The Impact of Digital Finance on the Green Utilization Efficiency of Urban Land: Evidence from 281 Cities in China

Abstract

In the era of digital economy, digital finance, as an innovative financial model, plays an important role in driving urban industrial transformation and development, technological innovation, industrial upgrading and sustainable utilization of energy, and has a significant impact on sustainable urban development. At present, in the process of building a new pattern of Chinese-style modernization in China, it is of great significance to improve the green use efficiency of urban land through digital finance and realize the sustainable use of land resources and the sustainable development of the city. The current study employed 281 Chinese cities from 2010 to 2020 as research samples to investigate the effects of technological financing on the productivity of city land green usage. Based on the ideas of responsible growth and efficient urban development, an assessment index system was developed. Comprehensive empirical tests, such as the Super-SBM model, fixed effect model, and mediation effect model, were implemented in the research. The study’s findings indicate that: (1) Throughout the research period, the benchmark model’s regression outcomes demonstrate that digital banking impacts urban land’s green development efficiency, with positive moderating effects offered by environmental legislation; the optimization of industrial structure has not yet played a positive regulating effect. (2) Urban area green usage performance is more clearly impacted by the extent of use and the degree of digitization, according to the regression results of digital financing heterogeneity. The positive effect of online financial services on a city’s green use efficiency occurs mainly in eastern cities and southern cities, given the results of urban development level difference. In light of resource endowment unpredictability, “non-resource cities” stand to gain more from global finance’s encouragement of resource-efficient urban land use than do “resource cities”. The results of the mechanism test indicate that there is a strong mediating influence from digital finance, city land environmental use productivity, and green technological breakthroughs. In consideration of these results, the following measures are suggested in this paper: (1) Persist in advocating for the transformation of traditional finance into online financing. (2) Intensify the impact of significant variables on the environmentally friendly use of urban areas. (3) Encourage technology creativity and execution through the application of technological economics.

1. Introduction and Literature Review

Following the implementation of economic reforms and liberalization, China’s economy has seen significant advancements. China’s gross domestic product (GDP) increased by an astounding 14.08% per year over 45 years, from CNY 367.87 billion to CNY 121 trillion between 1978 and 2022. In this context, using blockchain, artificial intelligence (AI), massive data sets, and other innovations in the financial sector, China’s financial industry is entering an era of digital transformation. A significant development in China’s economy has been the rise of digital finance, which was officially recognized in September 2016 with the release of the G20 High-Level Guidelines for Digital Financial Inclusion. Distinguished from “e-finance” and “Internet finance”, the prevailing perception of digital finance is that it is a brand-new category of economic services that combines traditional financial services with massive data sets, AI, the wireless Internet, and additional digital innovations [1]. China’s digital finance index rose 26.9% each year from 2011 to 2021 (http://www.199it.com/archives/1489063.html (6 September 2022)), and its fast growth is crucial to economic change and development [2,3,4,5,6], technological innovation [7] and industrial upgrading [8,9], and recognizing the city’s upward trajectory [10]. Numerous academic studies demonstrate how the growth in electronic financing can enhance the effectiveness of urban green innovation [11], curb carbon emissions [12,13,14], empower urban green economic growth [5,15], and assist with urban environmental improvement [16] and high-quality economic development [17].

As the urbanization construction process continues to advance, the demand for urban built-up land area in China shows a rapid upward trend. The city population of China is expected to grow by 23.6 thousand km² between 2010 and 2022, reaching 63.7 thousand km². Furthermore, all cities in the country have over 75% of their urban property efficiency categorized as medium-low [18]. The “three highs” of high energy consumption, pollution, and emissions are driving the fast increase in urban building land area and inefficient use, which threatens land carbon sequestration and sustainable use [19]. It contradicts the principle of sustainable development in China. The emergence of ecological civilization and the advancement of urban sustainability in China is strongly dependent on the efficient, ethical, and ecologically responsible use of the nation’s limited supply of urban areas.

As a new type of land utilization combining land elements and green development concept, for urban land utilization for green spaces, the main goal is to achieve optimal use of urban space via efficient economic growth, while prioritizing intensity, environmental friendliness, and long-term sustainability [20,21]. In an effort to construct an assessment index system, academics started making efforts based on the basic connotations, which at first only considered inputs and expected outcomes for the analysis and evaluation of city land use’s financial sustainability [22,23]. Subsequently, scholars have increasingly focused on the importance of non-desired outputs in economic activities, guided by the concepts of green, low-carbon, and sustainable development. They argue that true efficiency in urban land utilization can only be achieved by considering the green utilization efficiency while accounting for non-desired outputs [24]. Therefore, some scholars include industrial “three wastes” pollution elements into non-desired outputs on the basis of desired outputs [25], and some experts see urban population as an undesirable outcome while developing an urban land green use efficiency analysis rating method [26]. Following consideration of the undesirable result in the measurement process, the mainstream method has changed from the DEA model to the Super-SBM model [27,28]. Regarding the impact of certain data components on urban land use, several researchers have already carried out initial investigations and determined that the digital economy may improve the effectiveness of environmentally friendly green land use [29].

To increase the effectiveness of long-term urban land utilization, technical guidance is essential. Green technical innovation, distinct from general technological innovation, places more emphasis on environmental advantages. It provides an essential technological basis for the building of a sustainable society and accelerates the adoption of a green growth and transition pattern [30]. Innovation in environmentally friendly innovations is hampered by the high cost of financing, the presence of high risk and uncertainty, and the demanding needs for innovation. Consequently, securing conventional financial assistance for green technology innovation financing becomes challenging [31]. In this case, digital finance, with its advantages of universality and convenience, has injected new vitality into driving green technological innovation. Additionally, experts in related studies often hold the view that digital finance has the potential to facilitate the development of environmentally friendly technology innovation [32,33,34,35]. Thus, may the effective and sustainable usage of metropolitan areas be indirectly affected by the advances in ecological technology brought about by the increase in electronic financing?

In summary, urban area green usage efficiency measuring techniques, concept formulation, and other factors have yielded important and relevant study outputs for scholars both locally and globally, but there are also the following shortcomings: 1. The majority of previous studies on how the digital sector affects the efficiency of renewable land use in cities have focused on testing novel initiatives, with no literature investigating the specific mechanisms via which digital finance affects urban land green utilization efficiency; 2. Studies on the green usage potential of cities are scarce, especially when renewable energy innovation is the underlying process. This study incorporates findings from prior studies in light of China’s transition towards a green economy and the rapid growth of digital economics. It views industrial “three wastes” and labor, capital, and land as energy inputs; it views economic, social, and environmental outputs as undesirable; and it views energy inputs as inputs. The research thoroughly evaluates the effectiveness of urban land use, with an emphasis on green land use, via the Super-SBM model. The study examines panel data from 281 Chinese cities between 2010 and 2020. The analysis intends to provide practical recommendations for sustainable use of land resources, sustainable city development and the establishment of financial policies using the fixed impact and interactive outcome concept to promote the growth of the green economy.

2. Research Questions and Analytical Examination

2.1. Influencing the Effectiveness of Green Urban Area Use: A Theoretical Examination of Digital Financing

The advent of modern financing, which includes technologies including the blockchain, cloud-based computing, and massive data sets, has had a major effect on the property sector. The impact is most noticeable when it comes to ecological legislation and workplace optimization, which in turn affects how well urban land is used for green initiatives.

The ideal environmental system and policies are necessary to achieve ecological and sustained city land management. The local government needs to take advantage of financial innovation, using loans, financial services, and other means, and the objective is to achieve environmentally friendly, green and sustainable development of urban areas. Specifically, with the help of environmental standards, digital finance fosters the growth of businesses that safeguard the environment by accurately matching data. Any enterprise needs financial support to realize healthy development, and so do environmental protection enterprises. The use of AI, large-scale information, blockchain, online computing, fingerprints, and various other modern technologies allow financial services to precisely identify a firm’s attributes [16], due to the influence of environmental rules, banks are now placing more emphasis on the environmental advantages of firms, strengthening the financial support for enterprises in the field of green, to lessen the pollution of the land element, and offering finance options for businesses engaged with ecology and natural management while limiting the funding of heavily polluting companies [36] in order to enhance land use efficiency and minimize carbon emissions during land use, achieving sustainable use of land resources. In addition, it also can reduce the scope of the general congestion diseconomies, thereby improving the quality of life and positively affecting economic growth [37,38].

Optimizing and updating urban industrial structures using digital finance minimizes land element pollution. Digital finance has the potential to facilitate the modernization and organization of the commercial framework. Firstly, relying on the advantage of the low marginal cost of modern information technology, digital finance can amplify the scale effect of data elements, break the market barriers and spatial limitations, realize the efficient flow of information, guarantee the mobility and rate of return of factors of production, such as labor, capital, data, etc., optimize the distribution of factors of production among different industries, realize the updating of traditional industries, and make the urban industrial structure more reasonable. Secondly, the universality and convenience of digital finance can more accurately allocate capital factors and services to industries (such as high-tech industries) with low production costs, high productivity, low pollution outputs, and assistance with the growth of the technology sector, and this kind of enterprise also meets the preference of land investors, attracts land investors to invest in the land of this kind of industry, and creates a benign cycle of high-tech industry development, acknowledging the growth of urban improvement and refinement of the industrial framework [39]. By directing the rehabilitation of traditional sectors and supporting the expansion of high-tech firms, the use of technological financing may enable the improvement and automation of the city’s economic framework, inhibit the emission reduction of pollutants and reduce the pollution of land elements, thus enhancing the output benefits of the city’s overall green land use and reducing the negative impacts brought by non-desired outputs. On the other hand, industrial structure transformation provides preconditions for the proliferation and development of digital finance. An optimal and sophisticated industrial structure facilitates the integration and effectiveness of digital technologies, promotes the interactive integration of data elements and traditional elements, can more accurately and quickly identify the nature of enterprises, and can better carry out the function of digital finance on enterprise supervision. Specifically, the optimization and upgrading of industrial structure and the wide application of enterprise intelligent manufacturing technology provide convenience for the expansion of digital finance and technology integration, and can obtain corporate information more quickly and provide financial services to enterprises more accurately [40]. Thus, applying the favorable effect is more advantageous for financial technology. In conclusion, the application of technology finance could enhance and optimize urbanized commercial structures. Furthermore, cities with advanced and well-planned factories are more affected by digital money in terms of the environment. Therefore, maximizing the manufacturing sector in each city affects how well the financial system functions to raise the productivity of exploitation of environmentally friendly metropolitan areas. So, this study proposes the following hypothesis:

H1. 

The digital financial services can significantly boost the effectiveness of green urban land usage.

H1a. 

The environmental regulation may play a positive regulatory role.

H1b. 

The optimization of industrial structure may play a positive regulatory role.

2.2. Heterogeneity Effect Analysis

There is a significant discrepancy in China’s progress with digital financial services. The variety of urban development level influences how electronic funding influences the uptake of sustainable growth in these areas. Eastern cities have more developed economies and convenient access to transportation. However, the large amount of human capital that has been amassed as a result of the widespread migration to these eastern cities offers a strong basis for the development of digital banking. Therefore, digital banking is crucial for promoting a green attribute and effectively expanding the utilization of city land resources in the eastern area.

Similarly, the industrial structure of the central and western regions is mostly dominated by labor-intensive industries, while the population density is low and the ecological environment is fragile, coupled with the backwardness of the economic level compared with the eastern cities. Thus, financial technology may have far less effect on how well sustainable land is used in cities [41]. Moreover, since the financial, interpersonal, and environmental context in the southern and northern regions differs, there may be disparities in the influence of electronic financing on the efficacy of using city spaces for environmental purposes. Furthermore, there is a chance that the impact of the computerized financial system upon the efficacy of green land utilization in urban areas could vary. The digital financial structure mainly includes three aspects, specifically, the extent of coverage, the usage rate, and the extent of modernization; the last two pertain primarily to the financial sector’s radiation potential. The level of utilization indicates the extent to which the variety of digital financial goods and consumers are combined, while the level of digitization refers to the ease of digital financial products and services. The effectiveness of ecological development in metropolitan settings may be impacted differently by various online financing representations. Land utilization for green purposes is expected to be much more efficient in cities with a significant number of high-tech and service sectors and less reliance on main resources. These cities frequently possess a great ability to use enormous amounts of data, providing a solid hardware base for the rise of digital financial services. Conversely, resource-driven communities that depend on the exploitation of natural and mineral resources could be affected by variations in the extraction and utilization of these assets, which would affect their green plans for land use function. Resource-rich cities, often defined by a reliance on forests, minerals, and other natural resources, face difficulties when it comes to making green and sustainable urban land use a priority because of a phenomenon known as the “resource curse.” Furthermore, a route dependence that impedes the most environmentally beneficial use of land assets may result from the high concentration of secondary firms in these resource-based communities. Accordingly, this paper sets the following proposed hypotheses:

H2. 

The effect of internet banking on the usage of green space in the city differs based on urban development level difference, financial framework, and resource endowment.

2.3. Examining the Moderation Effect of Innovations in Green Technologies as the Mechanism of Action

Xiongbit’s innovation theory points out that financial development has a substantial driving impact on innovation [42]. Digital finance primarily impacts the efficient exploitation of urban land for green purposes via technical innovation in two primary areas: On one side, online finance utilizes big data technologies, is able to deal with huge amounts of data at low cost [43], and may readily overcome the constraints of time and location on the data transmission of innovative green technology to lessen the challenge of green technological advancement, and stimulate the initiative of green technology innovation. The emphasis is placed on advancing and protecting the environment via the optimization of manufacturing equipment and processes that are highly polluting and energy-intensive, and subsequently, decreasing energy use and minimize pollutant discharges. Green technology focuses on the advanced and environmentally friendly technology and equipment, optimizes highly polluting and energy-intensive production equipment and production processes, and minimizes energy consumption and pollutant emissions, while enhancing the effectiveness of eco-friendly land use within urban areas. Conversely, digital finance employs data analysis, intelligent investing, and other methods to lower the barrier to entry in the market, alleviate information asymmetry and moral risks, provide diversified financing channels for enterprises [44], facilitate the resolution of financial challenges throughout the course of technological advancement, and motivate businesses to engage in the development of environmentally friendly technologies [45], produce green and clean products, and ultimately accomplish efficient and ecological usage of city land. To summarize, this study presents the following hypothesis:

H3. 

Digital finance achieves green urban land use efficiency by promoting green technology innovation.

3. Study Design

3.1. Description of Variables and Data

The study presented here analyzes 281 Chinese cities as its subject. In addition, considering that although digital finance budded in the 1990s, rapid development started late, this paper selects 2010–2020 as the research time period, and some of the missing data are made up by the linear interpolation method.

(1)

Described Factors

Urban Land Green Utilization Efficiency (ULGUE): this study constructs an evaluation index system using resource inputs, intended outputs, and non-desired outcomes as the primary indicators [18,19,20], and the data are all from the China Urban Statistical Yearbook and China Statistical Yearbook of past years. The details are shown in

Table 1. Evaluation index system of urban land green utilization efficiency.

Level 1 Indicators	Secondary Indicators	Tertiary Indicators
Resource inputs	Land factor inputs	Construction site area
	Labor factor inputs	Total labor force + public employees in the labor force
	Capital factor inputs	fixed-asset investment
Expected outputs	Economic output	Gross Domestic Product (GDP)
	Social outputs	Average wage of employees
	Environmental outputs	Integrated solid waste utilization rate
		Centralized sewage treatment rate
		Non-hazardous treatment rate of domestic waste
Non-expected outputs	Environmental pollution	Industrial smoke emissions
		Industrial wastewater discharge
		Industrial sulfur dioxide emissions

.

The majority of the current research uses the farthest distance model to the frontier (SBM) as a means of quantifying efficiency; because the SBM model measures the value between 0 and 1, there will be a series of problems in the subsequent econometric analysis process, thus affecting the precision of the experimental results. The current study uses various dimensions, such as desired and non-desired outputs, to address the drawbacks of the SBM model, and the precise green consumption efficacy of urban plots and structures was measured using the Super-SBM model, and the specific model formulas are as follows:

$$\begin{gathered} min\rho =\frac{1+\frac{1}{m}{\sum }_{i=1}^m\frac{s_i^{-}}{x_{ik}}}{1-\frac{1}{q_1+q_2}\left({\sum }_{r=1}^{q_1}\frac{s_r^{+}}{y_{rk}}+{\sum }_{t=1}^{q_2}\frac{s_t^{b-}}{b_{rk}}\right)} \\ s.t.{\sum }_{j=1,j\ne k}^n{x}_{ij}{\lambda }_j-s_i^{-}\le x_{ik} \\ {\sum }_{j=1,j\ne k}^n{y}_{rj}{\lambda }_j+s_r^{+}\ge y_{rk} \\ {\sum }_{j=1,j\ne k}^n{b}_{tj}{\lambda }_j-s_t^{b-}\le b_{tk} \\ 1-\frac{1}{q_1+q_2}\left({\sum }_{r=1}^{q_1}\frac{s_r^{+}}{y_{rk}}+{\sum }_{t=1}^{q_2}\frac{s_t^{b-}}{b_{rk}}\right)>0 \\ \lambda ,s^{-},s^{+}\ge 0 \\ i=\mathrm{1,2},\cdots ,m;r=\mathrm{1,2},\cdots ,q;j=\mathrm{1,2},\cdots ,n(j\ne k) \end{gathered}$$

(1)

There are $n$ decision unit $DMU$s to be evaluated for efficiency, denoted by ${DMU}_j\left(j=\mathrm{1,2},\cdots ,n\right)$; each decision cell has $m$ inputs and $q$ outputs. $\rho$ is the efficiency value of the model, ${ x}_{ik}$ denotes the input variables of the model, ${ y}_{rk}$ denotes the model’s desired output variable, ${ b}_{rk}$ indicates the non-expected output variables, $\lambda$ is the decision making units, and the slack variables are represented by $s_i^{-}$, $s_r^{+}$, and $s_t^{b-}$, where a value exceeding 0 means that there is room for optimization of the efficiency value. In this paper, MaxDEA8.17 is used to calculate the urban land green use efficiency index using the current frontier, non-oriented, and super-efficiency model.

(2)

Key explaining variables

Digital finance (df). Three factors constitute the electronic financial indices: the rate of electronic records, cover breadth, and extent of usage. It is frequently employed in modern research and offers a thorough and impartial review of the overall state of advancement in financial technology. The “Peking University Digital Inclusive Finance Index (2011–2020)” was used as a stand-in for the primary explanatory variables in this study. To ensure the correctness of the regression results, the data are uniformly compressed by a factor of 100 for processing, and linear interpolation is used to deal with inaccurate data.

(3)

Mechanism variables

Green technological innovation (Gin): Existing literature often uses R&D investment or the number of patents to reflect the technological innovation capacity of cities. China’s official statistics already have relevant data specifically for “green” achievements. In this paper, we use the original data of green patents obtained by prefecture-level cities to characterize the level of green technological innovation in cities, and at the same time, we shrink the data by 1000 times. The data come from the CNRDS database [46].

(4)

Moderating variables

(1)

Environmental regulation (Er): The existing literature on environmental regulation alternative indicators for prefecture-level cities mostly uses pollutant removal rate characterization. Considering the correlation with land elements, this research use the complete utilization rate of industrial solid waste as a measure to assess the level of stringency in environmental regulation.

(2)

Advanced industrial structure (Ais): The index of advanced industrial structure can specifically reflect the state of development of the three major industries in different regions as well as the distribution results. In the index is a positive indicator, that is, as the value increases, it proves that the degree of optimization and development of industrial structure increases significantly. This article examines the correlation between the share of the three main industries and several other factors; the Tel index is used to calculate the index of advanced industrial structure.

(5)

Control variables

The following control variables are selected in this paper: (1) the degree of openness to the outside world (Fdi), for which the amount of foreign investment used in the current year is taken as a logarithmic characterization; (2) urban infrastructure (Utf), characterized by the road area per capita; (3) city size (LnDop), using population density to take a logarithmic view to reflect the city size; (4) the degree of government intervention (Gov), defined by the ratio of government public spending to GDP; (5) market activity (Market), defined by the logarithm of aggregate sales of consumer goods; and (6) urban education level (Edu), which is the ratio of the number of university students to the total resident population.

3.2. Modeling

(1)

Baseline linear model

To evaluate Hypothesis H1, the baseline linear model that follows is developed in the current study:

$${Ulgue}_{c,t}=\alpha +{\beta }_1{Df}_{c,t}+{\beta }_2{Controls}_{c,t}+{\gamma }_c+{\delta }_t+{\epsilon }_{c,t}$$

(2)

In the above model, index “$c\mathrm{”}$ denotes the city, index “$t\mathrm{”}$ is the time, and ${Ulgue}_{c,t}$ represents the time of the $\mathrm{c}$ city’s land green use efficiency in the year’s land green utilization efficiency, ${Df}_{c,t}$ represents the $c$ city’s digital financial development index in the year, ${Controls}_{c,t}$ represents a series of control variables, ${\gamma }_c$, and ${\delta }_t$ represent individual and time fixed effects in the model, respectively, and ${\epsilon }_{c,t}$ are randomized disturbance terms.

In order to further verify the environmental regulation and the industrial structure advanced adjustment effect, this study presents the incorporation of an interaction term into the benchmark model. The particular model is outlined as follows:

$${Ulgue}_{c,t}=\alpha +{\beta }_1{Df}_{c,t}+{\beta }_2{Rv}_{c,t}+{\beta }_3{Df\ast Rv}_{c,t}+{\beta }_2{Controls}_{c,t}+{\gamma }_c+{\delta }_t+{\epsilon }_{c,t}$$

(3)

In the above model, ${Rv}_{c,t}$ represents $c$ cities in that year, the moderating variables represent environmental regulation and industrial structure advancement, respectively. ${Df\ast Rv}_{c,t}$ represents the $c$ city’s and the year’s modifying factors and electronic financial association term, and the other variables’ interpretations align with the baseline regression model.

(2)

Mechanism testing

Based on the prior theoretical study, the environmental breakthrough mediated effect concept is built for studying the potential impact of online financing upon the efficient execution of urban land for sustainable objectives. The specific design is the one shown below:

$${Ulgue}_{c,t}={\beta }_1{Df}_{c,t}+{\beta }_2{Controls}_{c,t}+{\gamma }_c+{\delta }_t+{\epsilon }_{c,t}$$

(4)

$${Gin}_{c,t}={\beta }_1{Df}_{c,t}+{\beta }_2{Controls}_{c,t}+{\gamma }_c+{\delta }_t+{\epsilon }_{c,t}$$

(5)

$${Ulgue}_{c,t}={\beta }_1{Df}_{c,t}+{\beta }_2{Gin}_{c,t}+{\beta }_3{Controls}_{c,t}+{\gamma }_c+{\delta }_t+{\epsilon }_{c,t}$$

(6)

In the above model, ${Gin}_{c,t}$ represents $c$ city’s green technology innovation in the year, and the rest of the variable meanings are consistent with the benchmark regression model.

4. Empirical Analysis

4.1. Descriptive Statistics

When combined with the information shown in

Table 2. Variable statistics for descriptive purposes.

Variable Name		Sample Size	Average Value	Standard Deviation	Inimum Value	Maximum Values
Explanatory variable	Ulgue	3091	0.51	0.385	0.013	2.66
Core explanatory variables	df	3091	1.604	0.7974	0.0001	3.3448
Moderator variable	er	3091	0.7960	0.2304	0.0024	1
	ais	3091	6.522	0.3534	5.3108	7.8361
Mechanism variables	gin	3091	0.1058	0.4207	0	10.01
Control variable	fdi	3091	9.7962	2.3644	−4.6378	15.3078
	lndop	3091	5.738	0.939	0.683	7.882
	utf	3091	1.381	1.258	0.021	35.635
	edu	3080	0.0173	0.0198	1.63 × 10−6	0.1276
	market	3079	15.5408	1.0592	12.1556	18.8865
	gov	3091	0.1998	0.1049	0.0439	1.4852

, the findings indicate that the computed mean amount of green usage productivity of the cities is 0.51, with the highest value being 2.66, the lowest being 0.013, and the standard deviation being 0.385. The financial mean is 1.604, with a standard deviation of 0.7974. The two extremes are 0.0001 and 3.3448, respectively.

4.2. Benchmark Regression Results

In this paper, the bidirectional fixed effect model is used to estimate the benchmark model by adding the standard error of clustering file and clustering to the city level. Even in the absence of control variables,

Table 3. Benchmark model and moderating effect test results.

Benchmarking Models and Moderating Effects
	(1)	(2)	(3)	(4)
df	0.1492 ***	0.1598 ***	0.1369 ***	0.1469 ***
	(0.0500)	(0.0390)	(0.0456)	(0.0407)
er	-	-	0.1670 ***	-
	-	-	(0.0333)	-
er*df	-	-	0.0975 ***	-
	-	-	(0.0371)	-
ais	-	-	-	0.1318 **
	-	-	-	(0.0586)
ais*df	-	-	-	0.0208
	-	-	-	(0.0275)
edu	-	−2.3375 *	yes	yes
	-	(1.3491)
fdi	-	−0.0083
	-	(0.0053)
lndop	-	−0.0650
	-	(0.0557)
utf	-	−0.0079 **
	-	(0.0034)
market	-	0.0106
	-	(0.0384)
gov	-	0.1353
	-	(0.1324)
_cons	0.4254 ***	0.7376	0.8063	0.0917
	(0.0130)	(0.6555)	(0.6505)	(0.7314)
sample size	3091	3091	3091	3091
Individual fixation	yes	yes	yes	yes
fixed time	yes	yes	yes	yes
Adj-R2	0.0078	0.1208	0.1948	0.1545
F-statistic	34.11 ***	23.18 ***	23.18 ***	20.89 ***

Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

of Model (1) shows a strong positive link between the growth of online finance and the effective use of city space for green initiatives. This offers compelling proof that the two are positively correlated. Moreover, even with the introduction of control variables, Model (2) in Table 3 retains the direction of consequences of online banking and the significance of the regression coefficient. In particular, the adoption of digital financing results in a 1% improvement, which adds up to a significant 15.98% increase in the effectiveness of using urban land for green purposes. The consistent outcomes of Models (1) and (2) support the favorable and encouraging association between digital finance and the effective use of urban space for environmentally friendly endeavors. Some significant findings are revealed by the control variable results. At the 10% level, the urban education level (edu) is markedly negative. The reason may be that the advanced education level attracts foreign population, which increases the burden of urban environment to a certain extent and is not conducive to the improvement of the green use efficiency of urban land. Furthermore, at the 5% significant level, the urban infrastructure (Utf) shows a significantly negative association, suggesting that the growth of urban transportation infrastructure will limit the improvement of urban land green utilization efficiency.

This study adds interaction terms between the primary explanatory and moderating variables for empirical testing to better assess whether environmental law and industrial structure optimization have an impact on online finance and the efficacy of urban land green utilization. The results, which are displayed in Table 3 for Models (3)–(4), demonstrate that the impact on environmental legislation is favorable in Model (3), suggesting that it aids in encouraging the effective use of green space on city land. Environmental legislation has a positive effect on digital banking and increases the success of using cities for green initiatives, as shown by the significant linear effect observed in the interaction term between the two variables. Although the regression coefficient of industrial structure optimization passed the significance test, the interaction term did not pass the significance test. Therefore, the regulatory effect of industrial structure optimization on the improvement of urban land green use efficiency by digital finance has not yet formed. Hypothesis H1 and hypothesis H1a are verified, but Hypothesis H1b is not verified.

4.3. Robustness Tests

(1)

Replacement of explanatory variables: This research uses the entropy value approach to reevaluate the ecological efficiency of urban land usage, and the regression results are shown in

Table 4. Robustness test.

Variant	Robustness Check				Endogeneity Test
	(1)	(2)	(3)	(4)	(5)
df	0.0486 *	0.0827 ***	0.1409 *	0.1504 ***	0.5419 **
	(0.0278)	(0.0100)	(0.0847)	(0.0410)	(0.2267)
control variable	yes	yes	yes	yes	yes
_cons	−0.0230	2.4028 ***	0.2003 **	0.7351	1.6109 ***
	(0.0542)	(0.1698)	(0.5332)	(0.6567)	(0.5270)
sample size	3091	3091	3019	3035	3091
Individual fixation	yes	no	yes	yes	yes
fixed time	yes	no	yes	yes	yes
weak instrumental variables test (WIVT)	-	-	-	-	31.15
Adj-R2	0.2089	0.1957	0.1797	0.1209	0.7465
F-statistic	31.72 ***	73.67 ***	18.46 ***	22.44 ***	-

* p < 0.1, ** p < 0.05, *** p < 0.01.

for Model (1). The regression results demonstrate that the impact of digital finance aligns with the findings of the benchmark model, confirming the reliability of the benchmark regression model.

(2)

Replacement of estimation method: the benchmark model is re-regressed using the OLS estimation method, and the results of regression Model (2) in Table 4 are consistent with the results of Models (1) and (2), indicating that the benchmark model is robust.

(3)

Truncated treatment of the explanatory variables: In order to avoid extreme values affecting the results of regression analysis, this paper truncates the explanatory variables. It is evident based on the regression findings of Model (3) in Table 4, that it demonstrates the resilience of the benchmark regression model when it aligns with the benchmark regression outcomes.

(4)

Excluding municipality samples: The present research removes the municipality that is directly under the control of the central government from the sample data to further assure the consistency of the estimated findings and the reliability of the conclusions. Table 4 for Model (4) displays the results, with only the city samples at the prefecture level being retained. The results show that the direction and value of the role of online financing on the efficacy of green land usage in cities are consistent with the findings of the benchmark model, which also shows the resilience of the relevant conclusions drawn by the benchmark regression model and that the size of the city has little effect on the estimation results.

(5)

Endogeneity problem: Given the issue of endogeneity resulting from reverse causality, this study used the number of mobile phone users at the end of the year as the instrumental variable for the digital finance index, and at the same time, its secondary and tertiary terms were used as the instrumental variables of the secondary and tertiary terms of the digital finance development index for the two-stage least squares regression (2SLS). First, there is a need for a correlation between significant variables, and the rapid popularization of mobile terminals and the rise of digital financing cannot be separated; second, there is no direct correlation between the growing popularity of cell phones and the green use of urban land, which satisfies the requirement of homogeneity of instrumental factors. The regression findings of Model (5) in Table 4 show that the direction of the effect of the number of cellphone users at the end of the year on the efficacy of urban land green utilization is consistent with the benchmark model. Furthermore, the F value of 31.15 in the weak instrumental variable test supports the validity of the empirical variable, above the critical value of 10. The effective use of urban areas for green initiatives could profit from the presence of online money, even after accounting for the endogeneity problem.

4.4. Heterogeneity Analysis

(1)

Financial structure heterogeneity

To investigate the precise impact of the extent of coverage, intensity of use, and degree of digitalization of digital finance on the efficiency of environmentally friendly land use in metropolitan regions, the three dimensions of digital finance were separately regressed and analyzed, and the results are shown in

Table 5. Results of the test for structural heterogeneity in digital finance.

	(1)	(2)	(3)
cover	0.0016	-	-
	(0.0013)	-	-
use	-	0.0020 ***	-
	-	(0.0008)	-
digit	-	-	0.0011 ***
	-	-	(0.0003)
control variable	yes	yes	yes
_cons	0.8428	0.8596	0.8582
	(0.5769)	(0.5601)	(0.5744)
sample size	3091	3091	3091
Individual fixation	yes	yes	yes
fixed time	yes	yes	yes
R2	0.1321	0.1324	0.1091
F-statistic	21.94 ***	22.10 ***	21.44 ***

*** p < 0.01.

.

It can be seen that both the depth of use and the degree of digitization pass the significance level test, while the coverage breadth fails the significance test. It shows that digital finance can affect the green use efficiency of urban land mainly through the depth of use and the degree of digitization, among which the depth of use has the greatest impact. The depth of use reflects the diversity of digital financial products and the degree of integration with users. According to the regression results, digital finance has an impact on the green use efficiency of urban land through product diversity and deep integration with users. The influence of the digitization degree on the green use efficiency of urban land is relatively small, which may be due to the large gap in the level of digital development in China and the imbalance of digital infrastructure construction, leading to the weak influence of the digitization degree on the green use efficiency of urban land. The coverage breadth did not have an impact on the green use efficiency of urban land, which may be due to the differences in the radiation capacity of digital finance among cities. On the other hand, the regression results show that digital finance has financial structure heterogeneity in urban land green use efficiency.

(2)

Heterogeneity of urban locations

The previous benchmark regression findings indicate that digital finance may greatly enhance the efficiency of using urban land for green purposes. However, considering the fact that the socio-economic characteristics differ greatly between the north and south regions, and the east and west regions, this paper further develops heterogeneity analysis of regional locations (According to the National Bureau of Statistics for China’s regional division, the eastern region includes: Beijing, Tianjin, Hebei, Liaoning, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Hainan. The central region includes: Shanxi, Anhui, Jiangxi, Henan, Hubei, Hunan. The western region includes: Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang. The northern region includes: Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shandong, Henan, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang. The southern region includes: Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Hubei, Hunan, Guangdong, Guangxi, Hainan, Chongqing, Sichuan, Guizhou, Yunnan). The regression results are shown in

Table 6. Results of the test for heterogeneity of urban locations.

	Eastern City	Central Cities	Western Cities	Southern Cities	Northern Cities
df	0.1838 ***	0.2933	−0.0737	0.2175 ***	−0.1098
	(0.0339)	(0.1804)	(0.1105)	(0.0366)	(0.1178)
control variable	yes	yes	yes	yes	yes
_cons	4.7554 ***	−1.4899 ***	0.4113	1.5406	0.8791
	(1.5578)	(0.5572)	(0.7693)	(1.4226)	(0.7390)
sample size	1309	869	913	1672	1419
Individual fixation	yes	yes	yes	yes	yes
fixed time	yes	yes	yes	yes	yes
Adj-R2	0.0234	0.0800	0.0101	0.0135	0.2552
F-statistic	14.24 ***	12.45 ***	10.80 ***	19.41 ***	17.23 ***

*** p < 0.01.

.

First, from the perspective of east–west location division, after the urban sample is divided into eastern, central, and western regions, for the eastern cities, the digital finance coefficient passes the significance test; however, it has no significant effect on the central and western cities. The results show that the impact of digital finance on urban land green use efficiency mainly occurs in eastern cities. Compared with western cities, eastern cities and central cities have a higher economic development level and population agglomeration degree. However, in central and western cities, digital finance plays a more “lubricant” role for local development. Coupled with low population density and a fragile ecological environment, its impact on the green use efficiency of urban land is not obvious.

Regarding the north–south divide, the southern cities’ financial influence on the cities’ green usage efficiency has a statistically positive correlation at the 1% level, whereas the northern cities’ digital finance impact fails to reach the threshold for a significant criterion. This implies that in southern cities, online banking mostly improves the efficacy of green urban land use. The influence on urban land green use efficiency in the southern region may be explained by the difference in the growth rate of digital finance between the southern and northern regions.

(3)

Resource endowment heterogeneity

Two sorts of cities are distinguished in the National Sustainable Development Plan for Resource-Based Cities (2013–2020): “resource-based cities” and “non-resource-based cities”.

Table 7. Results of the test for heterogeneity of urban resource endowment.

	Resource-Based City	Non-Resource-Based Cities
df	0.1095	0.1523 ***
	(0.1292)	(0.0427)
control variable	yes	yes
_cons	1.1801	0.0248
	(0.7549)	(0.8737)
sample size	1243	1848
Individual fixation	yes	yes
fixed time	yes	yes
Adj-R2	0.2604	0.418
F-statistic	12.50 ***	16.38 ***

*** p < 0.01.

presents the particular regression results. In “non-resource-based cities”, the estimated coefficient of digital finance is significantly positive at the 1% level, indicating that the development of digital finance in “non-resource-based cities” can significantly improve the efficiency of urban land green use. However, the regression coefficient of “resource-based city” failed to pass the significance level test, indicating that “resource-based city” digital finance failed to have an improving effect on urban land green use efficiency. In light of city conditions and access to resources, this study highlights the variability in the financial structure and the influence of electronic finance on enhancing the effectiveness of using urban areas for green initiatives, which verifies Hypothesis H2.

4.5. Mechanism Testing

The earlier empirical work has already shown that digital finance can influence how effectively green space is used in urban areas; nevertheless, it was unable to determine how digital finance increases the effectiveness of green space use. This research employs the mediating effect model, uses green technology innovation (Gin) as a mechanism variable, and conducts an empirical Sobel test to validate the previous Hypothesis H3.

Table 8. Estimated results of intermediation effects.

	Ulgue	Gin	Ulgue
df	0.1596 ***	0.3622 ***	0.1391 ***
	(0.0390)	(0.0362)	(0.0349)
gin	-	-	0.0566 ***
	-	-	(0.0180)
control variable	yes
p-value	0.003
z-value	3.002
Individual fixation	yes	yes	yes
fixed time	yes	yes	yes
_cons	1.2709 ***	5.2566 ***	0.9734 **
	(0.3759)	(0.3967)	(0.3870)
N	3091	3091	3091
Adj-R2	0.7308	0.7996	0.7317
F-statistic	29.10 ***	42.30 ***	29.12 ***

** p < 0.05, *** p < 0.01.

displays the regression analysis’s findings. The first column in Table 8 shows that digital finance can significantly contribute to the efficiency of green urban land use, which satisfies the precondition of the mediation effect. The regression findings in the second column indicate that the coefficient for green technology innovation and digital finance is considerably positive, it is important to note that digital finance plays a crucial and beneficial role in advancing green technology innovation. The factor has a mediating influence since, as the third column demonstrates, both green innovation in technology and financial technology are important at the 1% level. Lastly, the Sobel test results, which indicate that the p value is below 0.003, and the Z score is 3.002, imply that online finance may improve the effectiveness of using land in cities for ecological purposes by promoting technological innovation in green practices, which verifies Hypothesis H3.

5. Conclusions of the Study and Recommendations for Countermeasures

5.1. Conclusions of the Study

The present research examines empirically the link between online financing and the effectiveness of ecologically friendly city land utilization. Using panel data, it explicitly examines the impact of virtual finance on the effectiveness of using city land for green purposes. The mechanism of the two variables is examined using the mediation effect model, which produces the following main findings:

(1)

Within the research period, the regression results of the benchmark model show that digital finance can have an impact on the green use efficiency of urban land. Specifically, whether or not control variables are added, digital finance has a significant promoting effect on the green use efficiency of urban land. After the addition of control variables, the digital finance index increases by 1%, and the green use efficiency of urban land increases by 15.98%. And the model passed a series of robustness tests. In addition, it is found that digital finance can enhance the effect of urban land green use efficiency under the regulatory effect of environmental regulations, while industrial structure optimization has not yet played a regulatory effect.

(2)

The heterogeneity regression results of digital financial structure show that the depth and digitization degree of digital finance have a significant impact on the green use efficiency of urban land, while the coverage breadth has no impact on the green use efficiency of urban land. The regression results of urban location heterogeneity show that the promotion effect of digital finance on urban land green use efficiency mainly occurs in eastern cities and southern cities, while the effect of digital finance on western cities and northern cities is not obvious. The results of resource endowment heterogeneity show that the effect of digital finance on improving the green land use efficiency in “non-resource-based cities” is stronger than that in “resource-based cities”.

(3)

The mechanism test’s outcomes highlight a significant mediating effect among digital finance, green technology innovation, and urban land green utilization efficiency. It shows that digital finance can take green technology innovation as a path to influence the green utilization efficiency of urban land.

5.2. Development Responses and Recommendations

(1)

Continue promoting the transformation of traditional finance into digital finance. Digital finance is conducive to the transformation of China’s urban land use to a green and sustainable development approach. It further contributes to enhancing the efficiency of green land use, and ultimately ensures that the expected goals of carbon peaking and carbon neutrality are attained. In this regard, governmental entities and relevant stakeholders should energetically promote digital finance. This entails prioritizing infrastructure investments, perpetually refining and optimizing digital finance, and maximizing the advantages and efficacy inherent in digital financial systems.

(2)

Strengthening the influence of key elements on the green utilization of urban land is crucial for maximizing the positive contributions of digital finance. It is necessary to further optimize cities’ industrial structure, strengthen the synergy between cities’ digital facilities, environmental regulation, city scale, and the development of digital finance, and provide a favorable environment for the development of digital finance. The optimization of urban industrial structure, urban population aggregation, urban economic development, and digital infrastructure construction should be continuously promoted through policy guidance. In terms of urban economy, digital finance should be combined with environmental policies to jointly guide enterprises towards green innovation, facilitating environmental sustainability throughout the production and consumption chain and, consequently, promoting the green utilization of urban land. Regarding urban scale, progressive policies should be devised to attract talent, gradually easing restrictions on urban registration to enhance the influx of high-quality individuals to cities. This approach aims to harness the demographic advantages associated with urban scale effectively. Through the coordination of key elements and digital finance, the green use efficiency of urban land is improved, and the sustainable use of land resources is realized.

(3)

Utilizing digital finance to promote technological innovation and application. Digital finance can provide financial support for low-carbon cutting-edge technologies, such as carbon capture, storage, and utilization, so as to realize the widespread adoption of advanced technologies such as renewable energy, energy conservation, and emission reduction, in the promotion and implementation of urban land use, avoid pollution of land resources. Ultimately, this ensures a continuous reduction in carbon emissions and further enhances the efficiency of green land use.