industrial center of gravity and the energy system differences are large, and some of the counties are still mostly high carbon energy and fossil energy and other traditional productivity as labor force, so the county has a more significant role in promoting green development. Fossil energy and other traditional productive forces as the object of labor, exacerbating environmental pollution and waste of resources, green low-carbon development path is difficult to synchronize the implementation of the actual more prominent. Nevertheless, there are still some counties based on their own resource advantages and industrial foundation while continuously achieving new results in green development and transformation. The industrial structure of Chinese counties is characterized by “steady decline in primary industry, slow decline in secondary industry, and continuous rise in tertiary industry”. The secondary industry is still an important pillar of the county economy, but the proportion of tertiary industry is gradually rising and becoming a new growth point. Such as Yichang City, Hubei Province, Wufeng Tujia Autonomous County in the mountainous areas of Southwest Hubei, with a superior geographical location and rich mineral resources, the county is surrounded by mountains on all sides, with a forest coverage rate of 81% and an average elevation of 1,100 m. The county has found coal, iron, barite, calcite and other minerals in a total of 137 (Dong and Liu, 2024). Excessive mining led to serious damage to the ecological environment, the original green mountains and green water into black hole black water. In order to reverse this predicament, Wufeng County learnt from the bitter experience, comprehensively shut down coal mining enterprises, digging deep into the resources to promote industrial integration, a large number of planting of economic forests and herbs, expanding the tea industry chain, the development of eco-circulation farming model, green development and transformation has achieved excellent results. There are some industrial chain is shorter, low value-added counties, in recent years are also actively expanding the industry in new areas, such as liling city in hunan province, the ceramic industry as the leading industry construction at the same time, vigorously develop the tourism industry and modern service industry, to enhance the green content of the industry, enhance the development of economic benefits.

3. Spatial difference analysis of the role of green technology innovation. From the regression coefficient, in 2000, green technology innovation had a positive driving effect on the green development of Chinese counties, mainly due to the positive guidance of national policies and the gradual promotion of green technology in counties. Sihong County in the lower reaches of the Huai River in western Jiangsu Province, Haiyan County and Deqing County in Zhejiang Province where the positive impact of technological innovation on green development is gradually increasing (Yuan et al., 2024), the impact of the increase in Research and Development investment in green technology on green development of counties in most regions has turned negative, and the degree of impact is gradually increasing from the south to the north and from the east to the west, indicating that the current increase in green technology Research and Development investment will instead reduce the level of county green development. Existing research shows that the increase in technological investment makes the impact of technological innovation on the efficiency of industrial green development present “U” type nonlinear relationship (Liu et al., 2018). Although the state’s attention to the green development of counties has been increasing, due to the lagging nature of technological innovation, and the increase in Research and Development investment makes the cost of enterprise production increase, it is difficult to take into account the economic and environmental benefits for economically underdeveloped counties, which results in the promotion of green technological innovation for the green development of counties is not obvious, and it may even result in the waste of resources and environmental pollution problems.

4. Spatial difference analysis of the role of human resource quality, the impact of human resource quality on county green development in 2000 and 2021 are both positive and negative, the negative effect is obvious and shows a decreasing trend from northeast to southwest, and positive high value area at the southwest region in 2021. From the perspective of factor allocation, improvements in human resource quality do not necessarily translate into enhanced green output or ecological performance. In regions dominated by heavy industries or resource based sectors, highly skilled labor may be disproportionately allocated to energy intensive and high-emission activities or to industries pursuing scale expansion. This may reinforce industrial expansion and resource exploitation in the short term, thereby increasing environmental pressure and resulting in human capital to pollution intensification effect rather than human capital to green efficiency enhancement effect. For example, counties such as Rongshui Miao Autonomous County, Ziyuan County, and Yongfu County exhibit pronounced resource-dependent characteristics, where local economic growth has long relied on mineral extraction, the intensive exploitation of tourism resources, or the expansion of forest land development. This resource driven growth pattern is inherently associated with ecological fragmentation, increased land disturbance, and declining ecosystem services. Consequently, environmental pressure intensifies during periods of economic expansion, while improvements in green development fail to keep pace, resulting in a short term negative relationship between economic development and green performance. Similarly, counties such as Yidu, Gongan and Binyang are experiencing rapid population mobility and accelerated urbanization. These processes generate substantial demand for infrastructure construction, accompanied by intensified building activities, transportation growth, and land conversion. In the short term, such dynamics lead to peaks in energy consumption and cumulative ecological disturbances, temporarily weakening green development performance and thereby producing negative coefficients for economic development. Moreover, current human capital indicators (such as average years of schooling) fail to adequately capture green skills and environmental awareness. Higher education levels may reinforce the traditional notion that development takes precedence over the environment. Additionally, environmental regulation costs within the green development index may suppress economic growth in the short term, leading to a statistically negative correlation between the two. Moreover, at the county level, highly educated talent often flows into non-green industries (such as finance and internet), creating a lag between human capital concentration and the spatial allocation of technology intensive clean industries. This disconnect between education and green transition further generates migration effects. Economically developed counties with high human capital face stricter regulations on their high-pollution industries, where increased short-term compliance costs lower their green composite scores. Conversely, counties with low human capital find it easier to adjust due to smaller industrial scales. From an endogenous perspective, reverse causality may exist. Pollution-intensive industrial clusters attract labor by raising wages, diluting average human capital. Moreover, regions with higher human capital can better afford environmental governance costs, proactively choosing a “governance first, development later” path. This creates complex dynamics of negative in the short term but positive in the long run. The other reason of the result may also be attributable to the human capital indicator is measured by the scale of secondary school enrollment, which makes it difficult to fully capture the skill structure and innovative capacity closely associated with green development. This indicator emphasizes the quantitative foundation of human capital while overlooking differences in education quality, skill types, and green technology oriented talent may lead to an underestimation of the role of human capital in green development.

5. Analysis of spatial differences in the role of government support strength. The impact of government support on all counties in 2000 was positive, the government’s regulatory role in economic development effectively compensated for the shortcomings of the market mechanism and promoted the development of green transformation. The overall distribution trend shows a gradual increase from northwest to southeast. As far as the regression coefficient is concerned, government support plays a key driving role in the green development of counties in most regions in 2021, and only counties such as Long’an County, Wanning City, Yao’an County, and Yuanmou County in the southwest region show a high negative sensitivity to the increase in government support. These counties are mostly economically underdeveloped and rich in natural resources but with more fragile ecological environments. It may be that local government support and funding is mainly used to strengthen the county’s economic construction, for the green transformation and development of the target is not strong, resulting in funds can not be effectively converted into green development momentum, the positive effect of promotion is difficult to be reflected. It may also be a mismatch between market mechanisms and corporate behavior. Despite the increase in government support, there are some enterprises that do not comply with the law, using non-compliant technology and equipment, increasing pollutant emissions, affecting the overall effect of green development. It is also possible that under the dual mechanism of the government and the market, high-quality human resources flow to economically developed regions, and the gap between traditional resource development paths and innovation strategies continues to widen, ultimately forming an “imbalance in the distribution of factors-innovation blocked-development stagnation”. The negative closed loop of “factor allocation imbalance-innovation obstruction-development stagnation” is eventually formed, which restricts the green development of the county.

Figure 6

Figure 6. Spatial distribution of regression coefficients of various drivers of green development in Chinese counties in 2000 and 2021. Note: Produced according to the standard map of the Ministry of Natural Resources Review No. GS(2019)1822, with no modification of the base map.

This study achieves multidimensional expansion and methodological breakthroughs in the spatiotemporal differentiation of green development at the county level compared to existing literature, specifically manifested in the following two dimensions. First, systematic transcendence of research scales. Han et al. (2024) focused on agricultural green development in the Yellow River Basin, Ning et al. (2024) were limited to agricultural ecological efficiency in Yunnan Province, and Liu et al. (2022) explored provincial-level green finance coupling coordination. Yet, none covered the complete spectrum of national administrative units. Existing county-level studies predominantly cover the period from 2010 to 2018, failing to encompass the complete policy cycle of ecological civilization institutional reforms before and after the 18th CPC National Congress. This paper constructs a 22-year panel dataset covering 916 counties, achieving China’s first nationwide county-level coverage. This breakthrough overcomes the regional and fragmented limitations of existing research, revealing the macro-level pattern of coexisting “high-value distribution zones in eastern coastal areas” and “low-value distribution zones in western inland areas”—a national-scale structural feature unattainable by existing regional studies. Second, methodological innovation in county-level adaptation. While some scholars employ geographic detector models to identify factor importance, these models fail to explain the mechanisms behind coefficient heterogeneity. This study applies the GTWR model to all counties nationwide. Its spatiotemporal coefficient characteristics precisely capture the spatial non-stationarity and temporal dynamics of each driving factor. Through robustness tests like the equal-weight method, it resolves estimation challenges posed by the high heterogeneity and complex spatial dependencies inherent in county-level data.

5 Conclusions and recommendations

This paper builds a county green development evaluation index system by choosing 19 specific indicators from the five dimensions of green innovation, green coordination, green efficiency, green openness, and green sharing. The entropy method is then used to measure the overall score of counties’ green development levels to reveal to determine the green and low-carbon development level of counties in Chinese. Agglomeration change and spatial and temporal differentiation characteristics are analyzed in detail, and the main conclusions are as follows: first, from the point of view of temporal change, the green development of China’s counties were at a low level during 2000–2009, and showed a trend of increasing year by year, but at the same time, there is a phenomenon of uneven regional development. Especially after 2015, the center position of kernel density and the overall right shift of the change interval, the level of green development in counties has increased by a large margin and the phenomenon of polarization of differences in green development in counties has intensified, and the level of green development in different counties has shown a trend of discrete distribution, and there are counties with higher levels of green development in counties such as those located in the eastern and southern regions of China. Second, from the perspective of spatial pattern. During the sample period, the green development level of most counties in China has been significantly improved, and the green development level of Chinese counties shows a spatial distribution pattern of “high in the east, low in the west, and moderate in the west”. In other words, counties with higher levels of green development are mostly clustered in the eastern and western regions, while the level of green development in the central region is relatively low. From the perspective of the quantitative trend of each level of development, the higher level of green development, the higher level of growth in the area, the lower level of growth in the area, indicating that the backward level of green development of the county has achieved rapid development, and the gap between the higher level of green development and the region is gradually narrowing. From the viewpoint of spatial distribution pattern, the green development level of Chinese counties shows the characteristic of “gathering first and expanding later”. From the perspective of center of gravity migration, the center of gravity of green development level in Chinese counties shows a spatial movement trend of “northwest-southeast”. Thirdly, from the perspective of spatial and temporal agglomeration, the Moran’s I index of green development of Chinese counties is greater than 0 from 2000 to 2021, with significant and positive global spatial autocorrelation, but the positive correlation continues to diminish with the passage of time, from 0.6309 in 2000 to 0.2779 in 2021, with a decrease of 57.06%, and the difference of green development among counties is more significant. Are more significant. The results of local autocorrelation analysis also show a certain spatial positive correlation, and the improvement of green development level of neighboring counties will have a positive effect on the green development level of the surrounding counties, and the green development level of the counties shows a spatial agglomeration trend, and the spillover effect is obvious. This shows that counties with high green development level have mature experience and technology, and the counties bordering them will be influenced by their policies, economy and technology in a subtle way, thus promoting the construction of green development level in the county. Fourth, the green and low-carbon development of Chinese counties is driven by the level of economic development, industrial structure upgrading, green technology innovation, quality of human resources and government support, and the same driving factor is spatially heterogeneous in different counties. The regression coefficient of the level of economic development is basically positive in the study period which has a strong driving effect on the green development of North China. The positive impact of industrial structure upgrading gradually diminishes from east to west, and the regression coefficient gradually becomes negative, transforming into a negative driving effect, which is insufficient for the driving ability of green development in Chinese counties. Xingwen County, Haiyan County and Deqing County in the southeast are more sensitive to the improvement and optimization of green technology innovation and government support. The regression coefficient of human resource quality is dominated by the negative value, which gradually increases from the central region to the surrounding area, which can be seen that the green development of China’s counties has already achieved stage-by-stage results.

Based on the above research conclusions, this paper puts forward the following countermeasure suggestions for further promoting the green and low-carbon development of Chinese counties:

First, deepen the concept of green low-carbon development and continuously promote the green construction of counties. County-level administrative units, as an important level in China’s administrative division, play a key role in promoting green development. The results of this paper show that the level of green development in China’s counties has shown a gradual upward trend over the past two decades, which proves that China has actively explored the green and low-carbon development mode and controlled greenhouse gas emissions in order to solve the environmental problems such as global warming and the frequency of extreme weather, and that the construction of regional green and low-carbon construction has made certain achievements. In addition, the natural environment and resource endowments of counties are superior to those of other administrative regions, and they naturally have advantages in practicing green low-carbon development mechanisms and policy systems. Therefore, we should actively promote the green transformation of the county, upgrade infrastructure and public service facilities, and improve the quality of the human environment.

Second, optimize the structure of the county’s industrial layout to achieve green synergistic transformation and development. Coordination is an inherent requirement for sustainable and healthy development, and to further enhance the integrity and stability of China’s county green development, it is necessary to recognize the current gap in the green development of counties in various regions. The conclusion of the study shows that the difference in the level of green development between the regions is a major constraint on the cause of China’s green and low-carbon development of the main factors for the further development of the cause. Therefore, the government should take into account local conditions, implement the concept of coordinated and shared development, and promote the concentration of relevant industries in areas with superior conditions by integrating the energy resources, environmental capacity, and market space of each county and region. At the same time, to deepen the cross-county division of labor, the lower level of green development of the county should learn from the surrounding higher level of green development, lower carbon emissions of the county development results, combined with their own regional and industrial characteristics, research suitable for the green development of the county’s new model, to promote the synergistic development of green and low-carbon industries in different regions, to accelerate the construction of a reasonable, efficient and coordinated modernization of the green development of the county pattern.

Third, adhere to the innovation-driven development strategy and enhance green low-carbon technological innovation. The green development of counties in the eastern region has been very effective and at the forefront, precisely by means of the key driving force of technological innovation. Through the introduction of clean energy technologies and energy-saving and emission reduction technologies, the use of traditional high-emission and high-pollution fossil energy is controlled at the source, and the energy structure is optimized. In order to create new advantages in green development, the counties in North China, represented by the “Four Mountainous Provinces”, should realize the green transformation of traditional industries, promote the efficient allocation and utilization of resources, actively introduce clean energy technologies, and gradually change the energy consumption structure dominated by coal.

In summary, as an important spatial unit, county level green development constitutes a critical foundation for achieving high-quality growth and ecological sustainability, serving as a vital bridge between macro-level policy objectives and micro-level implementation. From a theoretical perspective, county-level green development is shaped by the coordinated interaction among multiple factors, including economic growth, resource use efficiency, and environmental regulation. At the same time, counties often face practical constraints such as limited resource endowments and relatively weak industrial bases, which makes internal coordination among development elements particularly important. Therefore, counties should promote differentiated pathways of industrial transformation and upgrading in accordance with their resource endowments and stages of development. Enhancing resource use efficiency requires strengthening land-use regulation, promoting intensive and efficient land utilization, and guiding fiscal and financial resources toward green projects. In addition, it is necessary to establish and improve green finance mechanisms by supporting environmentally friendly enterprises through green credit instruments and fiscal incentives, thereby effectively alleviating financing constraints. Moreover, county governments should increase investment in digital infrastructure, support enterprises in undertaking green technological innovation, and facilitate technology diffusion and knowledge spillovers through regional cooperation. Through these integrated policy measures, county-level green development can be advanced along a more coordinated, efficient, and sustainable trajectory.

Despite providing a systematic analysis of the spatiotemporal evolution of counties green development in China, this study is subject to several limitations. First, due to data availability and space constraints, the analysis does not fully elaborate on key driving factors such as economic development, industrial structure, and green innovation. Although economic development is found to significantly affect green development in some counties, it is treated as a composite explanatory variable without further disaggregation. Second, industrial structure, which plays an important intermediary role between economic growth and environmental outcomes, is not examined in terms of its upgrading and rationalization dimensions, limiting a more comprehensive understanding of its influence on green development. Third, owing to difficulties in obtaining county-level data, green innovation indicators such as green patents and innovation efficiency are not explicitly incorporated into the empirical analysis. Future research may address these limitations in several ways. Non-linear or stage-based models could be applied to explore the heterogeneous effects of economic development across different development stages. Incorporating indicators of industrial upgrading and rationalization would help clarify the mechanisms through which industrial structural adjustment affects green development. And the indicators employed in this study balance theoretical relevance with data feasibility, they are nonetheless subject to certain measurement limitations. Future research should further deepen and refine the indicator system and expand data availability. In addition, integrating multi-source data on green innovation, including green patents would allow for a more thorough assessment of the role of technological progress in promoting county-level green development.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

Author contributions

XT: Conceptualization, Funding acquisition, Supervision, Writing – review and editing. RW: Data curation, Methodology, Software, Writing – original draft.

Funding

The author(s) declared that financial support was received for this work and/or its publication. XT acknowledges the grant of the Hebei Social Science Fund Project (Grant Number: HB24YJ013).

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Asumadu, S. S., Asantewaa, P. O., and John, T. (2023). Comprehensive green growth indicators across countries and territories. Sci. Data 10 (1), 413. doi:10.1038/s41597-023-02319-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Chen, M. Y., and Deng, Y. F. (2025). The impact effect of government public data of opening on regional entrepreneurial activity. J. Statistics Information. 40 (12), 118–128.

Google Scholar

Chen, H. F., Zhang, W. J., Du, R. Y., and Niu, D. X. (2025). Measurement of new quality productivity levels in Chinese cities, sources of differences and formation mechanisms. Statistics and Decis. (23), 29–35.

Google Scholar

Dong, Q. S., and Liu, J. (2024). Green mountains and land produce gold. Econ. Dly. (11).

Google Scholar

Fan, G., Wang, X. L., Zhang, L. W., and Zhu, H. (2003). Marketization index for China's provinces. Econ. Res. J. (3), 9–18+89.

Google Scholar

Fan, G., Wang, X. L., and Zhu, H. P. (2014). China marketization index. Economic Beijng: Science press.

Google Scholar

Gong, Y. X., Liu, X., Qi, Y., and Du, Y. (2025). Research on the resilience of urban logistics supply chain under the new development paradigm. Logist. Supply Chain Econ. 47 (11), 133–138.

Google Scholar

Griliches, Z. (1990). Patent statistics as economic indicators: a survey. J. Econ. Literature 28 (4), 1661–1707.

Google Scholar

Guo, J., and Yong, Z. T. (2023). Climate policy uncertainty and corporate green innovation-measured based on news media text analysis method. Finance Econ. (09), 75–86.

Google Scholar

Guo, F., and Zhang, J. (2025). “Deepening the transformation and development of resource-based economy in Shanxi: thickly planting the ecological background of,” in “North China Water Tower” to build the beauty of spatial layout, ecological harmony and green development. North China Natural Resources. (02), 2+1.

Google Scholar

Guo, Y. H., Mei, L., and Tong, L. J. (2020). Influence mechanism of industrial agglomeration on the efficiency of green development--The case of restricted development zone in Jilin Province. Geoscience 40 (09), 1484–1492.

Google Scholar

Guo, F. Y., Gao, S. Q., Tong, L. J., and Ren, J. M. (2022). Spatio-temporal evolution characteristics and influencing factors of green development efficiency in the Yellow River Basin. Geogr. Res. 41 (01), 167–180.

Google Scholar

Guo, D. M., Chen, B. K., and Wu, N. (2023). Research on the income and welfare effects of urban-rural integration: based on the perspective of factor allocation. Manag. World (11), 22–37.

Google Scholar

Han, X., Wang, Y., Liu, W. T., Yu, W. L., and Xia, X. L. (2024). Characteristics of spatial and temporal differentiation and driving factors of agricultural green development in the Yellow River Basin from the perspective of counties. Chin. J. Ecol. Agric. 32 (07), 1251–1263.

Google Scholar

He, X. Y., Li, G. P., and Liang, W. (2024). Research on the effect of green technology innovation on economic resilience in Beijing-Tianjin-Hebei. J. Beijing Jiaot. Univ. Sci. Ed. 23 (03), 76–88.

Google Scholar

Jiang, F. Z., and Liu, Z. T. (2024). Mechanisms and spatial effects of green technology innovation in curbing air pollution--an analysis based on spatial measurement and social network. Ecol. Econ. 40 (10), 212–221+229.

Google Scholar

Jing, J. X., and Gao, J. (2021). Contemporary implications of ecological modernization theory. Taishan Acad. J. (6), 42–46.

Google Scholar

King, R. G., and Levine, R. (1993). Finance and growth: schumpeter might be right. Q. J. Econ. 108 (3), 717–737. doi:10.2307/2118406

CrossRef Full Text | Google Scholar

Li, Y. Y., and Du, W. X. (2023). Analysis of spatial and temporal characteristics and heterogeneity of influencing factors of pollution reduction and carbon reduction in Beijing-Tianjin-Hebei urban agglomeration. J. Environ. Eng. Technol. 13 (06), 2006–2015.

Google Scholar

Li, H. L., and Li, C. Y. (2025). Research on the theoretical logic, influencing factors and practical paths of trade facilitation cooperation among countries along the Belt and Road. Northeast Asia Econ. Res. (12), 5–16.

Google Scholar

Li, X. H., and Wang, J. W. (2023). Regional disparities and influencing factors of green development of urban-rural construction in China under the goal of common wealth. J. Nat. Resour. 38 (02), 419–441. doi:10.31497/zrzyxb.20230210

CrossRef Full Text | Google Scholar

Li, J. L., Ran, G. H., and Ye, C. H. (2023). Market integration, industrial restructuring and green development efficiency—An empirical study based on the Yangtze River Delta Urban Agglomeration. Bus. Res. (03), 22–29+81.

Google Scholar

Liang, L. W., Xian, L., and Chen, M. X. (2022). Evolution of China’s regional population and economic centers of gravity since reform and opening up and their influencing factors. Econ. Geogr. 42 (02), 93–103.

Google Scholar

Liao, Y. L., and Jiao, X. (2005). Exploring regional sustainable development theory. Mark. Mod. (24).

Google Scholar

Lina, L., Heju, H., and Xiumei, T. (2023). Evaluation and driving factors analysis of agricultural green development level in Beijing from 1978 to 2022. Ecol. Evol. 11, 1233388. doi:10.3389/fevo.2023.1233388

CrossRef Full Text | Google Scholar

Liu, G. L., Ran, Q. Y., and Su, X. F. (2023). Heterogeneous local government competition, green technology innovation and upgrading industrial structure. Sci. Technol. Manag. Res. (1), 215–222.

Google Scholar

Liu, Y. Q., Quan, Q., Zhu, J. L., and Wang, F. (2018). Green technology innovation, industrial agglomeration and Eco-efficiency—The case of urban agglomeration in the Yangtze River Economic Belt. Yangtze River Basin Resour. Environ. 27 (11), 2395–2406.

Google Scholar

Liu, Q. L., Na, J. W., Li, Z. N., Liu, Y. X., and Xing, T. Y. (2022). Spatiotemporal variation in coupled coordination development of green finance and industrial structure in “Belt and Road” provinces. Ecol. Econ. 38 (10), 62–69.

Google Scholar

Ma, J., Zhang, Q., and Ha, M. F. (2025). The coupling coordination and spatial network collaboration characteristics of the “Production-living-ecological”functions of the urban agglomeration along the Yellow River in Ningxia. Sci. Technol. Manag. Land Resour. 42 (6), 59–72.

Google Scholar

Min, H., Zhenghui, L., and Benjamin, D. (2021). Do the green credit guidelines affect corporate green technology innovation? Empirical research from China. Int. J. Environ. Res. Public Health 18 (4), 1682. doi:10.3390/ijerph18041682

PubMed Abstract | CrossRef Full Text | Google Scholar

National Academy of Development and Strategy, RUC Capital Development and Governance institute (2018). Green road: China's green economic development report. Beijing: National Academy of Development and Strategy, RUC.

Google Scholar

Nie, Y. Y., and Yao, Q. Y. (2023). Productive service industry agglomeration and green development performance in Yangtze River Delta based on the perspective of “Local Neighborhood”. Res. World (06), 26–38.

Google Scholar

Ning, F. J., Lu, R. L. M., and Du, F. C. (2024). Spatiotemporal variation and influencing factors of agricultural ecological efficiency at the county level in Yunnan province. Chin. J. Agric. Resour. Regional Plan. 45 (02), 49–61.

Google Scholar

Ronaldo, R., and Suryanto, T. (2022). Green finance and sustainability development goals in Indonesian Fund village. Resour. Policy 78, 102839. doi:10.1016/j.resourpol.2022.102839

CrossRef Full Text | Google Scholar

Saeed, U. B., Adeel, T., Muhammad, F., Raza, A., and Iqbal, M. K. (2022). Green bonds for sustainable development: review of literature on development and impact of green bonds. Technol. Forecast. and Soc. Change 175, 121378. doi:10.1016/j.techfore.2021.121378

CrossRef Full Text | Google Scholar

Song, B. L., Bai, S. J., and Guo, Y. (2018). An empirical analysis of the relationship between economic growth, energy consumption, and industrial structure upgrading. Statistics and Decis. 20 (512), 142–144.

Google Scholar

Tang, Q. (2025). The impact of E-commerce to rural areas on county-level tourism economic development: evidence from a Quasi-natural experiment based on the national E-commerce demonstration counties program. J. Commer. Econ. (22), 177–180.

Google Scholar

The State Council the People’s Republic of China (2024). Opinions of the central committee of the communist party of China and the state council on accelerating comprehensive green transformation of economic and social development. Available online at: https://www.gov.cn/index.htm.

Google Scholar

Tong, X. (2020). The spatiotemporal evolution pattern and influential factor of regional carbon emissions convergence in China. Adv. Meteorology (3).

Google Scholar

Tong, X., Li, X. S., Tong, L., and Chen, K. (2019). Research of spatial and temporal evolution mechanism and countermeasures of Haze spatial pattern in China: visual field based on dynamic evolution and spatial agglomeration. Adv. Meteorology 2019 (3), 1–16. doi:10.1155/2019/7232481

CrossRef Full Text | Google Scholar

Wang, S. H., Zhang, W., He, M. X., and Li, Y. (2020). A literature review on supply-side reform-driven green development. J. Commer. Econ. (14), 186–189.

Google Scholar

Weng, Y., Yang, Y., Wen, Y. B., and Du, L. (2024). Evaluation and power mechanism of green transition development in mountainous counties of Zhejiang. J. Earth Sci. Environ. 46 (04), 470–485.

Google Scholar

Wu, H. Q., and Liu, Z. G. (2023). Influencing factors of ecosystem carbon storage in Jilin province based on spatial Durbin model. China Environ. Sci. (11).

Google Scholar

Wu, S., Fan, M., Wu, L., Liu, Z., and Xiang, Y. (2024). Path to green development: how do ESG ratings affect green total factor productivity? Sustainability 16 (23), 10653. doi:10.3390/su162310653

CrossRef Full Text | Google Scholar

Xie, L. Y., Yang, P., Yang, G. B., Xiang, J., Chen, P. F., Wang, R. R., et al. (2024). Measurement of spatial carbon sink conflict and characterization of spatial pattern in the national territory under multi-scale perspective-taking the Yangtze River Economic Belt as an example. Yangtze River Basin Resour. Environ. 33 (11), 2451–2460.

Google Scholar

Xiong, W., and Zhou, P. Y. (2024). Evaluation of agricultural green development level and analysis of spatial and temporal differences in the Yangtze River Basin. Ecol. Econ. 40 (07), 119–128.

Google Scholar

Xiong, L., Zhang, Q., Yan, Y. L., Jiang, S., and Su, L. Y. (2024). Spatio-temporal characteristics and driving factors of the coupled coordination between the level of numerical and intellectual development and carbon productivity in city clusters in the middle reaches of the Yangtze River. Econ. Geogr. 44 (10), 46–54.

Google Scholar

Xu, A. J., and Li, J. W. (2025). Research on R&D input performance of designated size industrial enterprises in 12 Western provinces. Gansu Sci. Technol. Inf. 54 (9), 11–18.

Google Scholar

Xu, K. N., Chen, F. L., and Liu, X. Y. (2015). The welfare cost of fiscal policy in China: a quantitative analysis in as two-sector structural change model. Econ. Res. J. (9), 17–29+57.

Google Scholar

Yang, Q., Xu, Q., and Chen, X. Y. (2022). Regional disparity and convergence test of green productivity of total factor water resources in China. Finance Trade Res. 33 (05), 15–30.

Google Scholar

Yang, Q. L., Chen, H. Y., and Wen, Q. (2024). The coupling coordination degree and enhancement path of digital economy and green development in municipal areas of the upper Yellow River region. Econ. Geogr. 44 (05), 22–32.

Google Scholar

Yuan, Q., Zhang, G., Wang, R., Ma, X., and Niu, J. (2024). Does technological innovation in National sustainable development agenda innovation demonstration zones promote green development? the case from Chengde City, China. PloS One 19 (5), e0300315. doi:10.1371/journal.pone.0300315

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhai, K. Z., and Hou, S. J. (2020). The green framework, implication and promotion program of China’s urban-rural integration and high-quality development during the “14th Five-Year Plan” period. Reform (11), 53–68.

Google Scholar

Zhang, K., and Xiong, Z. Y. (2024). Legal system improvement, cross-regional cooperation and green development in interprovincial border Areas-A Quasi-natural experiment from the implementation of the tourism law. Res. Quantitative Econ. Tech. Econ. 41 (12), 47–67.

Google Scholar

Zhang, R. B., and Zhong, C. B. (2022). Smart city pilot, nearby pollution transfer and green and low carbon development—New evidence from Chinese counties. China Popul. Resour. Environ. 32 (04), 91–104.

Google Scholar

Zhang, L. Z., He, Q., Qi, Y. B., and Deng, X. (2024). Regional differences and dynamic evolution of agricultural green development in Tuojiang river basin counties. China Agric. Resour. Zoning 45 (07), 54–65.

Google Scholar

Zheng, S. L., Zhou, L. A., and He, W. D. (2014). Telecommunications infrastructure and Chinese economic growth. Econ. Res. J. (5), 77–90.

Google Scholar

Zhirong, L., Kaiheng, Z., and Qikang, Z. (2022). Comprehensive evaluation and spatial-temporal pattern of green development in Hunan province, China. Sustainability 14 (11), 6819. doi:10.3390/su14116819

CrossRef Full Text | Google Scholar

Zhong, S. C., and Shao, J. H. (2022). Study on distribution dynamics, spatial differences and convergence of innovation development in the Yellow River Basin. Res. Quantitative Tech. Econ. 39 (05), 25–46.

Google Scholar

Zhong, J. W., Li, Y. X., and Zhao, Z. Y. (2024). Impact of China’s green tax system on regional economic “carbon lock-in”. China Population-Resources Environ. 34 (08), 13–21.

Google Scholar