EN中文

Journal of Geographical Sciences >

2024 , Vol. 34 >Issue 2: 289 - 308

DOI: https://doi.org/10.1007/s11442-024-2205-3

Research Articles

Spatial pattern evolution and driving factors of urban green technology innovation in China

  • LI Ying , 1 ,
  • FANG Yuanping , 2, * ,
  • MENG Qinggang 3
Expand
  • 1. School of Economics and Management, South China Normal University, Guangzhou 510006, China
  • 2. School of Geography, South China Normal University, Guangzhou 510006, China
  • 3. School of Statistics and Mathematics, Shandong University of Finance and Economics, Jinan 250014, China
*Fang Yuanping (1974-), PhD and Professor, specialized in economic geography and innovative geography. E-mail:

Li Ying (1993-), PhD Candidate, specialized in economic geography and portfolio selection. E-mail:

Received date: 2023-05-31

  Accepted date: 2023-11-17

  Online published: 2024-02-06

Supported by

National Natural Science Foundation of China(42171172)

Natural Science Foundation of Guangdong Province(2021A1515012248)

Major Program of the National Social Science Fund of China(21ZDA011)

Fold

Abstract

This study uses green patent data from 264 cities in China between 2006 and 2020 to examine the evolution of spatial patterns in urban green technology innovation (GTI) across the country and identify the underlying driving factors. Moran’s I index, Getis-Ord Gi* index, standard deviation ellipse, and geographical detector were used for the analysis. The findings indicate an increase in the overall level of GTI within Chinese cities. Provincial capitals, cities along the eastern coast, and planned cities emerge as the prominent “highlands” of GTI, whereas the “lowlands” of GTI predominantly lie in the western and northeastern regions, forming the spatial pattern of “hot in the east and center of the country, cold in the northwest and the northeast.” The distribution center of gravity of GTI is toward the southwest of China. The distribution pattern is in the “northeast-southwest” direction, which is characterized by “diffusion,” followed by “agglomeration.” Differences in economic development have the highest determining power on the spatial differentiation of GTI in Chinese cities, whereas differences in environmental regulation and industrial structure have the lowest degree of relative influence. The interaction between any two factors contributes to an amplified explanatory power in understanding the differences in GTI.

Key words： green technology innovation; spatial pattern; driving factors; geographical detector; Chinese city

Cite this article

LI Ying , FANG Yuanping , MENG Qinggang . Spatial pattern evolution and driving factors of urban green technology innovation in China[J]. Journal of Geographical Sciences, 2024 , 34(2) : 289 -308 . DOI: 10.1007/s11442-024-2205-3

1 Introduction

Modernization efforts in China require the harmonious coexistence between humanity and nature (Xi, 2022). Under the traditional crude development model, technological innovation aims to increase economic output and efficiency; however, economic growth is accompanied by high energy consumption and pollution (Wang et al., 2021; Chen et al., 2022), eventually leading to a series of problems such as resource shortage and ecological damage (Yang et al., 2013). The sustainability of economic development is confronted with serious challenges. The key to achieving a synergy between economic growth and environmental protection in the new development stage lies in embracing the concept of “greening” innovation. Green technology innovation (GTI) not only reduces the consumption of natural resources in economic activities but also acts as a significant instrument for environmental governance (Acemoglu et al., 2012; Fei et al., 2016; Guo et al., 2021; Wang and Yang, 2021). In October 2020, the Proposal of the Central Committee of the Communist Party of China on Formulating the 14th Five-Year Plan for National Economic and Social Development and the Visionary Goals for 2035 (the Proposal) explicitly emphasized the importance of supporting GTI. Accordingly, the questions posed are as follows: What is the current situation of GTI in Chinese cities? Are there any spatial characteristics and changing patterns? What are the determinants of spatial differentiation in GTI? Addressing these concerns can provide a comprehensive understanding of the spatial and temporal evolution patterns of GTI in Chinese cities and its determinants. This analysis holds theoretical value and practical significance in fostering the coordinated development of GTI in cities and Chinese-style modernization.
The concept of GTI was proposed by Fussler and James (1996). Subsequently, numerous scholars have explored its connotation from several dimensions (Rennings, 2000; Kemp, 2010; Driessen et al., 2013). Although the interpretations and descriptions of the concept of GTI given by scholars have been inconsistent, a consensus has been reached on the definition of GTI as technological innovations that positively affect the ecological environment. In recent years, the increasing availability of city-scale data has facilitated quantitative research on urban GTI. In the field of measuring GTI in cities, scholarly literature has explored two main approaches. One approach involves assessing the efficiency of GTI by using data envelopment analysis (DEA) methods (Dong et al., 2021; Sun and Shen, 2021; Chen et al., 2022; Dong et al., 2022; Yu et al., 2022; Peng et al., 2023; Wang et al., 2023). Another approach involves constructing GTI indices using comprehensive evaluation methods (Sun and Zhang, 2022; Tan et al., 2022). Existing literature primarily focuses on three main aspects within the realm of GTI research: convergence (Sun and Zhang, 2022), spatial distribution, and influencing factors (Dong et al., 2021; Sun and Shen, 2021; Chen et al., 2022; Dong et al., 2022; Tan et al., 2022; Yu et al., 2022; Peng et al., 2023; Wang et al., 2023). These studies encompass various geographical scopes, including national-level analyses (Dong et al., 2022; Sun and Zhang, 2022; Peng et al., 2023; Wang et al., 2023), examinations of significant regions (Dong et al., 2021; Sun and Shen, 2021; Yu et al., 2022), and investigations of urban agglomerations (Chen et al., 2022; Tan et al., 2022). In addition, some scholars have argued that the DEA-based research paradigm has focused only on the efficiency of resource use in the process of GTI. This approach may not fully capture the conceptual connotation of GTI. Moreover, the results of the comprehensive evaluation method heavily rely on the construction of the index system and the selection of the evaluation method. The complexity and variation in methodologies make it challenging to reach a unified conclusion when using this approach (Wang and Du, 2021).
Internationally, the OECD, on the other hand, uses green technology patents to measure GTI (Hascic et al., 2012; Hascic et al., 2015; Hascic and Migotto, 2015). Furthermore, some scholars in China have conducted studies based on green patent data. For instance, Dong et al. (2023) and Wang and Du (2021) examined the spatial correlation network of GTI in Chinese cities. Shao et al. (2022) conducted an analysis of the evolution of GTI output networks in three regions: the Yangtze River Delta, middle reaches of the Yangtze River, and Chengdu-Chongqing city cluster. Shang et al. (2021) explored the structural characteristics of innovation cooperation networks in different green technology areas within the Yangtze River Delta region. Wang and Du (2021) analyzed the distribution dynamics and spatial disparity of GTI in Chinese cities. Duan and Du (2022) subdivided green technologies into six categories and analyzed the subject’s characteristics, spatial distribution, and influencing factors for different types of GTI in China. Wang et al. (2020) employed the Shapley decomposition method to decompose the contribution of each factor to the variations observed in GTI in cities.
There is a scope for further exploration in the field of GTI research. Currently, most studies have focused on measuring GTI by constructing indices or assessing efficiency. While some quantitative analyses have utilized green patent data, a noticeable gap exists in the literature with regard to examining the spatial evolution of GTI in cities. Additionally, existing literature has used kernel density estimation, Dagum Gini coefficient, and social network analysis to reveal the distribution dynamics, spatial differences, and spatial association networks of GTI in Chinese cities. However, few studies have used the standard deviation ellipse technique to examine the spatial distribution direction and the migration path of the center of gravity of GTI. Furthermore, most of previous studies have predominantly relied on panel data regression models and spatial econometric models to investigate the factors affecting the spatial-temporal evolution of GTI. Traditional regression analysis methods are effective in testing the correlation between variables; however, they may not adequately capture the impact of complex interactions among multiple determinants on the spatial heterogeneity of GTI. In case of spatially stratified heterogeneity in the study sample, the results obtained from traditional econometric models may be biased.
Accordingly, this study adopts statistical methods such as Moran’s I index, Getis-Ord Gi* index, standard deviation ellipse, and geographic detector. Based on green patent data from 264 prefecture-level and above cities in China from 2006 to 2020, this study explores the spatiotemporal evolutionary characteristics and factors influencing urban GTI, with a view to provide theoretical support and decision-making basis for the formulation of synergistic development policies on urban green innovation and the realization of Chinese-style modernization in which human beings and nature coexist in harmony.

2 Methods and data

2.1 Research method

2.1.1 Exploratory spatial data analysis

In this study, Moran’s I index is used to assess a spatial correlation in China’s urban GTI. The calculation formula of Moran’s I index is as follows:
$\operatorname{Moran}^{\prime} s I=\frac{\sum_{i=1}^{n} \sum_{j=1}^{n} w_{i j}\left(x_{i}-x\right)\left(x_{j}-x\right)}{S^{2} \sum_{i=1}^{n} \sum_{j=1}^{n} w_{i j}}$
where xi and xj denote green innovation output of city i and j, respectively. $S^{2}=$ $\frac{1}{n} \sum_{i=1}^{n}\left(x_{i}-x\right)^{2}$ and $x=\frac{1}{n} \sum_{i=1}^{n} x_{i}$, wij denotes the spatial weight matrix; when i ≠ j, wij=1/d2 ij, and when i = j, wij = 0. dij denotes the geographical distance between the two cities. Moran’s I index takes values in the range [–1, 1]. Values greater than 0 indicate positive spatial autocorrelation, values less than 0 indicate negative spatial autocorrelation, and values close to 0 indicate no spatial autocorrelation.
Furthermore, the Getis-Ord Gi* statistic is used to identify the high-value area (hot spot) and low-value area (cold spot) of the spatial distribution of GTI. The calculation formula is as follows:
$G_{i}^{*}=\frac{\sum_{i=1}^{n} \sum_{j=1}^{n} w_{i j} x_{i} x_{j}}{\sum_{i=1}^{n} \sum_{j \neq i}^{n} x_{i} x_{j}}$
where the meanings of symbols are the same as above. Gi* statistics are standardized to obtain the following expression:
$Z\left(G_{i}^{*}\right)=\frac{G_{i}^{*}-E\left(G_{i}^{*}\right)}{\sqrt{\operatorname{Var}\left(G_{i}^{*}\right)}}$
where Z(Gi*) denotes the standardized value of the Gi* statistic. E(Gi*) and Var(Gi*) denote the expectation and variance of the Gi* statistic, respectively. A Z(Gi*) value that is significantly greater than 0 indicates a high-value clustering of GTI (hot spot area), whereas a Z(Gi*) value that is significantly less than 0 indicates a low-value clustering of GTI (cold spot area).

2.1.2 Standard deviation ellipse

This study uses the standard deviation ellipse analysis method to explore the spatial distribution pattern and locational characteristics of GTI in Chinese cities (Lefever, 1926). The main parameters are calculated as follows:
$\bar{X}_{w}=\sum_{i=1}^{n} w_{i} x_{i} / \sum_{i=1}^{n} w_{i} \quad \bar{Y}_{w}=\sum_{i=1}^{n} w_{i} y_{i} / \sum_{i=1}^{n} w_{i}$
$\sigma_{x}=\sqrt{\sum_{i=1}^{n}\left(w_{i} \Delta x_{i} \cos \theta-w_{i} \Delta y_{i} \sin \theta\right)^{2} / \sum_{i=1}^{n} w_{i}^{2}} \sigma_{y}=\sqrt{\sum_{i=1}^{n}\left(w_{i} \Delta x_{i} \sin \theta-w_{i} \Delta y_{i} \cos \theta\right)^{2} / \sum_{i=1}^{n} w_{i}^{2}}$
$\tan \theta=\left[\left(\sum_{i=1}^{n} w_{i}^{2} \Delta x_{i}^{2}-\sum_{i=1}^{n} w_{i}^{2} \Delta y_{i}^{2}\right)+\sqrt{\left(\sum_{i=1}^{n} w_{i}^{2} \Delta x_{i}^{2}-\sum_{i=1}^{n} w_{i}^{2} \Delta y_{i}^{2}\right)^{2}+4 \sum_{i=1}^{n} w_{i}^{2} \Delta x_{i}^{2} \Delta y_{i}^{2}}\right] / 2 \sum_{i=1}^{n} w_{i}^{2} \Delta x_{i} \Delta y_{i}$
$S=\pi \sigma_{x} \sigma_{y}$
where (xi, yi) denotes the geographical coordinates of city i. Furthermore, wi denotes the weight, and the weight is the GTI level of the city. ($\bar{X}_{w}$,$\bar{Y}_{W}$) denotes the weighted average center coordinates of Chinese cities, θ denotes the azimuth, and (Δxi, Δyi) denotes the difference between the geographical coordinates of city i and the weighted average center coordinates. σx and σy denote the standard deviation on the x-axis and y-axis, respectively, and S denotes the area of the ellipse.

2.1.3 Geographic detector

Factor detection and interaction detection are used to investigate the determinants of each detection factor and its interaction on the spatial differentiation of GTI in Chinese cities (Wang and Xu, 2017). The calculation formula for determining force q is as follows:
$q=1-\frac{1}{N \sigma^{2}} \sum_{h=1}^{L} N_{h} \sigma_{h}^{2}$
where h=1, 2, …, L denotes the stratification of the factor. N and Nh are the numbers of samples in the whole region and layer h, respectively. σ2 denotes the variance of the dependent variable Y in the whole region, σ2 h denotes the variance of the dependent variable Y in layer h, and q denotes the determining force of probe factor X on spatial differentiation of dependent variable Y; its value is [0, 1]. The probe factor should essentially be a type variable. Accordingly, the natural discontinuity method is used to convert continuous-type factor variables into type variables.
Furthermore, interactive detection is used to assess whether the explanatory power of interaction between the two factors X1 and X2 on the spatial divergence of the dependent variable Y is enhanced, diminished, or independent.

2.2 Sample data

This research focuses on China’s prefecture-level and above cities, covering the period from 2006 to 2020. However, cities with severe missing data and those that underwent administrative reorganization at the prefecture level during the sampling period were excluded from the study. For instance, Chaohu was excluded from the sample as it was reorganized from being a prefecture-level city to a county-level city in 2011. Counties under the jurisdiction of a prefecture-level city that underwent administrative changes, including name changes, were still included as the sample. Finally, a balanced panel data set consisting of 264 cities at the prefecture level and above was obtained. The sample included 35 municipalities directly under the central government, provincial capitals, planned cities and 229 general prefecture-level cities. Two years, 2006 and 2020, and three periods 2006-2010, 2011-2015, and 2016-2020 were selected as empirical research time nodes. The years 2006 and 2020 represent the start and end of the research period, respectively. The periods 2006-2010, 2011- 2015, and 2016-2020 correspond to the “11th Five-Year,” “12th Five-Year,” and “13th Five-Year” periods of China’s national economic and social development plan, respectively.
Drawing on the research conducted by Wang and Du (2021), the number of green invention patents has been recognized as a highly suitable indicator for the GTI capacity of urban areas. However, there is typically a lag of 1-3 years between the application and granting of patents (Hall and Harhoff, 2012), which implies that it may not fully reflect the current year’s level of green innovation in a city. In this study, the number of green invention patent applications (taking the natural logarithm) is used as a measure of the level of GTI in cities. Green patent information was obtained from the patent retrieval and analysis platform of the State Intellectual Property Office based on the “Green List of International Patent Classification” formulated by the World Intellectual Property Organization. By setting the patent type, International Patent Classification codes of green patents (including the seven subcategories of green patents and their respective detailed patent contents), the invention unit’s address, and green invention patent application data at the city level can be retrieved and aggregated by year and region.

3 Spatial pattern evolution of urban GTI in China

3.1 Spatial distribution changes of urban GTI

The sample data within the research period were first adjusted by 0.5% shrinkage both upwards and downwards to eliminate outliers. Then, using the equal spacing method, the cities in 2006 and 2020 were divided into four levels, low, medium-low, medium-high, and high, based on the level of GTI. Similarly, the cumulative amount of GTI in cities was processed separately for the periods 2006-2010, 2011-2015, and 2016-2020. As depicted in Figure 1, the spatial distribution map of GTI in Chinese cities was drawn using ArcGIS software for each year and time period.
Figure 1 Spatial distribution of green technology innovation in China
In 2006, the overall level of GTI in most cities was relatively low. The proportions of cities with low and medium-low GTI levels were 51.894% and 31.439%, respectively. These cities exhibited a contiguous spatial distribution. The cities with a medium-high GTI level mainly include provincial capitals such as Guangzhou, Chengdu, Xi’an, Zhengzhou, and Nanjing, as well as planned cities including Shenzhen, Dalian, and Qingdao. These cities are relatively few in number and have a scattered spatial distribution. Only two cities, Beijing and Shanghai, were classified as high-level innovation cities. Compared with 2006, the level of GTI in cities in 2020 differed significantly. The number of cities classified as having a low GTI level has greatly decreased as they transitioned to either medium-low GTI cities or medium-high GTI cities. Only five cities, namely Heihe, Hegang, Qitaihe, Liaoyuan, and Zhangjiajie, remained in the low-level category. On the other hand, the number of high-level cities increased from 2 to 38, which are primarily distributed in Jiangsu, Zhejiang, the Pearl River Delta region, and provincial capitals across different provinces.
By period, the spatial distribution of high-level and medium-high-level cities in the period 2006-2010 was similar to that in 2006. Furthermore, a small clustering phenomenon of medium-high-level cities centered around Wuxi emerges in the eastern region. Low-level cities are mainly distributed in Heilongjiang, Gansu, Inner Mongolia, Sichuan, Jiangxi, etc., whereas medium-low-level cities are mainly distributed in Hebei, Henan, Shandong, etc. From 2011 to 2015, the number of low-level cities significantly decreased, whereas the number of medium-high-level and high-level cities increased, and these cities were concentrated in the eastern coastal areas. From 2016 to 2020, the advantage of GTI in eastern cities was further highlighted, with Jiangsu and Zhejiang becoming the provinces with the highest number of high-level cities.
During the survey period, an overall improvement in the level of GTI in Chinese cities was observed. Provincial capitals, eastern coastal cities, and planned cities emerged as the “highlands” of GTI, while the “lowlands” were mainly concentrated in the western and northeastern regions. The aforementioned phenomenon may be ascribed to the fact that the economic development level in provincial capitals, eastern coastal cities, and planned cities is often high. These cities possess abundant R&D capital and have a siphoning effect on R&D personnel, attracting innovative talents from outside to relocate locally. The influx of foreign human capital brings diverse knowledge, culture, and skills, which contribute to the local output of GTI (Zhang, 2019). Conversely, the western and northeastern regions are economically underdeveloped and serve as major sources of outflow for innovation resources, especially in terms of R&D personnel (Liu and Wang, 2020). The outflow of R&D personnel, being a crucial input factor for innovation activities, has a disincentive effect on GTI. Thus, the output of GTI in the western and northeastern regions is relatively low.

3.2 Changes in agglomeration characteristics of urban GTI

3.2.1 Spatial autocorrelation

The Moran’s I index values for GTI in Chinese cities from 2006 to 2020 were calculated on the basis of a geographical distance weight matrix (Figure 2). The values were found to be positive and statistically significant at 1% level, suggesting that the GTI output in each city was not randomly distributed but was influenced by the GTI activities in neighboring areas. In other words, the spatial correlation between GTI activities in Chinese cities a significant and positive, consistent with the findings of Duan and Du (2022). This finding can be explained by the nature of green innovation activities, which rely on scarce innovation factors oriented toward seeking profits. In different innovation systems, innovation factors tend to be allocated to areas with higher marginal rates of return, and the cross-regional flow of innovation factors eventually results in a spatial correlation in urban GTI (Bai and Jiang, 2015). According to the changing trend of the Moran’s I index, the spatial correlation degree of GTI has increased, indicating that the spatial agglomeration of GTI in China’s cities may gradually strengthen.
Figure 2 Change trend of Moran’s I index of green technology innovation in China from 2006 to 2020

3.2.2 Analysis of hot and cold spots

High-value agglomerations (hot spots) and low-value agglomerations (cold spots) of GTI across the country could not be identified using Moran’s I index. Accordingly, the Getis-Ord Gi* index was applied to determine the spatial distribution of GTI agglomerations. By applying the Getis-Ord Gi* statistic, the spatial distribution pattern of cold-hot spots in urban GTI was determined and categorized into seven types based on the significance level (Figure 3).
Figure 3 Spatial distribution pattern of hot and cold spots of green technology innovation in China
In the early stage of the study, hot spots of GTI were primarily located in Liaoning, Hebei, Shandong, Jiangsu, and other provinces in the eastern region, as well as some cities in Anhui and Henan provinces in the surrounding central region. Local hot spots emerged in Ningde, Fuzhou, and Putian in Fujian province and Zhoushan in Zhejiang province. In 2006, the distribution of cold spots in GTI was relatively scattered, involving cities in some provincial-level regions such as Gansu, Ningxia, Shaanxi, Sichuan, Guizhou, Hunan, and Guangdong, as well as Heihe, Hegang, and Yichun in Heilongjiang province. In 2020, the range of hot spots expanded and shifted toward the south, exhibiting a clustered distribution in space. Cities located in Liaoning and Inner Mongolia, as well as Chengde, Qinhuangdao, Tangshan, Zhangjiakou, and Langfang in Hebei province, along with Beijing and Tianjin, disappeared from the hotspot area. At the same time, cities in Zhejiang, Fujian, and Jiangxi, as well as certain cities in Henan, Hubei, Hunan, and Guangdong, entered the hotspot area. This indicates an increase in GTI in the southern direction during this period. Compared with 2006, some cities in Shaanxi, Hubei, Hunan, Guangxi, and Guangdong exited from the cold spots in 2020. In Inner Mongolia, Bayannur, Ordos, and Baotou entered the cold spots, whereas in the northeastern region, the cold spots were mainly concentrated in Heilongjiang and Jilin provinces.
Looking at different time periods, from 2006 to 2010, the hots spots of GTI were mainly concentrated in eastern regions such as Shandong, Jiangsu, Zhejiang, and Fujian, as well as central regions including Hebei, Henan, Anhui, and Jiangxi. From 2011 to 2015 and 2016 to 2020, the hot spots expanded from the eastern coastal cities toward the inland cities in the southwest direction. From 2006 to 2010, the cold spots of GTI were primarily scattered across various provincial-level regions in the western region, such as Gansu, Ningxia, Shaanxi, Sichuan, Guizhou, Hunan, and Guangxi but were concentrated in Heilongjiang and Jilin provinces in the northeastern region. From 2011 to 2015 and 2016 to 2020, the range of cold spots decreased in the western region but remained relatively stable in the northeastern region.
Overall, Chinese cities exhibit spatial differentiation in GTI, with a “hot in the eastern and central regions, cold in the northwest and northeast” pattern. This may be related to the spatial agglomeration and distribution of economic activities (Wu et al., 2022). On the one hand, the geographical concentration of economic activities facilitates the free flow of technological factors, increasing the likelihood of knowledge spillover. Accordingly, different companies can learn from each other and share innovative resources, thereby enhancing their GTI capabilities. Spillover effects can also facilitate the diffusion of the impacts of R&D investments of certain companies to neighboring companies, accelerating the emergence of new technologies in the region (Gordon and McCann, 2005). On the other hand, surrounding conditions such as infrastructure can improve in areas with high economic agglomeration. Under the influence of economic radiation and spillover effects, the central areas can also moderately diffuse resource elements to the surrounding areas, promoting GTI in the peripheral regions (Martinus et al., 2020). Ultimately, the agglomeration and development of GTI activities may become consistent with the spatial agglomeration of economic activities, driven by positive externalities in the economic agglomeration spaces.

3.3 Distribution center of gravity and morphological evolution of urban GTI

This study subsequently determines the center of distribution and morphological evolution of GTI in Chinese cities by using the standard deviation ellipse technique (Figure 4). Table 1 presents the relevant parameters of the standard deviation ellipse.
Figure 4 Center of gravity - Standard deviation ellipse of green technology innovation in China
Table 1 Center of gravity - Standard deviation ellipse parameters of green technology innovation in China
Year 2006 2010 2015 2020
Center of gravity coordinates 115.486°E 115.059°E 114.877°E 114.853°E
33.446°N 32.967°N 32.815°N 32.783°N
Center of gravity city Zhoukou Zhumadian Zhumadian Zhumadian
Movement direction — Southwest Southwest Northeast
Moving distance (km) — 71.212 26.353 4.441
Movement speed (km/year) — 17.803 5.271 0.888
Short semi-axis (km) 674.534 694.942 689.058 687.773
Long semi-axis (km) 1096.13 1064.213 1082.258 1072.757
Azimuth (°C) 24.937 22.819 24.563 23.026
Elliptical area ratio 1 1 1.009 0.998
The center of distribution of GTI in Chinese cities shifted within the range of 114.853°E‒ 115.486°E and 32.783°N‒33.446°N within Henan province at different study points. The magnitude of the shift in the north-south direction was greater than that in the east-west direction, indicating a tendency for the center of gravity of GTI distribution to shift in the north-south direction. Additionally, the GTI level in southern cities surpassed that in northern cities. In terms of the movement of the distribution center, from 2006 to 2010, the center of gravity shifted southwest by 71.212 km, moving from Zhoukou to Zhumadian in Henan, with a migration rate of 17.803 km per year. This indicates a rapid improvement in the level of GTI in the southwest direction during that period. From 2010 to 2020, the center of gravity continued to move southwest, but at a gradually slower pace, being concentrated in Zhumadian city, Henan. Overall, from 2006 to 2020, the center of gravity of GTI in Chinese cities showed a trend of evolution toward the southwest direction. This could be attributed to the accelerated development of green finance and the continuous improvement of the financial support system for green technology research and development in regions such as Hunan and Guangxi, injecting vitality into GTI. In recent years, Guangxi has been actively promoting the construction of a green finance reform and innovation demonstration zone, expanding the scale of green credit and implementing special interest subsidy policies for green loans. This has provided tangible financing support to real enterprises, thereby increasing the enthusiasm for green technological research and development (Zhang et al., 2021; Tian et al., 2022).
From 2006 to 2020, the standard deviation ellipse of GTI in Chinese cities was located in the eastern and central regions, exhibiting a “northeast-southwest” spatial distribution pattern. With the improvement of green innovation efficiency, the output of GTI in the Beijing-Tianjin-Hebei region increased rapidly (Li and Ma, 2019). The ellipse rotated counterclockwise by 1.911° from the northeast-southwest direction to the north-south direction. The ellipse angle decreased from 24.937° in 2006 to 23.026° in 2020. Although there was a slight shift in the spatial distribution direction of urban GTI, the overall distribution pattern remained relatively stable. The range of the ellipse expanded from 2006 to 2015, with a 0.9% increase in the area. At the same time, the short semi-axis expanded from 674.534 km in 2006 to 689.058 km in 2015, while the long semi-axis narrowed from 1096.13 km in 2006 to 1082.258 km in 2015. This indicates a trend of diffusion in the spatial distribution of GTI in Chinese cities during this period. Although the northeast-southwest direction remained the main axis of spatial distribution for urban GTI, the development in the northwest-southeast direction was more prominent. From 2015 to 2020, the range of the ellipse showed a contraction trend, with the elliptical area ratio decreasing from 1.009 in 2015 to 0.998 in 2020. Additionally, the short and long semi-axis decreased from 689.058 km and 1082.258 km in 2015 to 687.773 km and 1072.757 km in 2020, respectively. This indicates that during this period, the growth of GTI within the ellipse surpassed that outside the ellipse, demonstrating a trend of agglomeration development.

4 Driving factors for urban GTI differences in China

The spatial distribution of GTI in Chinese cities is uneven. To investigate the causes of this spatial variation, the key determinants of spatial variation in GTI in Chinese cities and the impact of the interaction of each factor were examined using factor detection and interaction detection in the geodetector model.

4.1 Selection of driving factors

The spatiotemporal evolution of GTI in Chinese cities is a complex process influenced by multiple factors. Based on a review of literature, this study examines six key driving factors for spatial differentiation in urban GTI, namely environmental regulations (Er), economic development (Eco), financial development (Fin), industrial structure (Ind), fiscal technology expenditure (Fis), and government intervention (Gov) (Figure 5). (1) Environmental regulations (Er): The “Porter hypothesis” suggests that appropriate environmental regulations can stimulate GTI activities within companies. It posits that companies can improve their production efficiency while achieving environmental goals, leading to “innovation compensation” (Porter and Linde, 1995; Kneller and Manderson, 2012). However, the “compliance cost theory” argues that environmental regulations can increase the production costs and reduce R&D investment (Barbera and McConnell, 1990; Wagner, 2007). Some studies have also found a “U-shaped” relationship between environmental regulations and GTI (Zhang et al., 2019). Referring to the study by Zhang and Chen (2021), the frequency of environmental protection terms in government work reports is used as a proxy variable for environmental regulations. (2) Economic development (Eco): The level of economic development in a city is often indicative of the investment and accumulation of resources, including human and capital. A higher level of economic development corresponds to a greater ability to promote GTI, as measured by GDP per capita (Yu and Lyu, 2023). (3) Financial development (Fin): Financial development can increase the conversion rate of savings and investments in the economy. Furthermore, it can reduce the external financing costs of the R&D sector of companies and provide incentives for companies to increase their green R&D investments (Zhuang et al., 2020). This variable is measured using the number of loans from urban financial institutions as a share of GDP. (4) Industrial structure (Ind): Industrial structure influences regional technology choices. A higher proportion of clean industries within the industrial structure is generally more conducive to promoting GTI (Acemoglu et al., 2012; Jin and Li, 2013), as measured by the ratio of the output value of the tertiary industry to that of the secondary industry. (5) Fiscal technology expenditure (Fis): This variable reflects the government’s support for science and technology innovation. It can enhance the innovation output of enterprises by alleviating their budgetary constraints and improving their financial situation (Che et al., 2020). It is measured as the proportion of a city’s current year science expenditure to GDP. (6) Government intervention (Gov): Government intervention can, to a certain extent, correct distortions in the allocation of innovation resources due to market failure. According to the study by Dong and Wang (2021), it can guide the allocation of funds to green areas such as green talent training and green technology research and development. It is measured as the proportion of the government’s general budget expenditure to GDP. The corresponding data from 2006 to 2020 were obtained from the China Urban Statistical Yearbook and the statistical bulletins of various cities. The individual missing data were filled by interpolation.
Figure 5 Mechanisms of drivers of differences in green technology innovation in Chinese cities

4.2 Result analysis

4.2.1 Single-factor driving force

Table 2 reports the findings of the detection of spatial differentiation in urban GTI in Chinese cities, including the magnitude and significance levels of the determinants. Except a few indicators that were not statistically significant during specific periods, the majority of the driving factors passed the significance test at 1% level, indicating that these factors have a certain driving effect on urban GTI. By comparing the q-values of the driving factors for the years 2006 and 2020 and the periods 2006-2010, 2011-2015, and 2016-2020, it can be inferred that economic development, financial development, fiscal technology expenditure, and government intervention rank among the top four in terms of q-values. This suggests that these four variables are the core factors driving the spatial differentiation of GTI in Chinese cities.
Table 2 Determinants of China’s GTI differentiation detection factors
Year
Detection factors		 2006 2020 2006-2010 2011-2015 2016-2020
q Seq q Seq q Seq q Seq q Seq
Er 0.022 - 0.034** 5 0.032*** 5 0.017*** 6 0.017*** 5
(0.195) (0.057) (0.000) (0.000) (0.000)
Eco 0.413*** 1 0.470*** 1 0.361*** 1 0.301*** 2 0.426*** 1
(0.000) (0.000) (0.000) (0.000) (0.000)
Fin 0.289*** 2 0.231*** 4 0.301*** 2 0.316*** 1 0.141*** 4
(0.000) (0.000) (0.000) (0.000) (0.000)
Ind 0.003 - 0.016 - 0.024*** 6 0.091*** 5 0.011 -
(0.949) (0.720) (0.000) (0.000) (0.128)
Fis 0.094 - 0.270*** 3 0.228*** 3 0.179*** 4 0.259*** 3
(0.451) (0.000) (0.000) (0.000) (0.000)
Gov 0.134*** 3 0.386*** 2 0.199*** 4 0.228*** 3 0.304*** 2
(0.000) (0.000) (0.000) (0.000) (0.000)

Note: P-values are in parentheses, *** and ** indicate passing the significance level test of 0.01 and 0.05, respectively.

Analyzing the evolution of q-values for the core factors, the q-value for economic development is consistently ranked in the top two positions. This indicates that the differences in economic development are the primary driving factor for the spatial differentiation of GTI among cities nationwide. This is because economically developed cities are capable of attracting the agglomeration of innovative elements, while economically lagging cities face a scarcity of innovation resources, and their industrial development is primarily focused on heavy industries, resulting in insufficient driving force for GTI within their enterprises.
The q-value for financial development has decreased from 0.289 in 2006 to 0.231 in 2020. Moreover, it was 0.301 during the “11th Five-Year Plan” period and 0.141 during the “13th Five-Year Plan” period, indicating a decreasing trend. This indicates a weakened impact on GTI in Chinese cities. One possible reason for this is the multiple reductions in the required reserve ratio by the central bank, which has enhanced the credit resource allocation capacity of financial institutions and generally reduced the financing costs for technology research and development for real enterprises.
The q-value for fiscal technology expenditure did not pass the significance test in 2006 but became significantly positive in 2020, with a value of 0.270. Looking at different periods, the q-value for fiscal technology expenditure increased from 0.228 during the “11th Five-Year Plan” period to 0.259 during the “13th Five-Year Plan” period. This may be due to the continuous growth in China’s fiscal technology expenditure since the “11th Five-Year Plan” period, along with the widening regional differences in fiscal technology expenditure (Chen and Wang, 2021). As a result, the effect of fiscal technology expenditure on the spatial differentiation of GTI has strengthened.
The q-value for government intervention has increased overall, from 0.134 in 2006 to 0.386 in 2020. Moreover, it was 0.199 during the “11th Five-Year Plan” period and 0.304 during the “13th Five-Year Plan” period, indicating an increasing trend. This further indicates an enhanced impact on GTI in Chinese cities. This is because as the economy enters a stage of high-quality development, local governments begin pursuing innovation and green development, leading to the introduction of a series of new talent policies. These policies have prompted innovative talents to migrate to developed regions such as Guangdong, Beijing, Shenzhen, and Zhejiang, which has a positive impact on GTI (Wang et al., 2021; Gössling and Rutten, 2007).
The q-values for environmental regulation and industrial structure were relatively low during the study period, indicating that these two factors have a weaker driving effect force on the spatial differentiation of GTI in Chinese cities. This may be because different companies adopt different measures in response to local environmental regulatory policies. Heavy polluting industries and state-owned enterprises are more likely to increase R&D of clean technologies. On the other hand, some companies with weaker R&D capabilities or difficulties in technology transformation may choose to directly increase end-of-pipe pollution control expenses to reduce emissions and mitigate pollution in the short term (Liu et al., 2021). Additionally, some companies may relocate from areas with high environmental regulatory intensity to areas with lower intensity, thereby weakening the technological innovation effect of environmental regulation (Wu et al., 2017). As for the industrial structure, the emerging green industries are widely distributed across the three sectors of the economy, which include environmental protection industries, clean production industries, green services, as well as the green transformation of traditional industries. Therefore, the impact of an increase in the proportion of only the tertiary sector on GTI is relatively weak.

4.2.2 Interaction of each factor

Figure 6 presents the interaction results for the determinants of urban GTI spatial differentiation in China, allowing us to determine the degree of influence and the type of interaction between any two factors on the spatial differentiation of GTI. It can be observed that, irrespective of the study period, the interactions between factors were either two-factor enhanced or nonlinearly enhanced, indicating that the combined effect of any two factors enhances the explanatory power of a single factor for the spatial differentiation of urban GTI. This also indicates that the spatial differentiation of GTI in Chinese cities is the result of the combined effects of various factors, including environmental regulation, economic development, financial development, and industrial structure.
Figure 6 Detection results for the interaction among driving factors for green technology innovation in China
In 2006, the two factors with the strongest interaction effects on GTI in Chinese cities were economic development and financial development, whereas in 2020, these two factors were government intervention and financial development, with q-values after interaction being 0.574 and 0.638, respectively. This may be related to their relatively strong explanatory power as a single factor.
When examined by time period, the interaction between economic development and financial development had the highest q-value during the period 2006-2010. This indicates that the differences in economic development and financial development during the “11th Five-Year Plan” period were the main driving factors for the spatial differentiation of GTI in Chinese cities. This also suggests that the coordinated progress of the urban economic level and financial development may contribute to promoting GTI.
During the period 2011-2015, the interaction between government intervention and financial development had a q-value of 0.510. This indicates that the differences in government intervention and financial development during the “12th Five-Year Plan” period were the key driving factors for the differentiation of GTI in Chinese cities. In this period, when the output of GTI in a particular city was relatively low, the combined effect of appropriate government intervention and improved financial development resulted in a faster increase in the level of GTI compared with the single-factor effect.
During the period 2016-2020, economic development and industrial structure were the factors with the strongest interaction, with a q-value of 0.530. This indicates that during the “13th Five-Year Plan” period, the spatial distribution resulting from the interaction between economic development and industrial structure had a good consistency with the spatial distribution of GTI. In this stage, when a city reaches a certain level of economic development that no longer promotes further growth in GTI, government adjustments to the industrial structure can stimulate an increase in the level of GTI.
The single-factor driving force of environmental regulation and industrial structure was relatively low; however, their interaction with factors such as economic development, financial development, and government intervention significantly enhanced their impact. This indicates that the effect of environmental regulation and industrial structure on GTI needs to be realized by encouraging economic development, financial development, and other interventions.

5 Conclusions and discussion

Using 264 cities in China from 2006 to 2020 as subjects, this study employes green patent data to measure the level of GTI in cities and applies Moran’s I index, Getis-Ord Gi* index, standard deviation ellipse, and geographical detector to examine the spatial pattern evolution and determinants of urban GTI in China.

5.1 Conclusions

The present study has several key findings. First, the overall level of GTI in Chinese cities is increasing, showing spatial clustering characteristics. Hot cities are mainly concentrated in the east-central region, whereas cold cities are scattered in the western region comprising Gansu, Shaanxi, and Guizhou provinces. Conversely, Heilongjiang and Jilin exhibit a contiguous distribution.
Second, the center of gravity of GTI distribution in Chinese cities has been shifting toward the southwest. The spatial distribution pattern exhibits the coexistence of centrifugal dispersion in the northwest-southeast direction and spatial agglomeration in the southwest-northeast direction. Based on the magnitude of the movement of the center of gravity and the location of the spatial distribution, it can be seen that the development pattern of GTI in Chinese cities follows the trend of “strong in the south, weak in the north, high in the east, and low in the west.”
Third, during the study period, the difference in economic development was the key factor for the spatial differentiation of GTI in Chinese cities. In terms of interaction effects, during the “11th Five-Year Plan,” “12th Five-Year Plan,” and “13th Five-Year Plan” periods, the main interacting driving factors for the spatial differentiation of GTI in Chinese cities were economic development and financial development, government intervention and financial development, and economic development and industrial structure, respectively.

5.2 Implications

The present study proposes the following policy implications based on the conclusions:
(1) Supporting low-level regional innovation and gradually narrowing the differences in GTI across regions. The spatial distribution of GTI in Chinese cities exhibits significant disparities, with the south being stronger than the north and the east being higher than the west; specifically, the northwest and three northeastern provinces have a low level of GTI, indicating a large difference in GTI among cities in the region. To address these disparities, the spatial allocation of innovation resources should be optimized, green technology investment should be increased in western and northern cities, the GTI capability of low-level cities should be gradually improved, and the differences in GTI across regions should be narrowed.
(2) Reinforcing the advantages of high-level cities and enhancing the GTI of low-level cities. The Getis-Ord Gi* index reveals hot spots and cold spots for GTI in Chinese cities, leading to differences in GTI across cities. Cities with a high GTI level should strengthen their own GTI capability, maintain good development momentum, and enhance the GTI of low-level cities through spatial correlation. To promote the overall level of GTI, a multi-pronged approach must be adopted. This includes improving the GTI system, establishing a cross-regional GTI cooperation mechanism, and actively encouraging low-level cities to engage in exchanges and collaborations with cities having high GTI levels in areas such as talent, capital, and technology. Additionally, administrative barriers and spatial restrictions that hinder the flow of GTI factors should be eliminated.
(3) Optimizing the urban innovation environment and promoting the synergistic development of GTI. Based on the geographic probe model results, it can be concluded that economic development differences are the primary determinants of the spatial variation in GTI across cities. The coordination of economic development levels among cities should be enhanced by optimizing the spatial layout of production factors and deeply implementing regional coordinated development strategies. In addition, the interaction among economic development, industrial structure, financial and technological expenditure, financial development, and government intervention has a significantly stronger impact on GTI differences than an individual factor. When promoting the urban GTI development, the combined effect of the factors should be fully utilized to form a “1+1>2” green innovation coordination function.

5.3 Limitations

GTI plays a crucial role in achieving sustainable development. Some studies have evaluated GTI in the country, key regions, and urban clusters from the perspective of efficiency or level. The spatial association network, distribution dynamics, spatial disparity, and influencing factors for GTI based on green patent data have also been examined. However, no study has used patent data to examine the spatial distribution of the center of gravity and location of GTI at the city levels or to explore the driving factors for its spatial divergence. The present study thus fills this gap in literature. However, there are some limitations in this study. First, the data availability for some cities was limited. Second, the types of GTI were not subdivided. In addition, the development trends of urban GTI and the influence of spatial factors on its dynamics were not analyzed in detail, warranting further studies.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Acemoglu D, Aghion P, Bursztyn L et al., 2012. The environment and directed technical change. American Economic Review, 102(1): 131-166.

PMID

[2]
Bai J, Jiang F, 2015. Synergy innovation, spatial correlation and regional innovation performance. Economic Research Journal, (7): 174-187. (in Chinese)

[3]
Barbera J, McConnell D, 1990. The impact of environmental regulations on industry productivity: Direct and indirect effects. Journal of Environmental Economics and Management, 18(1): 50-65.

DOI

[4]
Che D, Wu C, Ren X et al., 2020. How does fiscal technology expenditure affect enterprise technology innovation? Heterogeneity, macro-micro mechanism and government incentive structure. China Soft Science, (3): 171-182. (in Chinese)

[5]
Chen B, Peng W, Liu Y, 2022. Spatio-temporal evolution and driving factors of green innovation efficiency of the urban agglomeration in the middle reaches of the Yangtze River. Economic Geography, 42(9): 43-49. (in Chinese)

DOI

[6]
Chen Y, Wang S, 2021. Research on the structure, effects and problems of my country’s fiscal technology expenditure. Scientific Management Research, 39(5): 140-149. (in Chinese)

[7]
Chen Y, Xu Y, Wang F, 2022. Air pollution effects of industrial transformation in the Yangtze River Delta from the perspective of spatial spillover. Journal of Geographical Sciences, 32(1): 156-176.

DOI

[8]
Dong H, Li X, Zhang R, 2021. Spatial-temporal characteristics and driving factors of green innovation efficiency in Guangdong-Hong Kong-Macao Greater Bay Area. Economic Geography, 41(5): 134-144. (in Chinese)

[9]
Dong S, Ren G, Xue Y et al., 2023. Urban green innovation’s spatial association networks in China and their mechanisms. Sustainable Cities and Society, 93: 104536.

DOI

[10]
Dong S, Xue Y, Ren G et al., 2022. Urban green innovation efficiency in China: Spatiotemporal evolution and influencing factors. Land, 12: 75.

DOI

[11]
Dong Z, Wang H, 2021. Urban wealth and green technology choice. Economic Research Journal, (4): 143-159. (in Chinese)

[12]
Driessen P H, Hillebrand B, Kok R et al., 2013. Green new product development: The pivotal role of product greenness. IEEE Transactions on Engineering Management, 60(2): 315-326.

DOI

[13]
Duan D, Du D, 2022. Green technology innovation in China city system: Dynamics and determinants. Acta Geographica Sinica, 77(12): 3125-3145. (in Chinese)

DOI

[14]
Fei J, Wang Y, Yang Y et al., 2016. Towards eco-city: The role of green innovation. Energy Procedia, 104: 165-170.

DOI

[15]
Fussler C, James P, 1996. Eco-innovation:A Breakthrough Discipline for Innovation and Sustainability. London: Pitman Publishing.

[16]
Gordon I, McCann P, 2005. Innovation, agglomeration, and regional development. Journal of Economic Geography, 5(5): 523-543.

DOI

[17]
Gössling T, Rutten R, 2007. Innovation in regions. European Planning Studies, 15: 253-270.

DOI

[18]
Guo J, Zhou Y, Ali S et al., 2021. Exploring the role of green innovation and investment in energy for environmental quality: An empirical appraisal from provincial data of China. Journal of Environmental Management, 292(1): 112779.

DOI

[19]
Hall B, Harhoff D, 2012. Recent research on the economics of patents. Annual Review of Economics, 4(1): 541-565.

DOI

[20]
Hascic I, Migotto M, 2015. Measuring environmental innovation using patent data. OECD Environ Working Papers, No. 89. Paris OECD: Publishing.

[21]
Hascic I, Silva J, Johnstone N, 2012. Climate mitigation and adaptation in Africa:Evidence from patent data. OECD Environment Working Papers, No. 50. Paris: OECD Publishing.

[22]
Hascic I, Silva J, Johnstone N, 2015. The use of patent statistics for international comparisons and analysis of narrow technological fields. OECD Science, Technology and Industry Working Papers, No.2015/05. Paris: OECD Publishing.

[23]
Jin B, Li G, 2013. Green economic growth from a developmental perspective. China Finance and Economic Review, 1-4.

[24]
Kemp R, 2010. Eco-innovation: Definition, measurement and open research issues. Economia Politica, 27(3): 397-420.

[25]
Kneller R, Manderson E, 2012. Environmental regulations and innovation activity in UK manufacturing industries. Resource and Energy Economics, 34(2): 211-235.

DOI

[26]
Lefever D, 1926. Measuring geographic concentration by means of the standard deviational ellipse. American Journal of Sociology, 32(1): 88-94.

DOI

[27]
Li J, Ma X, 2019. Comparative analysis of the time-space differences and influencing factors of cities green innovation efficiency in Beijing-Tianjin-Hebei. Systems Engineering, 37(5): 51-61. (in Chinese)

[28]
Liu B, Wang L, 2020. The impact of spatial mobility of innovation factors on regional innovation capacity: Foreign attraction and local dependence. Seeking Truth, (5): 66-75. (in Chinese)

[29]
Liu Y, Wang A, Wu Y, 2021. Environmental regulation and green innovation: Evidence from China’s new environmental protection law. Journal of Cleaner Production, 297: 1-10.

[30]
Martinus K, Suzuki J, Bossaghzadeh S, 2020. Agglomeration economies, interregional commuting and innovation in the peripheries. Regional Studies, 54(6): 776-788.

DOI

[31]
Peng W, Su X, Yang S et al., 2023. Spatio-temporal evolution and spillover effects of urban green innovation under environmental regulation. Scientia Geographica Sinica, 43(1): 41-49. (in Chinese)

DOI

[32]
Porter M, Linde C, 1995. Towards a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives, 4(4): 97-118.

DOI

[33]
Rennings K, 2000. Redefining innovation: Eco-innovation research and the contribution from ecological economics. Ecological Economics, 32(2): 319-332.

DOI

[34]
Shang Y, Wang Z, Mi Z et al., 2021. Structural characteristics and optimization strategies of green technology innovation network in the Yangtze River Delta. Resources and Environment in the Yangtze Basin, 30(9): 2061-2069. (in Chinese)

[35]
Shao X, Weng Z, Miao Q et al., 2022. Evolution and element analysis of regional green technology innovation output network: Evidence from the urban agglomeration of the Yangtze River Economic Belt. Geography Geo-information Science, 38(4): 40-49. (in Chinese)

[36]
Sun B, Zhang Y, 2022. A research of the distributional dynamic evolution and regional disparities of green innovation index in China. The Journal of Quantitative & Technical Economics, (1): 51-72. (in Chinese)

[37]
Sun Y, Shen S, 2021. The spatio-temporal evolutionary pattern and driving forces mechanism of green technology innovation efficiency in the Yangtze River Delta region. Geographical Research, 40(10): 2743-2759. (in Chinese)

DOI

[38]
Tan F, Gong C, Niu Z, 2022. How does regional integration development affect green innovation? Evidence from China’s major urban agglomerations. Journal of Cleaner Production, 379: 134613.

DOI

[39]
Tian C, Li X, Xiao L et al., 2022. Exploring the impact of green credit policy on green transformation of heavy polluting industries. Journal of Cleaner Production, 335: 1-12.

[40]
Wagner M, 2007. On the relationship between environmental management, environmental innovation and patenting: Evidence from German manufacturing firms. Research Policy, 36(10): 1587-1602.

DOI

[41]
Wang B, Zhang Y, Chen L et al., 2020. Urban green innovation level and decomposition of its determinants in China. Science Research Management, 41(8): 123-134. (in Chinese)

[42]
Wang J, Du G, 2021. The spatial difference and dynamic evolution of green innovation in China’s cities. Chinese Journal of Population Science, (4): 74-85. (in Chinese)

[43]
Wang J, Du G, 2021. Spatial association network of green innovation in Chinese cities and its impact effect. China Population, Resources and Environment, 31(5): 21-27. (in Chinese)

[44]
Wang J, Xu C, 2017. Geodetector: Principle and prospective. Acta Geographica Sinica, 72(1): 116-134. (in Chinese)

DOI

[45]
Wang K, Zhang F, Xu R et al., 2023. Spatiotemporal pattern evolution and influencing factors of green innovation efficiency: A China’s city level analysis. Ecological Indicators, 146: 109901.

DOI

[46]
Wang Y, Cui C, Wang Q et al., 2021. Migration of human capital in the context of vying for talent competition: A case study of China’s “first-class” university graduates. Geographical Research, 40(3): 743-761. (in Chinese)

[47]
Wang Y, Yang Y, 2021. Analyzing the green innovation practices based on sustainability performance indicators: A Chinese manufacturing industry case. Environmental Science and Pollution Research, 28: 1181-1203.

DOI

[48]
Wang Z, Liang L, Wang X, 2021. Spatiotemporal evolution of PM2.5concentrations in urban agglomerations of China. Journal of Geographical Sciences, 31(6): 878-898.

DOI

[49]
Wu C, Xu Y, Sun K, 2022. Impact of urban economic agglomeration on green technological innovation: Spatial econometric analysis based on the panel data of 232 prefecture-level cities. Economic Geography, 42(10): 25-34. (in Chinese)

[50]
Wu H, Guo H, Zhang B et al., 2017. Westward movement of new polluting firms in China: Pollution reduction mandates and location choice. Journal of Comparative Economics, 45(1): 119-138.

DOI

[51]
Xi J, 2022. Hold high the great banner of socialism with Chinese characteristics and strive in unity to build a modern socialist country in all respects. http://www.gov.cn/xinwen/2022-10/25/content_5721685.htm.

[52]
Yang J, Xu J, Wu X, 2013. Income growth, environmental cost and health problems. Economic Research Journal, (12): 17-29. (in Chinese)

[53]
Yu Y, Lyu L, 2023. Spatial pattern of knowledge innovation function among Chinese cities and its influencing factors. Journal of Geographical Sciences, 33(6): 1161-1184.

DOI

[54]
Yu Y, Xu Z, Shen P et al., 2022. Efficiency evaluation and influencing factors of green innovation in Chinese resource-based cities: Based on SBM-undesirable and spatial Durbin model. International Journal of Environmental Research and Public Health, 19: 13772.

DOI

[55]
Zhang C, 2019. High-skilled migrants, cultural diversity and urban innovation in China. The Journal of World Economy, (11): 172-192. (in Chinese)

[56]
Zhang J, Chen S, 2021. Financial development, environmental regulations and green economic transition. Journal of Finance and Economics, 47(11): 78-93. (in Chinese)

[57]
Zhang J, Geng H, Xu G et al., 2019. Research on the influence of environmental regulation on green technology innovation. China Population, Resources and Environment, 29(1): 168-176. (in Chinese)

[58]
Zhang K, Li Y, Qi Y et al., 2021. Can green credit policy improve environmental quality? Evidence from China. Journal of Environmental Management, 298: 1-11.

[59]
Zhuang Y, Chu Q, Ma Y, 2020. Financial development, firm innovation, and economic growth. Journal of Financial Research, (4): 11-30. (in Chinese)

Options
Outlines

/

〈 〉