Research on innovative mechanisms of financial agglomeration enabling green coordinated development in the Yangtze River Delta of China

Abstract

Exploring innovative mechanisms for financial agglomeration affecting the green coordinated development of China's Yangtze River Delta is important for the city to take advantage of financial and innovative resources to promote high-quality green development. Using panel data from 41 cities in the Yangtze River Delta region from 2003 to 2019, the intermediate effects model and spatial Durbin model are conducted to deeply explore the impact of financial agglomeration on coordinated green development and the intermediary role of innovation. Results show that, first, financial agglomeration can drive green technology innovation to significantly improve the coordinated development of “production greening - social optimization - environmental protection deepening”. Second, financial agglomeration significantly improves the green coordinated development in neighboring areas through inter-city innovative spatial connections, and the spatial spillover effect tends to decay with the critical point of the maximum impact range at 240 km. Third, the intermediate mechanism of green technology innovation performs well in the high-level financial agglomeration and financial capital agglomeration areas. The Banking and securities industries are the leading factors in financial capital agglomeration. Low-level financial agglomeration and financial personnel agglomeration cannot play the “driving green” role of the innovation mechanism.

Highlights

• •The role of financial agglomeration on the innovative mechanism of green coordinated development is explored.
• •Direct and spillover effects of influence factors are explored.
• •The spatial decay trend is verified.
• •There is heterogeneity in the innovation effect of financial agglomeration.
• •The agglomeration of the banking and securities industries is the leading factor in releasing the effect of innovation.

1. Introduction

In October 2019, China's Development and Reform Commission issued the “Yangtze River Delta Ecological Green Integrated Development Demonstration Zone Overall Program”, and the scope of the integrated demonstration zone contains Qingpu District of Shanghai, Wujiang District of Suzhou City, and Jiashan County of Jiaxing City. The guiding ideology of the program points out that the demonstration zone should realize the organic unity of green economy, high quality of life and sustainable development, and come out with a new path of cross-administrative region co-build and share, and eco-civilization and economic and social development complement each other. Breaking administrative boundary constraints with regional integration, promoting the cross-regional flow of green innovation elements, and realizing the integrated development of production, life, and ecology in the demonstration area not only represent the perfection of the overall layout of regional integration in terms of ecological synergistic governance but also serve as a strategic pivot point for realizing the demonstration area to a high-level green coordinated development and have exemplary significance for other city clusters in China to explore green innovative development.

Financial development not only affects the productivity of China’s cities and provides financial support for the new urbanization [1], but also plays a driving role in the ecological green development path [2]. In the context of the financial industry agglomeration, the intrinsic connection between the financial industry and the three spaces of “production, living, and ecology” in greening construction has become increasingly close. As one of the regions with the most openness and the most well-developed financial infrastructure, it is worth exploring whether the Yangtze River Delta region can take advantage of the agglomeration dividend and have a positive impact on ecological greening. In addition, the implementation of regional integration policy provides cost reduction advantages for the flow and accumulation of innovation resources in the Yangtze River Delta cities, and green innovation development has become an important strategic component of the overall program of the demonstration area. Under the policy goal of green coordinated development of "three-living space" in the Yangtze River Delta region, it is important to explore whether financial agglomeration can drive local green coordinated development by green innovation and whether it can promote neighboring cities' green coordinated development through accelerating innovation factor spillover, which is necessary for the Yangtze River Delta region to take advantage of financial endowment and innovation resources to promote high-quality green development. Therefore, this paper focuses on two key questions. The first is to verify whether financial agglomeration can promote the coordinated green development of local and neighboring regions by driving green technology innovation, and to explore the scope of its impact. The second is to explore the various functions played by various financial agglomerations, including those with varied subjects, industries, and levels, in fostering technological innovation and boosting green coordinated development.

2. Literature review

The research of domestic and foreign scholars on green coordinated development mainly includes two aspects of assessment methods and influence factors. First, we summarize the relevant literature on assessment methods, which focuses on three major factors related to green sustainable development: economy, life, and ecology. About the measurement methods of the green economy mainly include efficiency perspective and multi-dimensional indicator system construction. Many scholars have comprehensively analyzed the evolution trend characteristics, regional heterogeneity differences, and spatial influence mechanism of green economy efficiency in China based on the DEA model [3,4]. Other scholars have evaluated the green economy by constructing an indicator system from the relevant influencing factors [5]. Green life is measured from the perspective of low carbon efficiency and high quality, and the relevant factors include the low-carbon convenience of transportation, the popularity of infrastructure development, and the improvement of residents’ welfare [6,7]. Summarizing the influencing factors of green ecology, we found that the construction pressure, health status, and governance efficiency of ecology are the key issues being considered [8,9]. In the comprehensive evaluation of green development, constructing a multidimensional indicator system is a more common practice in academia. Cheng et al. (2020) considered the development of the economy and society, consumption of natural resources, and competitiveness of the ecological environment, and found that the spatial disparity of green development levels among countries along the Belt and Road was obvious [10]. Weng et al. (2020) devised an evaluation indicator system of green development based on a “five-circle” model and found that the level of green development in Beijing, Tianjin and Hebei weakened sequentially [11]. In summary, the above evaluation indicators all involve three aspects: the economy, residents’ life, and the ecological environment. The evaluation methods of green coordinated development mainly contain the social network analysis method, factor dimensionalities reduction method such as the entropy method and coupled coordination degree model. Zhou et al. (2020) used social network analysis to find that large cities have formed a “siphon effect” and that the polarization of eco-efficiency has become increasingly serious [12]. Pan et al. (2020) constructed a DPSIR framework and used a coordinated development index model to find that the level of green coordinated development in China’s major cities has increased, but their overall level is not high [13]. Second, we summarized the relevant literature on the influencing factors related to green coordinated development. Wu et al. (2022) found the traction role of economic agglomeration and regional openness, the resonance effect of transportation infrastructure and industrial structure, and the driving role of central cities [14]. Liao and Li (2022) found that the professionalization of the urban green technology innovation promotes urban green development and narrows the gap between China's central cities and non-central cities [15], highlighting the important role of technological innovation in green ecological construction in China.

Relevant studies on the relationship between financial agglomeration (FA) and green coordinated development have been conducted in two aspects: the economic effect of FA and its action mechanism to promote green ecological development. First, in terms of economic effects, FA not only has a certain stimulating effect on macro-regional economic growth [16], but also has a significant network effect on urban industrial upgrading [17]. In the Yangtze River Delta region, Chen and Zhang (2021) found that financing support, risk diversification, and information transmission as intermediary mechanisms for FA can significantly promote the innovative development of urban agglomerations [18]. Second, in terms of action mechanisms, FA has obvious direct, indirect, threshold, and spatial effects. Yuan et al. (2019) analyzed the role of FA in promoting green development efficiency using a spatial Durbin model and a panel threshold model and found that its spatial impact was characterized by a stepwise enhancement [19]. The role of FA in promoting green-growth and resource-environment has also confirmed in Peng et al. (2022)’s research [20]. In the studies of intermediary mechanisms, Feng et al. (2022) found that industrial structure upgrading, labor force upgrading, and technological innovation are intermediate variables of FA affecting green development [21]. Yuan et al. (2021) also found that technological innovation is one of the important channels through which FA reduces environmental pollution [22]. However, in the study of exploring the influencing factors on green total factor energy efficiency, FA has a negative effect, technological innovation has a positive effect, and the interaction of FA and technological innovation has no effect [23].

The above can be found that technological innovation is a key factor affecting ecological green development and an important intermediate variable for FA affecting green development. Domestic and foreign scholars have conducted rich research on the role of technological innovation in green development. Sharif et al. (2022) use research data from the G7 countries and find that green technology innovation as well as green financing can significantly reduce CO₂ emissions [24]. Mohd et al. (2022) use research data from the ASEAN-6 countries and find the importance of technological innovation in reducing carbon emissions while finding the positive effects of replacing non-renewable energy with renewable energy [25]. Lbrahim et al. (2022) use Chinese time series data and find that technological innovation is a fundamental factor in achieving sustainable development in China and that green finance plays an essential role [26]. Chen et al. (2023) use data from 28 Chinese provinces and find that green finance significantly raises the degree of green productivity and that financial development and technological innovation contribute significantly to green production [27]. China has the advantage of green finance legislation in green development, which can help accelerate the development of green finance. Although these studies confirm the importance of financial and technological innovation for local green development, their spillover effects on neighboring regions, especially technology spillovers, are not addressed. Wang et al. (2021) use the spatial Durbin model and find that green technology innovation has a positive spatial spillover effect on green total factor productivity in China [28]. However, this study lacks clarity on the relationship between finance, technology spillover, and green development. Academics often study technological innovation spillovers together with knowledge spillovers as an important way to improve resource efficiency and achieve sustainable development goals [29,30]. The relevant empirical tools to measure spillover effects mainly include the Jaffe weight matrix, Mahalanobis approach, and citation asymmetric index. The Jaffe weight matrix could combine geographical distance and knowledge or R&D technology to consider its spillover effects, and is a common method used in academia [31]. Finance is an important tool not only to promote green development but also to accelerate technology spillover. It is necessary to clarify the role of FA in green development from the perspective of green innovation technology spillover.

The following three shortcomings are found by summarizing the above studies on the relationship among FA, technological innovation, and green coordinated development. First, although the relevant literature is rich in measuring and evaluating the relevant green development, it does not define the green coordinated development (EGI) of the Yangtze River Delta. Second, the size, direction and scope boundary of the spatial impact of FA on urban EGI under the perspective of green innovation technology linkage are not considered. Third, the heterogeneous impact of FA is ignored. In the context of green innovation linkage, besides regional heterogeneity and city size heterogeneity, different financial industry agglomeration and different levels of FA may have various effects on green coordinated development. Therefore, using the panel data of 41 cities in the Yangtze River Delta, this paper measures their EGI levels, verifies the intermediary role of technological innovation in FA affecting EGI, explores the size and scope of the spatial influence of FA on EGI using a spatial Durbin model based on the constructed green innovation correlation matrix, and discusses the heterogeneity characteristics of FA.

The marginal contributions of this paper lie in two aspects. First, this paper explores the influence mechanism of FA on EGI and the role of technological innovation in the path influence, contributing to the literature. Second, by classifying different financial subject agglomerations (including financial capital and financial personnel), different financial industry agglomerations (including banking, insurance, and security), and different financial level agglomerations (including high-level and low-level), this study compares the effectiveness of FA, which identifies the pivot and power point in optimizing EGI of Yangtze River Delta cities.

3. Theoretical analysis

3.1. The influence mechanism of financial agglomeration on green coordinated development

EGI requires the organic unity of a green economy, high quality of life, and sustainable development, which involves all fields of economy, society, and ecology. FA, with its great advantages and essential functions of optimal resource allocation, alleviation of financing constraints, and significant cost reduction, effectively solves the problems of unbalanced efficiency and equity in ecological greening construction, alleviates the contradiction between green living needs and insufficient production, and promotes the formation of economies of scale, economies of scope and green sharing. FA can not only encourage the green transformation of the economy, enhance the construction of new urbanization, and improve the welfare of residents through the implementation of the green finance strategy [32], but also force the transformation of enterprises to promote the upgrading of industrial structure to block the source of pollution, thereby driving the simultaneous improvement of economic efficiency, living standards and environmental quality, and promoting EGI.

To clarify the intermediate role played by innovative technologies in the direct impact of FA on EGI, it is necessary to focus on clarifying two problems about how FA stimulates the accumulation of green innovative technologies and whether technologies can promote the ecological green effect. First, FA can not only proactively attach innovation dynamics through financing condition constraints but also passively stimulate innovation potential by intensifying healthy competition among firms [33]. Specifically, by restricting high-pollution and high-emission enterprises (two-high enterprises) and supporting high-tech enterprises, financial institutions not only force the two-high enterprises to achieve green transformation, but also effectively solve the shortcomings of financial constraints of emerging enterprises, adding to the green innovation momentum, and also ensuring the sustainability and stability of innovation output. In the context of financial capital preferring low carbon and environmental protection industries, the financing demand intensifies industry-wide competition in the external market and accelerates the conversion of green innovation results. The high efficiency and practicality requirements of green technologies also intensified the internal competition in the high-tech industry and implement the quality of green technology products [34]. Second, technological innovation is the endogenous driving force to reverse the production mode from crude to high-quality, the core advantage to building a more intelligent and convenient social life, and the key to promoting industrial upgrading to achieve a higher ecological civilization [35]. Therefore, building a green innovation platform with FA can boost a synchronous trend of economic, social, and environmental green development, realizing the EGI of the Yangtze River Delta.

Research Hypothesis 1: FA can promote EGI of Yangtze River Delta cities.

Research Hypothesis 2: FA can drive technological innovation to promote EGI.

3.2. Spatial spillover effect of financial agglomeration on green coordinated development

In the context of regional integration, the strength and efficiency of the cross-regional flow of green innovation factors have strengthened. There is a significant positive spatial correlation between the EGI in the Yangtze River Delta cities, which provides an important channel for FA to exert spatial spillover effects. To clarify the spatial effect of FA on EGI of neighboring regions under the green innovation correlation, we need to focus on the mechanism channels of FA in releasing the spatial spillover effect, the possibility of innovation factor flow, and the role it plays. First, the concentration of financial institutions also forms capital, knowledge, information, and talent centers. With the core advantage of spider-web coverage of financial nodes, the four centers easily generate a spatial impetus from the center to the periphery, release the trickle-down effect, and promote the EGI of the surrounding cities [19]. Second, FA can enhance the ecological greening of surrounding cities by accelerating technology spillover. In essence, green innovation technology is the “knowledge” that is given value and is the core power for enterprises to use their technology monopoly to make greater profits and gain a foothold as industry leaders [35]. The spatial transfer of technology is purposeful and conscious rational behavior. Both the profit-chasing market-oriented transfer and the policy-supported government-oriented transfer provide a realistic basis for green technology spillover to release externalities [36]. The close connection between cities and the dense distribution of financial networks provide guaranteed support for technology spillovers. Finally, the diffusion of green innovation technology has an obvious “latecomer advantage”. Whether the innovative technology is shared within the region or introduced outside the region, their diffusion from local to other regions eliminates monopolistic competition at the market level, realizes the progression of green products from singular to diversified and universal, and releases the surplus value of consumers exploited by the technology monopoly, which not only essentially crosses the old production methods but also brings greater social welfare. Further, technology spillover can rapidly establish and expand the green industry chain, promote green development outward, and improve the coverage of greening technology. Local localization and normalization of green technology can be realized through the principle of “learning by doing” [37], driving EGI from both economic and environmental aspects. In addition, according to the first law of geography, the spatial correlation between different subjects decays with increasing distance [38]. The characteristics of capital, knowledge, information, talent and technology decaying with distance will lead to regional boundaries in the spatial impact of FA on EGI.

Research Hypothesis 3: FA can promote EGI of neighboring cities in the context of innovation linkages.

Research Hypothesis 4: There is a regional boundary for the spatial spillover effect of FA on EGI.

3.3. Heterogeneity analysis of the agglomeration of different financial entities

• 1.Different financial subjects’ agglomeration

FA has an important role in promoting the development of new industries, but at different stages of industrial development, the allocation ratio of financial capital and financial personnel will have different effects on EGI. Especially in the field of science and technology innovation, the traditional development mode of increasing factor inputs can no longer meet the development needs of high-tech industries, and the flow of science and technology innovation talents has become an important factor affecting the development of urban industries. Therefore, the FA represented by financial personnel and financial capital would have different impacts on EGI.

• 2.Different financial industries’ agglomeration

Different financial industries serve different social objects, play different ways of resource allocation, and support different industrial fields. Therefore, distinguishing different financial industries such as banking, insurance and securities is of great significance to optimize the allocation of financial capital, reasonably promote the agglomeration of financial industries, and effectively serve the real economy. Specifically, industrial upgrading, technology diffusion and capital transfer correspond to different industry development needs. In the process of continuous segmentation and optimization of financial fields, the corresponding different agglomeration of banking, insurance and securities will have different effects on the EGI of the Yangtze River Delta.

• 3.Different levels of financial agglomeration

Based on the spatial spillover effect of FA on EGI, different levels of FA will inevitably have different intensities of effect on EGI. On the one hand, due to the networked characteristics of FA, the development of related productive service industries represented by the financial industry will avoid the two-high enterprises, and then achieve EGI through clean and efficient production methods. On the other hand, the levels of FA also represent the maturity of financial system development. Cities with higher FA are bound to stimulate industrial structure upgrading and technological innovation of neighboring cities through the knowledge spillover effect, while cities with lower FA have a weaker spillover effect.

Research Hypothesis 5: The local and spillover effects of FA from different perspectives may be different.

The mechanism is illustrated in Fig 1.

Fig. 1.

The mechanism of FA affecting EGI.

4. Study design

4.1. Baseline regression model

To test the above research hypotheses, first, the intermediary effect of technological innovation in FA affecting EGI is tested by constructing the following model.

$${EGI}_{it}={\alpha }_0+{\alpha }_1{FA}_{it}+{\alpha }_c{Z}_{it}+{\mu }_i+{\delta }_i+{\epsilon }_{i,t}$$ (1)

$${TI}_{i,t}={\beta }_0+{\beta }_1{FA}_{i,t}+{\beta }_c{Z}_{i,t}+{\mu }_i+{\delta }_i+{\epsilon }_{i,t}$$ (2)

$${EGI}_{it}={\gamma }_0+{\gamma }_1{FA}_{i,t}+{\gamma }_2{TI}_{i,t}+{\gamma }_c{Z}_{i,t}+{\mu }_i+{\delta }_i+{\epsilon }_{i,t}$$ (3)

Among equations (1), (2), (3), ${EGI}_{it}$ is the EGI degree of city I at time t， ${FA}_{i,t}$ is the FA level of city i at time t, ${TI}_{i,t}$ is the level of green innovation of city i at time t， $Z_{i,t}$ are a series of control variables. ${\mu }_i$ and ${\delta }_i$ are time and individual fixed effects, ${\epsilon }_{i,t}$ is a random disturbance term. The significance of regression coefficients such as ${\beta }_1$, ${\gamma }_1$, ${\gamma }_2$ is used to determine whether there is an intermediary effect of green technological innovation.

4.2. Spatial econometric model

4.2.1. Spatial weight matrix construction

There are spatial correlations of green innovation factors among different cities in the Yangtze River Delta. To overcome the limitation that the geographic distance matrix only considers distance correlation, drawing on Bai and Jiang (2015) [39], the gravitational model is first used to link the strength and distance of green innovation elements between two cities, and then the matrix is used to carry the spatial correlation strength of innovation elements between any two cities. The green innovation correlation matrix is constructed as follows.

$$W_{ij}=\{\begin{array}{c}\frac{I_i{I}_j}{D_{ij}},i\ne j\\ 0,i=j\end{array}$$ (4)

In equation (4), i and j are different cities, I is the number of green innovation elements, and D is the distance between cities. $\frac{I_i{I}_j}{D_{ij}}$ is the spatial association strength of green innovation elements between any two different cities, which is proportional to I and inversely proportional to D.

4.2.2. Spatial model construction

After constructing the green innovation correlation matrix, a spatial Durbin model is set to test the spatial spillover impact of FA on EGI under this matrix.

$${EGI}_{it}={\alpha }_0+\rho W{EGI}_{it}+{\beta }_1{FA}_{it}+{\beta }_{2}W{FA}_{it}+{\beta }_3{Z}_{it}+{\beta }_{4}W{Z}_{it}+{\mu }_i+{\delta }_i+{\epsilon }_{it}$$ (5)

In equation (5), $\rho$ is the spatial autoregressive coefficient, and $W$ is the green innovation correlation matrix. ${\beta }_1$ and ${\beta }_3$ are the local effect coefficients of FA and control variables on EGI, ${\beta }_2$ and ${\beta }_4$ are the spatial effect coefficients, ${\mu }_i$ and ${\delta }_i$ are the time and individual fixed effects, and ${\epsilon }_{i,t}$ are random error terms. The spatial Durbin model requires the spatial error coefficients of ${\epsilon }_{i,t}$ to be zero.

4.2.3. Variable measurement and description

• 1.Explanatory variable: green coordination development (EGI).

The core connotation of EGI lies in the greening of economic production, the facilitation of social public services and the deepening of ecological environment governance to achieve coordination, unity, and synchronization. In this context, this paper constructs three sub-systems to evaluate the levels of economic, social and ecological green development and uses the coupled coordination model to evaluate the degree of EGI. The specific model and the measurement method of each subsystem are set as follows.

$$C={\left[\frac{3U_1{U}_2{U}_3}{{\left(U_1+U_2+U_3\right)}^3}\right]}^{1/3}$$ (6)

$$T=\alpha U_1+\beta U_2+\gamma U_3$$ (7)

$$D=\sqrt{C\times T}$$ (8)

Among equations (6), (7), (8), $U_1$, $U_2$, and $U_3$ are the production greening, social optimization, and environmental protection deepening indices, C is the coupling degree, D is the coordination degree, and $\alpha 、\beta 、\gamma$ are the coefficient to be determined. Considering that the size of these three systems is equal, set $\alpha =\beta =\gamma =1/3$.

4.2.3.1. Production greening subsystem

To achieve green development of the economy, China needs to focus on solving the problem of production efficiency. The greening of production methods is a fundamental need to achieve high-quality economic development in line with the context of China as a large manufacturing country [40]. Therefore, this paper uses the Super-SBM model with non-expected output to measure the production greening index. This model includes the three first-level indicators of input, expected output, and non-expected output. Labor, capital, innovation, and energy are selected as input indicators along with year-end total employment, capital inventory, government expenditures on science and technology, and total natural gas supply. Among these indicators, capital inventory is measured by the perpetual inventory method following the practice of Cui et al. (2022) [41]. To avoid the correlation of input indicators affecting the accuracy of the model, principal component analysis is used to measure the total input indicators [42]. Urban GDP is selected as the expected output indicator, whereas industrial wastewater emissions, industrial sulfur dioxide emissions, and industrial smoke (dust) emissions are selected as non-expected output indicators.

4.2.3.2. Social optimization subsystem

Convenient services and universal welfare are the foundation of social stability, and the optimization and upgrading of society require constantly improving communication, transportation convenience, and residents’ welfare. Therefore, this paper constructs an index system to evaluate social optimization following the practice of Feng et al. (2022) [21]. This index system includes the three first-level indicators of convenient communication, convenient transportation, and welfare protection. Internet subscribers and mobile phone subscribers are selected as convenient communication indicators. Public bus (electric) passenger volume and the number of public buses and cabs are selected as convenient transportation indicators. GDP per capita, the employees’ average salary, hospital beds, and library collections are selected as welfare protection indicators. Principal component analysis is conducted to measure the social optimization index.

• (3)Environmental protection deepening subsystem

The connotation of environmental protection deepening lies in the ecological environment governance to achieve high quality and high efficiency [43,44]. Therefore, this paper constructs an index system to evaluate the environmental protection deepening, which includes the two first-level indicators of high environmental quality and high environmental efficiency. The greening coverage rate of the built-up area and park green area are selected as high environmental quality indicators [45]. The solid waste, sewage, and garbage treatment rates are selected as the high environmental efficiency indicators.

• 2.Explanatory variable: Financial agglomeration (FA).

Domestic and foreign scholars mostly use the location entropy method to measure the level of FA, but in the current context of diversified financial development, the proportion of securities and insurance institutions in the financial system is increasing year by year. Using the location entropy method based on the balance of deposits and loans or financial employees to measure FA is not consistent with the real situation of financial development in each region and does not accurately reflect the real level of FA [46]. Following Wang et al. (2019) [46], this paper constructs a comprehensive evaluation indicator system that includes different financial industries and financial personnel to measure FA and assigns certain weights to the selected indicators. These financial industries include banking, securities, and insurance, and the overall agglomeration of these industries is regarded as financial capital agglomeration (CAP). Banking industry agglomeration (BAN) includes three indicators: the deposit and loan balance ratio to the built-up area, the deposit and loan balance ratio to the population, and the savings ratio to the population [47]. Securities industry agglomeration (SEC) includes three indicators: the number of IPO companies issued in the A-share market ratio to the total number in the Yangtze River Delta, the number of securities companies ratio to the total number, and the number of public funds ratio to the total number. These ratios are presented in year-to-date values. Insurance industry agglomeration (INS) includes one indicator: the number of medical insurance participants ratio to the total number. Meanwhile, financial personnel agglomeration (PEO) includes one indicator: the number of financial industry employees ratio to the total number. It should be noted that FA includes BAN, SEC, INS, and PEO totaling 8 indicators, CAP includes BAN, SEC, and INS totaling 7 indicators, and BAN and SEC each include 3 indicators, all measured by using principal component analysis.

• 3.Intermediate Variable: green innovation (TI).

Following Lin and Ma (2022), and Li et al. (2022) [48,49], green innovation patent application is selected as a green innovation indicator (TI). In addition, to verify the robustness of this model, a green innovation patent grant is selected for robustness testing.

• 4.Control variables.

In this paper, we set some control variables including foreign investment (FDI), talent pool (TR), science and technology investment (STI), government intervention (GI), and environmental manpower density (EE). The stock of FDI with logarithm is selected as a foreign investment indicator following the principle of objectivity, which is consistently calculated using the perpetual inventory method. The university students per 10,000 people are selected as the talent pool indicator. The ratio of government expenditure on science and technology to total fiscal expenditure is selected as the science and technology investment indicator. The ratio of fiscal expenditure to GDP net of science and education expenditure is selected as the government intervention indicator following Shao et al. (2013) [50], which is eliminate the multicollinearity with the science and technology input variables. The ratio of employees in the water, environment, and public facilities management industry to the total employees is selected as the environmental manpower density indicator.

4.2.4. Data sources and descriptive statistics

This paper uses 41 prefecture-level and above cities in the Yangtze River Delta region of China from 2003 to 2019 as the research sample. Securities industry data in the FA is collected from the Cathay Capital Database (CSMAR). TI data is collected from the China Research Data Service Platform (CNRDS). The other sample data are collected from the China City Statistical Yearbook, China City Construction Statistical Yearbook, and statistical yearbooks of provinces and cities. Smoothing is performed for individual missing data.

Table 1 reports the descriptive statistics for each indicator. Among them, the maximum and mean values of EGI are 0.837 and 0.342, indicating that the EGI level in the Yangtze River Delta region is relatively balanced. The maximum and mean values of FA are 0.960 and 0.081, indicating the large differences in FA levels. The central city has a concentrated financial industry, whereas other cities have weak financial development. CAP, PEO, BAN, SEC, and INS also exhibit the characteristics of “small mean and large error”. FDI and GI exhibit the characteristic of large mean, indicating that local governments have obvious intentions to promote urban development by increasing the degree of opening and financial support.

Table 1.

Descriptive statistics.

	Variable Name	Variable Symbols	Number of observations	Average value	Standard deviation	Minimum value	Maximum value
Explained variables	Green coordinated development	EGI	697	0.342	0.123	0.089	0.837
Explanatory variables	Financial agglomeration	FA	697	0.081	0.100	0.006	0.960
	financial capital agglomeration	CAP	697	0.071	0.108	0.000	0.993
	Banking industry agglomeration	BAN	697	0.096	0.126	0.000	0.991
	Securities industry agglomeration	SEC	697	0.022	0.081	0.000	1.000
	Insurance industry agglomeration	INS	697	0.085	0.142	0.000	1.000
	financial personnel agglomeration	PEO	697	0.282	0.152	0.000	1.000
Intermediate variables	Green innovation	TI	697	1.937	0.953	0.000	4.078
Control variables	Foreign investment	FDI	697	6.330	0.887	2.917	8.234
	Talent pool	TR	697	0.018	0.020	0.000	0.127
	Science and technology investment	STI	697	0.024	0.020	0.000	0.130
	Government intervention	GI	697	3.006	0.192	2.572	4.163
	Environmental manpower density	EE	697	0.015	0.007	0.001	0.075

5. Empirical analysis

5.1. Baseline regression analysis

Table 2 reports the mechanism test results of FA on EGI. Firstly, model (1-1) is the direct effect of FA on EGI. It shows that the regression coefficient of FA is significantly positive at the 1% level, indicating that FA can significantly promote EGI of the Yangtze River Delta, which verifies hypothesis 1. Secondly, model (1–2) verifies that FA can promote TI. Finally, putting the intermediary variable TI back into the regression equation of FA on EGI. The result shows that the regression coefficient of FA on EGI in the model (1–3) is not significant, whereas the regression coefficient of TI is significant, indicating that TI plays a full intermediary effect and is an important intermediary mechanism for FA promoting EGI, verifying the Hypothesis 2. In addition, the intermediary indicator TI is replaced for the robustness test, and the regression results are shown in model (1–3) and model (1–4). In model (1–3), the regression coefficient of FA on TI is significant. In model (1–4), the regression coefficients of FA and TI are insignificant and significant. This verifies that TI is an important intermediary mechanism of FA empowering EGI.

Table 2.

Intermediary mechanism test of FA on EGI.

Variables	EGI	TI	EGI	TI	EGI
	Model (1-1)	Models (1–2)	Models (1–3)	Models (1–3)	Models (1–4)
FA	0.035*** (0.012)	0.115*** (0.003)	−0.029 (0.021)	0.121*** (0.002)	−0.024 (0.026)
TI	–	–	0.562*** (0.151)	–	0.492** (0.191)
FDI	−0.001 (0.001)	−0.001 (0.001)	−0.001 (0.001)	−0.001 (0.001)	−0.001 (0.001)
TR	−0.013 (0.128)	0.004 (0.033)	−0.016 (0.126)	0.046* (0.026)	−0.036 (0.127)
STI	0.427*** (0.056)	0.138*** (0.014)	0.349*** (0.059)	0.066*** (0.011)	0.394*** (0.057)
GI	0.013* (0.007)	0.002 (0.002)	0.011 (0.007)	0.001 (0.001)	0.013 * (0.007)
EE	−0.833*** (0.117)	−0.098*** (0.030)	−0.778 *** (0.117)	−0.082*** (0.024)	−0.792 *** (0.118)
Constant term	0.214*** (0.023)	−0.012** (0.006)	0.221*** (0.023)	−0.004 (0.004)	0.216*** (0.023)
Fixed effects	YES	YES	YES	YES	YES
R2	0.966	0.819	0.967	0.866	0.967

Note: Robust standard errors are reported in parentheses.

5.2. Spatial econometric analysis

5.2.1. Moran’s I test

Before the spatial econometric analysis, the potential spatial correlation in the EGI of cities needs to be tested. In this paper, the spatial effect is calculated for each year under the green innovation correlation matrix using Moran's I method. Table 3 shows that Moran's I index of EGI from 2003 to 2019 is all significantly positive at the 1% level, indicating that the EGI of the cities in the Yangtze River Delta shows a significant positive spatial correlation.

Table 3.

Moran’s I test.

Year	EGI	Year	EGI	Year	EGI	Year	EGI
2003	0.037** (2.131)	2008	0.066*** (3.114)	2013	0.065*** (3.113)	2018	0.069*** (3.282)
2004	0.055*** (2.748)	2009	0.064*** (3.014)	2014	0.059*** (2.908)	2019	0.069*** (3.314)
2005	0.053*** (2.668)	2010	0.060*** (2.908)	2015	0.057*** (2.827)
2006	0.060*** (2.897)	2011	0.057*** (2.794)	2016	0.053*** (2.732)
2007	0.064*** (3.043)	2012	0.057*** (2.795)	2017	0.071*** (3.382)

Note: The z-values are reported in parentheses.

5.2.2. Spatial regression analysis

To verify the local and spillover effects of FA affecting EGI under the spatial association of green innovation factors, the LM test, Hausman’s test, and SDM model simplification test need to be passed. The spatial Durbin model (SDM) with time-fixed and entity-fixed effects is selected for analysis. The point estimation results in the SDM model have an error probability in explaining the spatial effects, whereas the direct and indirect effects obtained from partial differential decomposition can well explain the local and spatial effects. Therefore, this paper reports the partial differential estimation results, whereas the point estimation results report only the estimates of FA [32].

The results of model (2-1) in Table 3 show that: (1) the direct and spatial effect coefficient of FA are positive at the 5% significance level both in point estimation and partial differential estimation, indicating that FA can improve local and neighboring EGI under the green innovation linkage of the YRD cities, which verifies hypothesis 3. The explanation for this phenomenon is that FA can accumulate green innovative technologies by adding innovation power and stimulating innovation potential, optimizing traditional production methods, reducing pollution at source, and empowering local green ecology. In addition, FA can accelerate green technology spillover under both market and government guidance, release agglomeration dividends, and bring positive externalities to neighboring areas. (2) In terms of control variables, the direct and spatial effects of FDI are significantly positive and negative, indicating that the introduction of foreign investment can bring positive local effects on EGI by bringing advanced innovative technologies, whereas the out-migration of pollution leads to negative spillover effects. The direct and spatial effects of TR are insignificant and negative because there is a certain constraint and lag in releasing the greening effect by talent-driven innovation. The migration of talents to the central region leads to a serious brain drain from the surrounding areas, especially the lack of professional talents, which has a significant negative impact on ecological construction. The direct effect of STI and GI are positive, whereas the spatial effects are negative. It indicates that all financial interventions by local governments can bring innovative guidelines for local development, but the "self-interest intention" and obvious disorderly competition among local governments in the Yangtze River Delta region have not opened the pattern of integrated development [51]. The direct effect of EE is negative due to the poor quality of environmental manpower inputs, which tends to breed corruption in the environmental sector [52], rendering the innovation-driven role of eco-greening largely ineffective.

5.2.3. Robustness test

First, to verify the robustness of the model, this paper replaces the TI indicator using a green innovation patent grant and recreates the spatial weight matrix for spatial econometric analysis. The regression results of model (2-2) in Table 3 show that: the coefficients and significance of the core explanatory variables FA are enhanced and pass the robustness test. Second, this paper adopts the dynamic SDM model to solve the endogeneity problem caused by the spatial lag term and time lag term of the explanatory variables and adopts the one-period lag term of FA as the instrumental variable to solve the associative endogeneity between FA and EGI [53]. The regression results of models (2–3) and (2–4) in Table 4 show that both the direct and spatial coefficients of FA on EGI are positive at the 5% significance level, which verifies the stability of the model.

Table 4.

Spatial regression results of FA on EGI.

Variables		Model (2-1)		Model (2-2)		Models (2–3)		Models (2–4)
Baseline return	FA	0.034**	(0.014)	0.042***	(0.014)	0.031*	(0.016)	0.038**	(0.015)
	W*FA	0.127***	(0.049)	0.162***	(0.051)	0.159***	(0.057)	0.125**	(0.053)
	Control variables	YES	–	YES	–	YES	–	YES	–
Direct effect	FA	0.034**	(0.014)	0.041***	(0.014)	0.030*	(0.016)	0.037**	(0.014)
	FDI	0.007***	(0.001)	0.006***	(0.001)	0.007***	(0.002)	0.007***	(0.002)
	TR	−0.036	(0.123)	−0.042	(0.124)	0.129	(0.144)	0.067	(0.144)
	STI	0.396***	(0.055)	0.392***	(0.055)	0.386***	(0.056)	0.321***	(0.056)
	GI	0.014*	(0.008)	0.014*	(0.008)	0.010	(0.008)	0.003	(0.008)
	EE	−0.892***	(0.105)	−0.891***	(0.106)	−0.901***	(0.107)	−0.625***	(0.107)
Indirect effects	FA	0.094**	(0.038)	0.122***	(0.041)	0.124**	(0.048)	0.092**	(0.040)
	FDI	−0.041***	(0.013)	−0.040***	(0.014)	−0.042*	(0.017)	−0.047***	(0.017)
	TR	−1.147***	(0.301)	−1.160***	(0.290)	−1.359***	(0.444)	−1.430***	(0.405)
	STI	−1.008*	(0.342)	−1.103***	(0.371)	−1.063***	(0.361)	−0.796**	(0.331)
	GI	−0.141*	(0.079)	−0.165**	(0.082)	−0.148*	(0.084)	−0.128	(0.081)
	EE	0.200	(1.134)	−0.110	(1.198)	−0.152	(1.228)	−0.154	(1.188)
Overall Effect	FA	0.128***	(0.040)	0.163***	(0.043)	0.155***	(0.052)	0.130***	(0.044)
	FDI	−0.034***	(0.013)	−0.033**	(0.014)	−0.034**	(0.017)	−0.040**	(0.017)
	TR	−1.184***	(0.326)	−1.203***	(0.316)	−1.230***	(0.478)	−1.363***	(0.436)
	STI	−0.609*	(0.345)	−0.710*	(0.373)	−0.677*	(0.364)	−0.474	(0.334)
	GI	−0.127	(0.077)	−0.150*	(0.080)	−0.137*	(0.083)	−0.125	(0.080)
	EE	−0.692	(1.126)	−1.002	(1.189	−1.054	(1.222)	−0.780	(1.183)
Fixed effects		YES	–	YES	–	YES	–	YES	–
$R^2$		0.556	–	0.494	–	0.439	–	0.489	–
LogL		2107.780	–	2108.880	–	1999.373	–	1996.213	–

Note: Robust standard errors are reported in parentheses.

5.2.4. Spillover boundary exploration

The above study verifies the direct and spatial effect of FA on EGI, whereas the scope of this spatial influence should be further analyzed. This study assumes that the shortest initial distance between cities in the Yangtze River Delta is 80 km, and the geographical distance setting in the green innovation association matrix $W^{*}$ is limited by a threshold of 20 km increments. So, $W^{*}=\{\begin{array}{c}\frac{I_i{I}_j}{d_{ij}},\mathrm{W}\mathrm{h}\mathrm{e}\mathrm{n}{d}_{ij}\mathrm{i}\mathrm{s}\mathrm{o}\mathrm{u}\mathrm{t}\mathrm{s}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}\\ 0,\mathrm{W}\mathrm{h}\mathrm{e}\mathrm{n}{d}_{ij}\mathrm{i}\mathrm{s}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}\end{array}$. By observing the SDM model regression results under different thresholds, this paper records the spatial effect coefficients, significance levels, and standard errors, and plots the spatial effect decay trend (see Fig. 2).

Fig. 2.

The spatial decay process of financial agglomeration affecting green coordinated development.

The regression results show that the spatial coefficient of FA is significantly positive at the 1% level in the regional range of 160–280 km, and the impact strength shows a decaying trend of increasing and then decreasing. The spatial coefficient is no longer significant for distances smaller than 160 km and larger than 280 km, verifying hypothesis 4 that the spatial impact of FA on EGI is decaying. Specifically, the spatial impact of FA on EGI is negative and insignificant for distances smaller than 160 km, probably because the “polarizing” effect of FA on the capital of neighboring cities is stronger, showing the phenomenon that financial centers cannot coexist in neighboring areas [52]. FA gradually shows a positive spatial effect on EGI when the distance threshold is 160–280 km, and the promotion effect is strongest at 240 km, which is exceeded the range of two neighboring areas. This indicates that the “trickle-down” effect of FA has achieved cross-city propagation, and the provincial central cities are able to achieve radiation coverage of provincial remote areas. The spatial effect gradually weakens and the “trickle-down” effect dividend gradually disappears when the distance threshold increases further, verifying the spatial decay hypothesis 4. Whereas, the spatial coefficient is negative and insignificant for distances larger than 280 km.

6. Further exploration: the heterogeneous impact of financial agglomeration

6.1. Sub-financial capital and financial personnel agglomeration

In a micro sense, FA is the spatial development trend of financial capital and financial personnel tending to gather in the central city, and the agglomeration of financial capital is mainly attached to financial institutions. In this paper, the agglomeration of financial resources in the banking, securities (including the fund industry), and insurance industries are grouped as CAP. Considering the different development patterns of CAP and PEO in terms of geographical location, it is necessary to further explore whether these two different financial subjects empower EGI.

The results in Table 5 show that: (1) the direct and spatial effect coefficient of CAP is significantly positive, and the direct effect coefficient is larger than that of FA, indicating that CAP can drive green technology innovation and generate technology spillover to promote local and neighboring EGI. CAP is directly related to the micro subjects of investment and financing, which is the most important reason for FA to release the local effect. (2) The direct and spatial effect coefficient of PEO are significantly negative, indicating that PEO does not form the trend of driving green innovation to influence ecological greening but has a negative impact. These phenomena are mainly due to the mismatch of factor resources between financial personnel and financial capital. On the one hand, the over-saturation of financial personnel input hinders the development process of FA promoting EGI [6]. On the other hand, the poor quality of financial personnel input has not played a role in promoting EGI.

Table 5.

Spatial regression results of financial heterogeneity.

		FA=CAP	FA=PEO	FA= BAN	FA=SEC	FA=INS	High-level FA	Low-level FA
Direct Effect	FA	0.042*** (0.013)	−0.033*** (0.005)	0.034*** (0.012)	0.042*** (0.012)	0.018* (0.010)	0.038** (0.019)	−0.162 (0.248)
	FDI	0.007*** (0.001)	0.007*** (0.001)	0.006*** (0.001)	0.007*** (0.001)	0.006*** (0.001)	0.001 (0.002)	0.025*** (0.005)
	TR	−0.055 (0.123)	−0.044 (0.124)	−0.064 (0.123)	−0.038 (0.123)	−0.003 (0.124)	0.721*** (0.190)	−0.343 (0.319)
	STI	0.387*** (0.054)	0.401*** (0.054)	0.382*** (0.055)	0.390*** (0.054)	0.416*** (0.054)	0.350*** (0.088)	0.415** (0.167)
	GI	0.015* (0.008)	0.015** (0.008)	0.015* (0.008)	0.012 (0.008)	0.014* (0.008)	0.011 (0.013)	−0.043* (0.023)
	EE	−0.885*** (0.105)	−0.685*** (0.105)	−0.877*** (0.105)	−0.896*** (0.104)	−0.854*** (0.107)	−1.609*** (0.203)	−0.441*** (0.164)
Indirect Effect	FA	0.097*** (0.036)	−0.097* (0.050)	0.089** (0.038)	0.123*** (0.035)	0.011 (0.027)	0.223** (0.100)	−7.506** (3.830)
	FDI	−0.034*** (0.013)	−0.058*** (0.012)	−0.037*** (0.013)	−0.038*** (0.012)	−0.055*** (0.013)	−0.045 (0.033)	0.218*** (0.070)
	TR	−1.166*** (0.296)	−1.000*** (0.281)	−1.087*** (0.298)	−1.178*** (0.296)	−1.154*** (0.304)	4.597*** (1.445)	−2.491 (3.974)
	STI	−1.020*** (0.338)	−0.768** (0.305)	−1.005*** (0.343)	−0.932*** (0.333)	−1.054*** (0.347)	−1.671* (0.870)	0.954 (1.687)
	GI	−0.154** (0.077)	−0.044 (0.068)	−0.126 (0.080)	−0.167** (0.076)	−0.132* (0.078)	−0.133 (0.186)	−0.868** (0.350)
	EE	0.032 (1.107)	2.783*** (1.038)	0.534 (1.099)	0.005 (1.058)	1.208 (1.184)	−7.350*** (2.705)	1.163 (3.196)
Overall Effect	FA	0.140*** (0.038)	−0.130*** (0.049)	0.122*** (0.041)	0.166*** (0.037)	0.029 (0.026)	0.262** (0.113)	−7.668* (3.992)
	FDI	−0.026** (0.012)	−0.051*** (0.011)	−0.030** (0.013)	−0.030** (0.012)	−0.049*** (0.012)	−0.045 (0.034)	0.243** (0.074)
	TR	−1.221*** (0.321)	−1.045*** (0.301)	−1.152*** (0.325)	−1.216*** (0.320)	−1.157*** (0.332)	5.319*** (1.590)	−2.834 (4.203)
	STI	−0.633* (0.340)	−0.366 (0.308)	−0.622* (0.345)	−0.542 (0.334)	−0.637* (0.349)	−1.320 (0.926)	1.370 (1.829)
	GI	−0.139* (0.076)	−0.028 (0.067)	−0.110 (0.078)	−0.154** (0.075)	−0.117 (0.077)	−0.121 (0.189)	−0.911** (0.369)
	EE	−0.852 (1.098)	2.097** (1.033)	−0.342 (1.090)	−0.890 (1.049)	0.353 (1.179)	−8.959*** (2.782)	0.721 (3.271)
Fixed effects		YES	YES	YES	YES	YES	YES	YES
R2		0.492	0.776	0.480	0.537	0.763	0.721	0.786
LogL		2109.338	2120.049	2108.078	2111.662	2105.751	1240.614	932.363

Note: Robust standard errors are reported in parentheses.

6.2. Sub-financial agglomeration of different industries

The above paper finds that CAP is the main reason for releasing the agglomeration effect. In the context of the gradual enrichment of financial institution subjects in China, it is necessary to divide CAP into BAN, SEC, and INS according to specific carriers, and further examine the possible different impacts of different CAP in driving green innovation to empower EGI. The results in Table 5 show that: (1) BAN and SEC can significantly promote local and neighboring EGI through green innovation drive and spillover, and the effective coefficient of SEC is greater. (2) The direct and spatial effects of INS on EGI are positive and insignificant, whereas the strength of the direct effect is weaker than that of BAN and SEC. Explanations for the above phenomenon: (1) the securities industry mainly relies on the form of green credit can better support local EGI, and due to its stronger spatial spillover effect leads to stronger promotion of neighboring EGI. The banking industry shows a relatively stable economic effect, and its role in EGI is more similar to that of the securities industry. (2) Financial and insurance inputs have obvious local regulatory characteristics [54], so the promotion effect of INS on EGI is mainly reflected in the local effect.

6.3. Different levels of financial agglomeration

Further, the cities of the Yangtze River Delta are divided into two regions with high-level FA and low-level FA to test whether different levels of FA have a different impact on EGI. The results in Table 5 show that: (1) regions with high-level FA have higher intensity in promoting EGI, and consistent estimates are obtained for both direct effect and spatial effect. (2) Regions with low-level FA show negative effects on EGI, and the spatial effect is significant. The main reason for this is that the region with high-level FA forms stronger siphoning and polarization effects on the region with low-level FA, which is more common in the early stage of FA. However, with the increasing level of FA in the Yangtze River Delta city cluster, the problem of regional coordination has become a major obstacle to the green and coordinated development of the city cluster. Thus, the government needs to address the issue of regional balanced development under the framework of urban cluster integration in the future.

7. Conclusions, policy implications, and research outlook

Using panel data from 41 cities in China’s Yangtze River Delta region from 2003 to 2019, this paper estimates the indexes of FA and EGI, examines the direct, intermediary, and spatial effect of FA on EGI based on the intermediary effect model and the SDM model. The main findings are summarized as follows. First, FA can significantly improve local EGI, and green technology innovation plays an important intermediary mechanism role. Second, in the context of green innovation factor linkage among cities in the Yangtze River Delta, FA has a positive direct and spatial effect on EGI, and the spatial spillover intensity shows a trend of “increasing first and then decreasing”, with positive spatial spillover area within 160–280 km and spatial spillover effect disappearing outside 280 km. Third, in promoting EGI, CAP has positive direct and spillover effects, BAN and SEC are the most important forces of CAP, INS has only direct effect, whereas PEO has negative direct and spillover effects. In addition, high-level FA has a stronger positive effect, whereas low-level FA has a significant negative effect.

Accordingly, this study puts forward the following policy recommendations: First, the cities in the Yangtze River Delta should continue to strengthen financial institutions’ fund support for enterprises’ green innovation, optimize financing constraints, support the development of high-tech industries, force high-polluting enterprises to innovate and upgrade, and bring into play the linkage effect of FA and green innovation in the Yangtze River Delta. Second, cities should improve financial infrastructure construction, strengthen cross-regional links among financial institutions, and fully release the trickle-down dividends of FA to neighboring regions. Enhancing the advantages of agglomeration can start from the spatial spillover boundary, pay attention to the financial development within the critical range of spatial decay, improve the linkage effect between the center and the boundary, and broaden the financial coverage. In addition, cities need to accelerate regional integration and break the constraints such as administrative and trade barriers to the flow of innovative factors. Third, the cities need to reasonably optimize the allocation ratio of financial personnel and financial capital, focus on promoting the level of BAN and SEC, strengthen the role of policy support and technology leadership, and build a reasonable and orderly financial agglomeration industry. Cities need to optimize the structure of financial personnel, introduce high-quality talents, and improve the professionalism of financial personnel, to eliminate the negative effects of PEO. In addition, regions with low-level FA should focus on implementing the needs of local industrial development, learn from the experience of developed regions, and give full play to their “post-emergence advantages”.

This study mainly analyzes the innovative mechanism role of FA on EGI in the Yangtze River Delta region, which enriches the related literature. However, there are still some shortcomings, such as other acting mechanisms and regional heterogeneity that can be further expanded for this study. It is well known that cities are more closely connected with Internet development, and information transfer and technological linkage play important roles. The Yangtze River Delta region of China contains many small, closely connected groups, such as the metropolitan areas. A comparative study of the key driving factors and their action mechanisms for EGI of the Yangtze River Delta metropolitan areas is thus essential to find a strategic support point that suits each of these metropolitan areas. Therefore, our research outlook is to find more intermediate mechanisms of FA on EGI and to further enrich this study by comparing the “greening effectiveness” of FA from provincial and metropolitan area heterogeneity.

Author contribution statement

Fuqiang Wu: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Xiaoli Yang: Contributed reagents, materials, analysis tools or data.

Yujia Chen: Conceived and designed the experiments; Wrote the paper.

Funding statement

This work was supported by Jiangsu Province Higher Education Teaching Reform Top Priority Project (2021JSJG004); Jiangsu Provincial Postgraduate Research and Practice Innovation Program Project "Measurement and Comparison of Regional Center Urban Resilience Development" (KYCX212484).

Data availability statement

Data will be made available on request.

Declaration of interest’s statement

The authors declare no conflict of interest.