Research on the Impact of Digital Inclusive Finance on Green Innovation of SMEs

Abstract

Green innovation is an effective driving force for high-quality development in the new era. As a new financial service model, digital inclusive finance provides a new way to solve the financing dilemma of green innovation. In order to investigate the impact of digital financial inclusion on the green innovation of small and medium-sized enterprises (SMEs), based on the panel data of SMEs in China from 2011 to 2021, this paper empirically tested the influence and mechanism of the two by using the panel fixed effect model and threshold regression model. The findings demonstrate that digital inclusive finance is effective in promoting the green innovation of SMEs and alleviates their financing constraints. The digitization level of digital inclusive finance also has a significant positive and non-linear effect of increasing “marginal effect” on the green innovation of SMEs. Notably, it has a greater significant role in driving green innovation for non-state-owned enterprises, enterprises located in the central and western regions, and SMEs with dispersed equity. Consequently, several policy implications are derived from these findings, which can provide a sustained impetus to help SMEs carry out green innovation activities and also provide a scientific basis for governments to improve digital financial inclusion policies and achieve economic equity.

1. Introduction

Green innovation is the organic combination of the two new development concepts of green development and innovation-driven development, which can effectively coordinate the relationship between economic growth and ecological environmental protection [1], and is also an important means for enterprises to obtain competitive advantages. Under the guidance of the strategic goal of “carbon peak and carbon neutral”, green innovation is different from traditional innovation, paying more attention to the symbiosis of economic and environmental values, aiming to promote the sustainable development of the economy, ecology, and society, and becoming an effective driving force to lead the high-quality development in the new era [2]. However, green innovation activities, which are characterized by high investment, high risk, and long-term sustainability, require long-term and stable capital investment [3]. Only internal liquidity will not be able to meet the needs of green innovation activities for enterprises, especially small and medium-sized enterprises (SMEs), which have natural weaknesses. Additionally, smaller enterprise scale and limited collateral further exacerbate financing constraints [4,5]. As a result, solving financial problems is one of the key issues for SMEs to carry out green innovation activities [6,7,8].

With the continuous development of information technology, digital inclusive finance, a new model of financial services, has gradually emerged, and it provides a new way to solve the financing dilemma of corporate green innovation [9]. Digital financial inclusion is a new business form and business model, an action to use digital financial services to promote financial inclusion, and also the fourth stage of the financial revolution after microcredit, microfinance, and inclusive finance [10]. Different from traditional finance, digital inclusive finance pays more attention to technology and has many advantages, such as “wide coverage, low cost, and high efficiency” [11]. Relying on big data and popularity, digital inclusive finance can break the limitations of traditional financial institutions in space and information transmission [12,13], reduce the cost and threshold of financial services, help to alleviate the financing obstacles faced by small and medium-sized enterprises (SMEs) in the process of green innovation, and bring new space for the development of SMEs’ green innovation [14]. However, it is a key issue of concern to make full use of digital inclusive finance to effectively solve the problem of SMEs’ green innovation activities stagnating due to the difficulty of financing and the high cost of financing for the government, academia, and the business sector.

Based on this, this paper takes SMEs as an example to scientifically and objectively evaluate the effects of different dimensions of digital inclusive finance on SMEs’ green innovation and carries out an in-depth study on the relationship between digital inclusive finance and SMEs’ green innovation, as well as the mechanism of its role, with a view to helping SMEs, as the main players in the market, to solve the problems of financing difficulties and lack of resources in the green innovation process and provide the sustained impetus for them to carry out green innovation activities, and at the same time, it can also provide a scientific basis for the government to improve the policy of digital inclusive finance and realize economic fairness.

The rest of this article is divided into five parts. Section 2 reviews the relevant domestic and foreign research status. Section 3 mainly puts forward the relevant hypothesis of this paper. Section 4 introduces the measurement methods and data sources of relevant variables. In the Section 5, descriptive statistical analysis, benchmark regression analysis, intermediary effect results analysis, robustness analysis threshold effect test, and heterogeneity analysis are carried out. Section 6 introduces relevant discussions of the empirical result analysis. Section 7 summarizes the conclusions based on the research results and corresponding policy implications and proposes the limitations of the research and future research directions.

2. Literature Review

Green innovation, a key element of the company to maintain environmental management [15,16], was first proposed by Fussler and James (1996), who considered it to be the production of new products using new technologies that bring economic value while reducing negative impacts on the environment [17]. Green innovation is not only important for achieving environmental sustainability and economic profitability [18,19,20] but also leads companies to gain competitive advantages [21,22]. Afterward, scholars conducted a lot of research on green innovation and currency. The definition of green innovation is interpreted from two main perspectives. One is process-oriented, which considers green innovation as a series of means adopted by enterprises in the production process to improve the greenness of products, including energy saving, consumption reduction, emission reduction, and improvement of management techniques [23,24]. The other approach is goal-oriented, which considers green innovation as a mode of technological innovation in which firms have the primary goal of improving environmental benefits [25]. However, regardless of whether green innovation is process-orientated or goal-orientated, its characteristics, such as quasi-public goods attributes and a high degree of uncertainty, have led to larger problems, such as long investment cycles and insignificant direct economic benefits for enterprises in the process of green innovation [26], and many enterprises have faced significant financing constraints due to the limitations of their own resources [27,28], thus making it difficult to carry out green innovation independently [29], leading to restrictions on the development of green production and green consumption.

As an important way to break the limitations of traditional financial institutions in terms of space and information transmission [30], reduce the cost and threshold of financial services [31], and empower the development of corporate innovation, the emerging concept of digital financial inclusion has received extensive attention from academics, but most of it focuses on the study of the mechanism of the impact of digital financial inclusion on corporate innovation. Scholars generally believe that digital inclusive finance mainly promotes enterprises to innovate in a variety of ways, such as easing financing constraints [32,33,34], enhancing the motivation of non-controlling shareholders [35], expanding market potential [36], mitigating information asymmetry [37], improving the efficiency of loan approval by financial institutions [38], alleviating resource misallocation, and incentivizing enterprises to increase R&D investment [5]. Some scholars have also studied it from a spatial perspective and found that the development of digital inclusive finance not only improves the level of innovation in the local cities but also drives the innovation capacity of neighboring cities, and its innovation incentive effect will be characterized by increasing marginal returns as the level of regional economic development increases [39].

Although existing studies have explored the relationship between digital inclusive finance and corporate innovation in depth from a number of perspectives, it is not difficult to find that there is a lack of sufficient research in the area of the role of digital inclusive finance in promoting green innovation among SMEs. Based on the existing research results, the marginal contribution of this paper may lie in the following aspects: First, from the perspective of the digital level, this paper deeply discusses the non-linear relationship between digital inclusive finance and green innovation of SMEs, so as to develop new ideas for digital inclusive finance to better serve the green innovation of SMEs. Second, from the internal and external perspectives, this paper is committed to exploring the heterogeneity of the driving effect of digital inclusive finance on green innovation of SMEs with different natures, different regions, and different ownership structures so as to provide more targeted and scientific empirical evidence for digital inclusive finance to promote green innovation of SMEs. The research in this paper can not only explore a new perspective on how to carry out green innovation for SMEs but also provide an effective solution to the financing problems of SMEs and provide references for SMEs with different business natures, regions, and ownership structures to carry out green innovation with the help of digital inclusive finance.

3. Theoretical Analysis and Hypotheses

Digital inclusive finance, as a profound integration of financial revolution and technological revolution, serves as an important lever for enterprise green innovation [40]. The inherent advantages of digital inclusive finance, such as low threshold, high efficiency, and low cost, have disrupted the traditional finance service model, effectively bringing new room for development for SMEs with active green innovation projects and strong demand for loans in transition. Its specific roles are demonstrated as follows:

On the one hand, digital inclusive finance, as a new financing method [41], expands the financing channels for green innovation of SMEs. Serving as a powerful supplement to traditional finance [42], digital inclusive finance, leveraging technology applications in big data, cloud computing, and artificial intelligence fields, can break the time and space limitations of traditional financial services quickly while ensuring broad coverage, deep extension, and risk control [43]. By continuously lowering the financing threshold for green innovation entities, it opens up new financing channels for green innovation activities, thereby solving the problem of lack of financing channels for SMEs and ensuring the smooth implementation of green innovation projects for SMEs.

On the other hand, digital inclusive finance helps reduce the financing costs of green innovation for SMEs. The service process of traditional finance is lengthy and cumbersome, which can prolong the financing cycle and R&D cycle of green innovation activities, exacerbating the financing costs of green innovation for SMEs that already have low financing levels and increasing the uncertainty of their green innovation behaviors. Leveraging new-generation information technologies such as big data, artificial intelligence, and blockchain, digital inclusive finance can rapidly integrate and process massive data on a low-cost, low-risk, and low-threshold basis [8], effectively reducing the time costs for both borrowers and lenders and quickly resolving the information asymmetry constraints that traditional credit activities cannot avoid, thereby effectively reducing the financing costs of green innovation entities.

However, in the initial stage of digital technology construction, inclusive finance has a high initial cost, low integration degree of digital technology and traditional finance, and limited available data, resulting in low predictive analysis ability, low technology spillover effect, and ultimately weak promotion effect of digital inclusive finance on green technology innovation. With the continuous improvement of the digital level, the degree of informatization, networking, and intelligence of digital inclusive finance has been deepened [44], which has improved the efficiency of financial resource allocation [45], enhanced the participation of green innovation subjects, made up for the disadvantages of traditional finance in information collection, improved the efficiency of information transmission between lenders and borrowers, and reduced information asymmetry [46]. This will help financial institutions to effectively identify the relevant information on SMEs’ green innovation, reduce their “adverse selection”, enhance the external financing capacity of SMEs’ green innovation, and enable more green innovation entities to fully enjoy the spillover dividends of green innovation on a larger scale and at a lower cost, thus greatly promoting their enabling role in green innovation. Therefore, the level of digitization may contribute differently to green technologies at different stages of development. Based on this, the following hypotheses are proposed:

Hypothesis H1.

The development of digital inclusive finance significantly promotes the level of green innovation for SMEs.

Hypothesis H2.

Digital inclusive finance promotes green innovation for SMEs by alleviating financing constraints.

Hypothesis H3.

The digitalization level of digital financial inclusion has a non-linear impact on the green innovation of SMEs.

4. Research Design

4.1. Model Establishment

By referring to the test method of [47], the following benchmark regression model is established to explore the relationship between digital financial inclusion and green innovation of SMEs and to verify hypotheses H1:

$${GI}_{ijt}={\alpha }_0+{\alpha }_1{DIF}_{jt}+{\alpha }_3{controls}_{ijt}+{\mu }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(1)

In order to further clarify the impact of digital inclusive finance on green innovation for SMEs, that is to verify the hypothesis H2, the intermediary effect model of [48] is further adopted for analysis, and the regression equation is as follows:

$${GI}_{ijt}={\beta }_0+{\beta }_1{DIF}_{jt}+{\beta }_2{controls}_{ijt}+{\mu }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(2)

$${Med}_{ijt}={\theta }_0+{\theta }_1{DIF}_{jt}+{\theta }_2{controls}_{ijt}+{\mu }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(3)

$${GI}_{ijt}={\delta }_0+{\delta }_1{DIF}_{jt}+{\delta }_2{Med}_{ijt}+{\delta }_3{controls}_{ijt}+{\mu }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(4)

In order to explore the non-linear relationship between digital financial inclusion and green innovation of SMEs, the following threshold effect test model is constructed by referring to the research method of [47]:

$$\begin{gathered} {GI}_{ijt}={\delta }_0+{\delta }_1{DIF}_{ijt}\times I\left(q_{ijt}\le {\mu }_1\right)+{\delta }_2{DIF}_{ijt}\times I\left({\mu }_1\le q_{it}\le {\mu }_2\right)+\dots +{\delta }_n{DIF}_{ijt}\times I\left(q_{ijt}\ge {\mu }_n\right)+ \\ \sum {\delta }_3{Control}_{ijt}+\sum Year+\sum Industry+{\epsilon }_{ijt} \end{gathered}$$

(5)

Among them, i, j, and t represent enterprises, regions, and years, respectively; $GI$ indicates the level of green innovation of enterprises; DIF represents digital inclusive finance; Control represents a series of control variables; Med represents intermediary variables; ${\mu }_i$ indicates the individual fixed effect of the enterprise; ${\gamma }_t$ indicates the time fixed effect; ${\epsilon }_{ijt}$ is a random error item; I (·) is a conditional function, if the conditions in parentheses are valid, I is 1, otherwise it is 0; ${\delta }_0$ is the intercept term; ${\delta }_1$, ${\delta }_2{,\delta }_n$_ are the estimated coefficients of government subsidies. $q_{ijt}$_ it is the threshold variable; ${\mu }_1$, ${\mu }_2{,\mu }_n$ are threshold values to be measured.

4.2. Variables Selection

4.2.1. Dependent Variable: Green Innovation of Enterprises

There are three main methods for measuring the green innovation of enterprises in current research. The first method is using questionnaires to measure on a scale. This method is greatly influenced by the subjective factors of the participants. The second method is input-output ratio analysis. Due to the characteristics of high investment, high risk, and low conversion rate in green innovation activities, using input and output to measure green innovation may lead to underestimation. The third method is the quantity of green patent applications or authorizations. In this study, we draw on the research methods of [49] to measure green innovation. The number of green patent applications by listed companies on China’s SME board from 2011 to 2021 is selected as the measure of green innovation of SMEs (GI). On this basis, the total sum of corporate green innovation patents is logarized by adding 1 with the corporate patent information provided.

4.2.2. Independent Variable: Digital Inclusive Finance

Based on the report “Digital Inclusive Finance Index (2011–2021)” compiled by the Digital Finance Research Center of Peking University, this study measures the level of digital inclusive finance in different regions from two dimensions: inclusiveness and digitization, which are shown in Figure 1. Regarding inclusiveness, this study selects two indicators, namely the breadth of digital financial coverage (BDIF) and the depth of digital financial usage (DDIF). Regarding digitization, this study selects the indicator of the digitization level of inclusive finance (GDIF). The breadth of digital financial coverage evaluates the degree of coverage of digital financial services, emphasizing the provision of sufficient financial services. The depth of digital financial usage reflects the actual use of digital inclusive financial services, emphasizing effective demand. The digitization level of inclusive finance measures the value of digital inclusive finance through the convenience of digital inclusive financial services, the level of borrowing costs, and the degree of creditization. Moreover, this paper draws on relevant studies on digital inclusive finance, and these variables are transformed into the logarithmic scale [31,50,51].

4.2.3. Control Variables

Since the research on green innovation of SMEs is correlated with their financial conditions and management aspects, based on the extensive literature [52,53], this study selects six control variables to eliminate the potential influence of these variables on enterprises’ green innovation. They are as follows and is shown in the

Table 1. Specific description of control variables.

Variable	Variable Name Symbol	Measurement Method
Enterprise scale	Size	Obtained by taking the logarithm of the total assets of the enterprise.
Asset–liability ratio	Lev	Measured by the ratio of the total liabilities and total assets of the enterprise in the year.
Equity concentration	Top1	It is expressed by the shareholding ratio of the largest shareholder.
Management cost rate	MCF	The current management expenses of the enterprise are divided by the operating income.
Return on assets	ROA	Measured by the total profit of the enterprise in the year and the ratio of total assets.

:

Enterprise Size (Size): The size of an enterprise’s development is closely related to the implementation of green innovation activities. The value of this variable is obtained by taking the logarithm of the total assets of the enterprise.

Debt Ratio (Lev): The debt capacity of an enterprise affects its cash flow situation and level of research and development investment. The ratio of total liabilities to total assets in the current year is used to measure this variable.

Equity Concentration (Top1): The development of a company is closely related to the decisions made by its management. The decision-making power of the management has a significant impact on the company’s development. Therefore, this variable is represented by the proportion of shares held by the largest shareholder.

Management Expense Ratio (Mf): This is an important factor influencing the profitability of an enterprise and reflects the level of its operational management. This variable is represented by dividing the current period’s management expenses by operating income.

Return on Assets (Pro): The higher the profitability level of an enterprise, the better it can ensure the smooth operation of its cash flow and provide financial support for research and development activities. Therefore, the ratio of annual profit to total assets is used to measure this variable.

4.2.4. Intermediate Variable

In this study, financing constraints are chosen as the intermediate variable. Generally, there are three representative methods for measuring financing constraints: the SA index, KZ index, and WW index. This study uses the KZ index to measure financing constraints. The KZ index is a method proposed by [54] to measure the level of financing constraints faced by companies. The KZ index can have both positive and negative values, and a higher index indicates a higher degree of financing constraints faced by the company.

The calculation formula is:

$$\begin{gathered} KZ=-1.001909\times \frac{OCF}{Asset}+3.139193\times Lev-39.3678\times \frac{Dividends}{Asset}- \\ 1.314759\times \frac{cash}{Asset}+0.2826389\times TobinQ \end{gathered}$$

(6)

Among them, OCF refers to net operating cash flow, Dividends refers to dividends, cash refers to cash holding level, and all are standardized by total assets at the beginning of the period. Lev refers to the asset–liability ratio, and Tobin Q refers to Tobin Q value.

4.3. Sample Selection and Data Sources

This study selects listed companies on the Chinese SMEs Board from 2011 to 2021 as the research subjects. The company data are sourced from the CSMAR database, while the green patent data are obtained from the website of the National Intellectual Property Administration and the World Intellectual Property Organization (WIPO) official website. The indices for the level of digital inclusivity in inclusive finance, breadth of digital financial coverage, and depth of financial usage are sourced from the “Peking University Digital Inclusive Finance Index” compiled by the Peking University Digital Finance Research Center [11].

Furthermore, to ensure the scientific integrity and reasonability of the data and mitigate the impact of outliers, the following data treatments were carried out prior to the empirical analysis:

First, companies listed as “ST” and those that have been delisted were excluded from the sample.

Second, companies in the financial industry were excluded from the sample.

Third, companies with missing data for certain variables were excluded.

Fourth, a winsorization method was applied to the entire sample data to trim extreme values within the upper and lower 1%.

As a result, a panel dataset with 4927 observations was formed, and the flowchart of the research is shown in Figure 2.

5. Empirical Results Analysis

5.1. Descriptive Statistical Analysis

In order to understand the distribution and characteristics of variables, the Stata 17.0 version is used to conduct descriptive statistical analysis on relevant variables. The descriptive statistical results of the sample data are shown in

Table 2. Descriptive statistical results.

Variable	N	Mean	SD	Min	Max	P50	Range
GI	4927	0.70	0.99	0	5.52	0	5.21
BDIF	4927	5.45	0.51	2.64	6.00	5.64	3.36
DDIF	4927	5.46	0.36	3.14	5.92	5.55	2.78
GDIF	4927	5.43	0.38	2.64	5.87	5.60	3.23
KZ	4927	0.11	2.51	−12.91	7.58	0.48	20.49
Size	4927	21.13	0.90	18.39	25.91	21.04	7.52
Lev	4927	0.29	0.17	0.0081	0.97	1.28	6271
Top1	4927	0.31	0.13	0.041	0.90	29.06	85.91
ROA	4927	0.07	0.05	−0.03	0.43	0.07	0.47
MCF	4927	0.11	0.07	0.0013	1.07	0.09	1.07

. From Table 2, it can be seen that, on average, companies in the sample apply for 0.70 green patents annually, with a maximum value of 5.52 and a minimum value of 0. The standard deviation is 0.99, and the range is 5.21, indicating significant differences in green innovation among companies, with some displaying strong green innovation capabilities. The mean values of digital financial coverage breadth, digital financial usage depth, and inclusive finance digitalization level are 5.45, 5.46, and 5.43, respectively, with standard deviations of 0.51, 0.36, and 0.38, indicating varying levels of development in digital financial coverage, usage depth, and inclusive finance digitalization among SMEs in different regions within the research sample.

The average value of the KZ index is 0.11, with a standard deviation of 2.51, highlighting significant discrepancies in financing constraints faced by different SMEs. Among the control variables, the average company size is 21.13, with a standard deviation of 0.90, indicating significant differences in company size within the sample. The average asset–liability ratio is 0.29 with a standard deviation of 0.17, showing considerable diversity in capital structure among different SMEs. The mean value of the proportion of shares held by the largest shareholder is 0.31, with significant differences observed in the equity structure of sample companies. The minimum asset return rate is −0.33, while the maximum is 0.43, indicating varying profitability levels among companies, with significant differences in profit-making abilities. The mean management expense ratio is 0.11, with a notable gap between the maximum and minimum values of 1.07 and 0.0013, signifying significant disparities in operational management efficiency among different companies.

5.2. Benchmark Regression Analysis

The regression results of the impact of digital inclusive finance on green innovation in SMEs are shown in

Table 3. Benchmark regression results.

Variable	(1)	(2)	(3)
	GI	GI	GI
BDIF	0.36 ***
	(2.76)
DDIF		0.27 *
		(1.89)
GDIF			0.19 *
			(1.91)
Size	0.18 **	0.17618 **	0.18 **
	(4.91)	(4.91)	(4.93)
Lev	0.54 ***	0.54 ***	0.55 ***
	(3.98)	(4.00)	(4.02)
ROA	0.22	0.25	0.30
	(0.72)	(0.80)	(0.97)
MCF	−0.16	−0.11	−0.07
	(−0.57)	(−0.39)	(−0.25)
Top1	−0.003	−0.005	−0.009
	(−0.02)	(−0.03)	(−0.06)
Constant	−5.15 ***	−4.74 ***	−4.28 ***
	(−5.23)	(−4.70)	(−4.89)
Year-fixed	control	control	control
Ind-fixed	control	control	control
N	4927	4927	4927
R2a	0.23	0.23	0.23

Note: Regression T-values are in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% confidence levels, respectively.

. The above regression results indicate that the regression coefficients of the level of digital inclusive finance, the breadth of digital financial coverage, and the depth of financial usage are positive and pass the 1% and 10% significance levels, respectively. It can be observed that digital inclusive finance significantly promotes green innovation in SMEs, thus validating hypothesis H1. Comparing the impact of the level of digital inclusive finance, the breadth of digital financial coverage, and the depth of financial usage on green innovation, it is found that the breadth of digital financial coverage has the greatest impact on green innovation in SMEs, followed by the level of digital inclusive finance, with financial usage depth having the lowest impact.

Company size, asset–liability ratio, and asset return rate all have a significant positive impact on green innovation in companies. In addition, the asset return rate and asset–liability ratio pass the 1% significance level, while company size passes, the 5% significance level.

5.3. Robustness Test

To ensure the reliability of the benchmark regression results, a robustness test is conducted by replacing the dependent variable and lagging the core independent variable by one period.

Replacement of the dependent variable. By drawing on [55], the robustness analysis is conducted by using the number of green invention patent applications (GIA) as a measure of the level of green innovation in SMEs instead of the quantity of green patents. The regression estimation results are shown in columns (1), (2), and (3) of

Table 4. Robustness test results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	Replace the Interpreted Variables			The Core Interpretation Variable Lags Behind One Period
	GIA	GIA	GIA	GI	GI	GI
BDIF	1.54 ***
	(3.58)
DDIF		0.89 **
		(1.98)
GDIF			1.161 ***
			(2.02)
L.BDIF				2.30 ***
				(3.77)
L.DDIF					1.18 *
					(1.86)
L.GDIF						1.89 *
						(1.68)
Constant	9.98 ***	8.53 ***	11.59 ***	12.91 ***	11.91 ***	15.24 ***
	(−5.02)	(−4.73)	(−4.71)	(−3.80)	(−4.07)	(−3.14)
Controls	YES	YES	YES	YES	YES	YES
Year-fixed	YES	YES	YES	YES	YES	YES
Ind-fixed	YES	YES	YES	YES	YES	YES
N	4927	4927	4927	4927	4927	4927
R2a	0.15	0.15	0.15	0.18	0.18	0.18

Note: Regression T-values are in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% confidence levels, respectively.

. It can be observed that the regression coefficients of the three indicators of digital inclusive finance are all significant and positive, confirming the robustness of the previous results.

Lagging the core independent variable. Considering the potential time lag in the impact of digital inclusive finance on green innovation in SMEs and to control for reverse causality, the digital inclusive finance-related indicators are lagged by one period and further included in the regression analysis. The results, shown in columns (4), (5), and (6) of Table 4, indicate that even after considering the lagged effects, the regression coefficients remain significant and positive, further validating the robustness of the results. This suggests that digital inclusive finance still significantly promotes green innovation in SMEs.

Overall, the robustness test results indicate that the regression coefficients of the three indicators of digital inclusive finance are significant and positive, indicating that the improvement of digital inclusive finance significantly promotes green innovation in SMEs, consistent with the performance of the benchmark regression.

5.4. Mechanism Test

To examine the intermediating effect of financing constraints, this study adds financing constraints to the benchmark model for further testing. First, a regression test is conducted on the relationship between digital inclusive finance and financing constraints. Then, a regression test is performed on the relationship between digital inclusive finance, financing constraints, and green innovation, yielding the results shown in

Table 5. Intermediating effect test results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	KZ	KZ	KZ	GI	GI	GI
BDIF	−0.232 **			2.26 ***
	(−2.04)			(3.67)
DDIF		−0.27 **			1.11 *
		(−2.30)			(1.71)
GDIF			−0.05			2.17 *
			(−0.22)			(1.84)
KZ				−0.20 ***	−0.21 ***	−0.21 ***
				(−2.61)	(−2.66)	(−2.71)
Constant	−11.40 ***	−11.41 ***	−11.57 ***	−6.87 ***	−6.77 ***	−6.79 ***
	(−15.34)	(−15.38)	(−15.55)	(−16.66)	(−16.42)	(−16.43)
Controls	YES	YES	YES	YES	YES	YES
Year-fixed	YES	YES	YES	YES	YES	YES
Ind-fixed	YES	YES	YES	YES	YES	YES
N	4927	4927	4927	4927	4927	4927
R2a	0.32	0.32	0.32	0.17	0.16	0.16

Note: Regression T-values are in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% confidence levels, respectively.

.

The regression results in columns (1), (2), and (3) of Table 5 indicate that the breadth and depth of digital financial coverage have significantly negative effects on financing constraints. Specifically, for every 1% increase in the breadth of digital financial coverage, financing constraints decrease by 0.23%. Similarly, for every 1% increase in the depth of digital financial usage, financing constraints decrease by 0.27%. Both effects pass the 5% significance level. However, the digitalization level of inclusive finance has a negative impact on financing constraints, but the effect is not significant. In general, increasing the breadth and depth of digital finance can reduce the financing difficulties faced by SMEs to some extent.

The results in columns (4), (5), and (6) of Table 5 show that under the influence of financing constraints, the breadth of digital financial coverage, the depth of digital financial usage, and the digitalization level of inclusive finance have significant positive effects on green innovation capabilities. Furthermore, financing constraints have a significant negative effect on the green innovation capabilities of SMEs. This suggests that digital inclusive finance can reduce the impact of financing constraints, thereby alleviating the financing difficulties of enterprises and promoting their green innovation capabilities.

5.5. Threshold Effect Test

In order to explore the non-linear relationship between digital financial inclusion and green innovation of SMEs, based on [56], a panel threshold effect was carried out to test the non-linear impact of digital financial inclusion on green innovation of SMEs. First, the significance of the threshold effect is tested, and the number of thresholds is found. The results are shown in

Table 6. Results of the threshold test.

Threshold Variables	Threshold Type	Threshold Values	F-Statistics	p-Value	Lower Limit	Upper Limit
GDIF	First threshold	5.7794 **	20.39	0.0167	5.7781	5.78
	Second threshold	5.5252	5.47	0.52	5.5141	5.5292
	Third threshold	5.6934	6.12	0.7667	5.6881	5.6986

Note: ** indicates significance at the 5% confidence level.

and Figure 3. In the threshold effect test with digitization degree as the explanatory variable and threshold variable, the single threshold p-value is less than 5%, and the significance passes; the double threshold p-value is 0.46, and the value is greater than 10%, and the significance fails. There is a single threshold for the impact of digital financial inclusion on green innovation of SMEs, and the regression results are shown in

Table 7. Threshold regression test results.

Variables	GI
GDIF ≤ 5.7794	0.521 *
	(1.898)
GDIF ≥ 5.7794	0.470 *
	(1.731)
Size	0.123 *
	(1.815)
Lev	0.002
	(0.333)
ROA	−0.934 *
	(−1.694)
MCF	−0.715
	(−1.125)
Top1	0.009 *
	(1.689)
Constant	−4.952 ***
	(−3.079)
Observations	1062
Number of ID	177
R2a	0.155

Note: Regression T-values are in parentheses; * and *** indicate significance at the 10% and 1% confidence levels, respectively.

. As can be seen from the results in the table, the single threshold value of GDIF is 5.78. When the GDIF is <5.78, its regression coefficient is 0.47, which is significantly positive at the 10% level. When the GDIF is ≥5.78, its regression coefficient is 0.52, which is significantly positive at the 10% level. It can be seen that with the improvement of the digitization level, its influence on the green innovation of SMEs presents a significant positive and non-linear feature of increasing “marginal effect”. H3 is proven.

5.6. Heterogeneity Analysis

5.6.1. Heterogeneity Analysis of Enterprise Nature

The differences in enterprise nature can affect the impact of digital inclusive finance on green innovation in SMEs. Therefore, the sample data are further divided into two sub-samples, state-owned SMEs and non-state-owned SMEs, for regression analysis. The results are shown in

Table 8. Test results of enterprise heterogeneity.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	State-Owned	Non-State-Owned	State-Owned	Non-State-Owned	State-Owned	Non-State-Owned
	GI	GI	GI	GI	GI	GI
BDIF	0.41	2.41 ***
	(0.19)	(3.32)
DDIF			1.86	1.10
			(0.67)	(1.40)
GDIF					0.76	1.87
					(0.15)	(1.58)
Controls	YES	YES	YES	YES	YES	YES
Constant	−2.93	−6.93 **	−2.34	−9.57 **	−6.58	−4.25 **
	(−0.49)	(−2.41)	(−0.27)	(−2.21)	(−1.09)	(−2.00)
Year-fixed	YES	YES	YES	YES	YES	YES
Ind-fixed	YES	YES	YES	YES	YES	YES
N	320	4607	320	4607	320	4607
R2a	0.27	0.11	0.28	0.11	0.27	0.11

Note: Regression T-values are in parentheses; ** and *** indicate significance at the 5%, and 1% confidence levels, respectively.

.

It can be observed that for non-state-owned enterprises, the level of digitalization in inclusive finance, the breadth of digital financial coverage, and the depth of digital financial usage all have a positive promoting effect on their green innovation. Furthermore, the breadth of digital financial coverage passes the 5% significance level. However, for state-owned SMEs, the promoting effects of the breadth of digital financial coverage, the depth of digital financial usage, and the level of digitalization in inclusive finance on green innovation are not apparent. The possible reason for this is that the state-owned SMEs have obvious advantages in green innovation under the current institutional background in China. They are not only favored by various policies and resources but also have priority access to more financing resources under the credit endorsement of the government. Therefore, the marginal effect of digital inclusive finance is relatively small. On the other hand, digital inclusive finance can improve their financing environment, effectively alleviate their financing constraints, and facilitate their green innovation activities for non-state-owned SMEs. Consequently, the impact of digital inclusive finance on green innovation in state-owned SMEs is weaker than that in non-state-owned SMEs.

5.6.2. Regional Heterogeneity Analysis

In order to test the regional heterogeneity of the impact of digital inclusive finance on green innovation in SMEs, this study divided the sample into three groups—eastern, western, and central regions—based on China’s economic division standards combined with company registration information for group regression analysis. The results are shown in

Table 9. Regional heterogeneity test results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	East Regions	West Regions	Central Regions	East Regions	West Regions	Central Regions	East Regions	West Regions	Central Regions
	GI	GI	GI	GI	GI	GI	GI	GI	GI
BDIF	3.02 ***	6.28 **	7.89 *
	(3.08)	(2.15)	(1.83)
DDIF				1.11	7.10	16.38 *
				(1.01)	(1.61)	(1.84)
GDIF							0.88	3.19	17.70 *
							(0.57)	(0.94)	(1.92)
Controls	YES	YES	YES	YES	YES	YES	YES	YES	YES
Constant	3.04 ***	10.47 *	10.51	3.61 ***	10.79 *	11.00	2.40 **	9.80 *	10.81
	(−2.66)	(−1.82)	(−0.91)	(−3.08)	(−1.81)	(−0.94)	(−2.23)	(−1.81)	(−0.93)
Year-fixed	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind-fixed	YES	YES	YES	YES	YES	YES	YES	YES	YES
N	3289	337	503	3289	337	503	3289	337	503
R2a	0.15	0.09	0.20	0.14	0.09	0.21	0.14	0.09	0.20

Note: Regression T-values are in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% confidence levels, respectively.

.

The results indicate that the breadth of digital financial coverage has a significant impact on green innovation in the eastern, western, and central regions. Compared to the eastern and western regions, the impact is greatest in the central region, followed by the western region. In comparison to the central region sample, the depth of digital financial usage and the level of digitalization in inclusive finance have an insignificant impact on green innovation in the eastern and western region samples, suggesting that digital inclusive finance benefits innovation development in the central region but has limited influence on green innovation development in SMEs in the eastern and western regions.

Possible reasons for this include the better digital infrastructure and digital integration capabilities in the central region compared to the western region, where digital infrastructure is relatively weaker. Additionally, the lower economic development and financial level in the central region compared with the eastern region result in a weaker overall financial system and higher financing constraints faced by businesses. Therefore, the demand for digital inclusive finance is greater in the central region, and when digital inclusive finance effectively alleviates financing constraints, the green innovation effect in enterprises in the central region is more significant compared to those in the eastern region.

5.6.3. Heterogeneity Analysis of Equity Structures

To examine the heterogeneous effects of empowering green innovation in SMEs through digital inclusive finance under different equity structures, this study refers to [57] research on the relationship between government subsidies, equity structure, and corporate green innovation. The ratio of the total shareholding of the second to fifth largest shareholders to that of the largest shareholder is used as a key variable to measure the degree of equity dispersion in SMEs. Groups with ratios higher than the sample average are categorized as concentrated equity groups, while those with ratios lower than the sample average are categorized as dispersed equity groups. The results of the group regression analysis can be seen in

Table 10. Test results of equity structure heterogeneity.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	Equity Dispersion Group			Equity Concentration Group
	GI	GI	GI	GI	GI	GI
BDIF	1.86 **			2.73 ***
	(2.06)			(2.69)
DDIF		1.76 *			0.68
		(1.83)			(0.60)
GDIF			4.47868 ***			0.47
			(2.61)			(0.28)
Controls	YES	YES	YES	YES	YES	YES
Constant	−6.61 ***	−6.58 ***	−6.64 ***	−6.78 ***	−6.60 ***	−6.55 ***
	(−3.16)	(−3.14)	(−3.15)	(−2.92)	(−2.86)	(−2.81)
Year-fixed	YES	YES	YES	YES	YES	YES
Ind-fixed	YES	YES	YES	YES	YES	YES
N	2419	2419	2419	2508	2508	2508
R2a	0.12	0.12	0.13	0.12	0.12	0.12

Note: Regression T-values are in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% confidence levels, respectively.

.

From the results in Table 8, it can be observed that under both concentrated and dispersed equity situations, the breadth and depth of digital financial coverage show significantly positive effects, and pass the 10% and 5% significance levels, respectively. The level of digitization in inclusive finance has a positive impact, and passes the 5% significance level in the case of dispersed equity, but no significant impact in the case of concentrated equity. However, under dispersed equity conditions, the coefficients of the three indicators of digital inclusive finance on corporate green innovation are greater than those under concentrated equity conditions. The main reason for this is believed to be the improvement in the balance of power among shareholders, leading to mutual checks and balances among shareholders. In situations where a shareholder makes a decision detrimental to the company, other shareholders will promptly intervene to prevent the reduction of investments in green innovation activities by major shareholders seeking personal gain.

Moreover, the enhanced proactiveness and supervisory willingness of major shareholders promote the smooth implementation of green innovation activities, reduce the risk of failed green innovation investment decisions, and provide solid support for improving corporate performance. Therefore, a higher degree of equity balance enhances mutual constraints among shareholders, contributing to the strengthening of green innovation in SMEs.

6. Discussions

The empirical analysis shows that digital inclusive finance has a significant role in supporting SMEs’ green innovation. This finding is consistent with the study by [47]. In addition, we find that financing constraints play a partial mediating role in the relationship between digital financial inclusion and SMEs’ green innovation. On the one hand, digital inclusive finance, as a new financing method, broadens the financing channels of green innovation for SMEs [58]. On the other hand, digital financial inclusion can help reduce the financing cost of green innovation for SMEs [59]. Relying on big data and popularity, digital inclusive finance can break the limitations of traditional financial institutions in space and information transmission, reduce the cost and threshold of financial services [60], help alleviate the financing obstacles faced by SMEs in the process of green innovation, and bring new development space for SMEs in green innovation. This is similar to the findings of [53]. The threshold effect results show that the green innovation of SMEs is affected by the digitalization level of digital inclusive finance, and the higher digitalization level has a greater role in promoting the green innovation of SMEs. With the continuous improvement of the digital level, digital inclusive finance improves the allocation efficiency of financial resources [11], enhances the participation of green innovation subjects, helps financial institutions effectively identify the relevant information of SMEs’ green innovation [46], and enhances the external financing capacity of SMEs’ green innovation. It has greatly promoted its enabling effect on green innovation.

7. Conclusions and Implications

7.1. Conclusions

Empirical research is carried out using data from SMEs listed on the SME board from 2011 to 2021. Conclusions can be drawn from the study:

• There is a significant positive correlation between the breadth of digital financial coverage, depth of usage, level of inclusive finance digitization, and green innovation of SMEs. Therefore, improving the level of digital inclusive finance can promote green innovation in SMEs to a certain extent.
• The intermediating effect test reveals that digital inclusive finance has a significant negative impact on the financing constraint index. Digital inclusive finance indirectly stimulates SMEs to enhance their green innovation capabilities by affecting their financing constraints and reducing the difficulty of financing.
• The threshold effect results show that the dynamic impact of digital inclusive finance on SMEs’ green innovation is influenced by the digital level of digital inclusive finance, showing a significant positive and non-linear feature of increasing “marginal effect”.
• Compared with state-owned SMEs, digital inclusive finance has a greater promotional effect on non-state-owned SMEs; compared with the central region, the promotion effect of digital inclusive finance on the eastern and western regions is rather weaker; compared with SMEs with concentrated equity ownership, digital inclusive finance has a greater positive impact on green innovation in SMEs with dispersed equity ownership.

7.2. Implications

Based on the research and the obtained conclusions, the following recommendations are proposed:

• The depth and breadth of digital finance usage, as well as the level of inclusive finance digitization, will become important driving forces for the rapid growth of digital inclusive finance. To promote the transformation of digital inclusive finance from “extensive” development to efficient and in-depth expansion, it is necessary to deepen the exploration of various functions of digital inclusive finance, introduce digital inclusive finance services to SMEs and underdeveloped regions through financial support, policy bias, etc., expand the breadth and depth of digital inclusive finance coverage and the level of inclusive finance digitization, optimize the efficiency of financial resource allocation, make it play a more important role in the development of green innovation in SMEs, and improve the regional development imbalance.
• Tailored use of digital inclusive finance to drive green innovation in SMEs. In terms of enterprise nature, efforts should be made to further promote the reform of state-owned SMEs, remove the label of “distorted competition”, promote fair participation in market competition, and enhance the awareness of green innovation in state-owned enterprises while encouraging and supporting non-state-owned SMEs to unleash their green innovation vitality through digital inclusive finance. Differentiated policies on digital inclusive finance should also be formulated according to the characteristics of each region: the central region should accelerate the digital transformation of the manufacturing industry, promote the optimization and upgrading of industrial structure, narrow the digital gap with the eastern region, and achieve green innovation development through digital inclusive finance services; the eastern region should rely on mature financial markets, talent, and technology clustering advantages to amplify the green innovation effects of digital inclusive finance, striving to build a highland of green innovation; the western region should take advantage of policy biases, leverage the “latecomer advantage”, and gradually reduce the gap with regions with higher levels of digital inclusive finance development. Meanwhile, in the process of green innovation development in SMEs, to effectively prevent the infringement of the interests of minority shareholders by the largest shareholder, it is necessary to increase the proportion of non-largest shareholders so that large shareholders can check and balance each other, making decisions more scientific and creating a good internal environment for the enhancement of green innovation capabilities in SMEs.
• SMEs should enhance their own awareness of green innovation, strengthen their own digital construction, and rationally use digital inclusive financial policies. According to the above analysis, digital inclusive finance can significantly promote the green innovation of SMEs. At the same time, with the improvement of the digital level of digital inclusive finance, its green innovation of small and medium-sized enterprises also presents a significant positive and non-linear feature of increasing “marginal effect”. Therefore, resource-based enterprises should enhance their awareness of green innovation. With the help of digital inclusive financial policies, green innovation activities should be vigorously carried out, but at the same time, attention should be paid to strengthening their own digital construction and improving the ability to acquire and control corporate data, thereby improving the efficiency of information transmission between borrowers and reducing information asymmetry, and thus helping financial institutions to effectively identify information related to green innovation of SMEs. The external financing capacity of SMEs for green innovation will be enhanced so that more green innovation entities can fully enjoy the spillover dividends of green innovation on a larger scale and at a lower cost.
• By strengthening the flow and cooperation of resources within regions to expand the inclusiveness of digital inclusive finance and improve the environment for green innovation in SMEs. Regions should rationally use policy means to promote the cross-regional flow of resources, technology, and talents, deepen interregional exchanges and cooperation, and fully utilize spatial effects to promote the development of green innovation in SMEs within the region through the development of digital inclusive finance.

7.3. Limitations and Future Research

There are still two noteworthy deficiencies in this study, which can be further studied in the future. First of all, this study uses data samples of listed SMEs, which may have more obvious reference significance for listed SMEs; however, its significance for unlisted SMEs is limited. In the future, we think SMEs should enhance their own awareness of green innovation, strengthen their own digital construction, and rationally use digital inclusive financial policies, so there may be an option to expand the scope of sample selection to reach more generally applicable conclusions. Secondly, this study mainly investigates from the overall micro level and only evaluates the promotion effect of digital inclusive finance on the green innovation of SMEs on the whole but does not reflect the effect of digital inclusive finance on the green innovation of enterprises in different industries. In the future, it can be discussed through comparative studies of various industries.