Digital Financial Inclusion and Innovation of MSMEs

Abstract

Digital inclusive finance is a new type of financial service that combines digital finance with inclusive finance. It is of great significance for improving the accessibility of financial services for small- and medium-sized enterprises and promoting their development. In this paper, we merge the Peking University Digital Financial Inclusion Index of China with the innovation data of micro-, small-, and medium-sized enterprises (MSMEs) from the National Bureau of Statistics, verify the facilitating effect of the development of digital financial inclusion on the technological innovation of MSMEs, and explain the mechanism of the influence of digital finance on the innovation of MSMEs from the perspective of alleviating financing constraints and promoting consumption. Digital inclusive finance has a promoting effect on the technological innovation of enterprises of different technological levels, but the support for high-tech enterprises is still insufficient. The heterogeneity analysis found that there are significant differences in the regional effects of digital financial inclusion; the central and western regions compared to the eastern region have better performance of digital inclusive finance to promote enterprise innovation.

1. Introduction

Inclusive finance, which refers to the provision of appropriate and effective financial services at an affordable cost to all social classes and groups in need of financial services, has achieved remarkable developmental results since it was first proposed by the United Nations in 2005. The Third Plenary Session of the 18th CPC Central Committee in 2013 explicitly proposed “developing inclusive finance”, and at the end of 2015, the State Council issued the “Plan for Promoting the Development of Inclusive Finance (2016–2020)” (referred to as the “Plan” below), which made the development of inclusive finance an important national development strategy. The plan emphasises the need to build a comprehensive inclusive financial services system, noting that the development of inclusive finance in China should be based on the principles of equal opportunity and economic sustainability and should be market-driven to provide appropriate and effective financial services to all segments of society. At present, inclusive finance has gradually expanded from small loans to comprehensive financial services. Digital financial inclusion is the digitalisation of inclusive financial services. The development of digital finance has improved the accessibility and convenience of financial services and has become an important means of achieving inclusive finance.

A large number of publications in the literature show that innovation is the main driver of economic development. Innovative enterprises tend to be high risk and thus prone to financing constraints. In order to promote high-quality economic development and technological progress, the financial system must be improved. Micro-, small-, and medium-sized enterprises (MSMEs) play an important role in technological innovation and promoting economic development. According to the Ministry of Industry and Information Technology’s 2022 report, MSMEs contribute more than 50 per cent of China’s fiscal revenue, more than 60 per cent of the GDP, and more than 70 per cent of technological innovation, accounting for more than 90 per cent of the number of enterprises. While MSMEs are impressive in terms of quantity and quality of development, they tend to face serious financing constraints in the development process due to the small scale of financing, poor qualification guarantees, imperfect credit records, and information asymmetry. Improving the financial capability of vulnerable groups and promoting the development of MSMEs is one of the main strategic objectives of inclusive finance policies. The development of inclusive finance contributes to the improvement of the financial infrastructure, providing guarantees for the technological innovation of MSMEs and the development of the economy.

As a new type of financial service, digital inclusive finance reduces the threshold of financial services and the cost of financial transactions, improves the coverage and efficiency of financial services, and brings new opportunities to ease the financing constraints of small- and medium-sized enterprises (Huang and Huang) [1]. The Central Financial Work Conference in 2023 proposed to “Do a good job in five major articles on Finance Technology, Green Finance, Inclusive Finance, Pension Finance, and Digital Finance”. It is of strategic significance to China’s financial development and high-quality economic development to promote the development of digital inclusive finance and enhance the capability of digital inclusive financial services for the real economy. Therefore, this paper explores the effect and mechanism of digital inclusive finance on enterprise innovation and provides evidence and feasible suggestions for a broader understanding of the development results and development obstacles of digital inclusive finance.

In this article, we mainly focus on four vital questions: (1) What are the effects of digital financial inclusion on MSMEs’ innovation? (2) Through what mechanisms does digital financial inclusion affect MSMEs’ innovation? (3) What are the heterogeneous characteristics of digital financial inclusion on MSMEs’ innovation? (4) As a new type of financial service, what improvements can be made to digital inclusive finance?

This paper merges the Peking University Digital Financial Inclusion Index of China with the innovation data of micro-, small-, and medium-sized enterprises (MSMEs) from the National Bureau of Statistics and empirically verifies the facilitating effect of the development of digital financial inclusion on the technological innovation of MSMEs with a double fixed-effects model and the instrumental variables method. This paper finds that digital inclusive finance has a positive effect on the technological innovation of MSMEs. Digital inclusive finance promotes innovation in MSMEs by alleviating financing constraints and stimulating consumption demand. The innovative effect of digital inclusive finance does not show significant differences for firms with different levels of technology, implying a lack of targeted incentives for innovation. Significant regional differences exist in the effects of digital inclusive finance. The central and western regions outperform the eastern regions, indicating that digital inclusive finance plays a strategic role in promoting balanced regional development.

The remainder of this paper is organized as follows. Section 2 analyses the conclusions and shortcomings of existing studies. Section 3 presents the theoretical analysis concerning how digital financial inclusion affects MSMEs’ innovation. Section 4 introduces the model and variables. Section 5 discusses the results of the benchmark regression, robustness test, mechanism analysis, and heterogeneity analysis. Finally, Section 6 presents conclusions and recommendations for further research.

2. Literature Review

Digital inclusive finance is an important form of digital finance and inclusive finance, and the existing literature has conducted extensive research on the development of digital inclusive finance.

First, many studies in the literature have investigated the impact of digital inclusive finance on corporate economic activities. Some of them suggest that digital inclusive finance improves the function of the financial system, reduces the cost of financial services, alleviates financing constraints for enterprises, increases productivity, and stimulates the innovation and entrepreneurship of enterprises (Bech et al., 2017; Wan et al., 2020; Candraningrat et al., 2021; Liu et al., 2021) [2,3,4,5]. Digital financial services may have a positive impact on the performance of MSMEs through technological, organizational, and environmental (TOE) characteristics (Dewi and Wiksuana, 2023) [6]. Digital inclusive finance has alleviated the liquidity constraints of rural credit, improved agricultural technology levels, and reduced the use of traditional energy products (Li and Zhang, 2023) [7]. In the multinational economy, digital financial services can help enhance e-commerce revenue, provide loan convenience for physical enterprises, especially small- and medium-sized enterprises, and promote economic growth (Sadigov et al., 2020) [8]. Digital finance also has a positive effect on green innovation, as the use of digital technology for green supply chain management can promote enterprise ecological technology innovation and improve enterprise performance (Liu et al., 2022; Nureen et al., 2023) [9,10]. Some articles worry about the regulatory risks associated with digital finance and suggest that appropriate regulatory rules are necessary to improve financial inclusivity and corporate financing (Reshetnikova, 2021; Pavlidis, 2021) [11,12]. Furthermore, the impact of digital financial services on the economy may be non-linear. In the early stages of the development of digital finance, international remittance inflows will significantly increase, but as the penetration rate increases, the remittance inflow rate may decrease (Emara, 2021) [13].

Second, many studies have focused on the impact of digital inclusive finance on household economic activities. Most of them suggest that digital inclusive finance can increase residents’ consumption, promote the upgrade of the consumption structure, improve the convenience of household consumption and investment, smooth intertemporal consumption, and reduce poverty (Feng and Zhang, 2021; Yang et al., 2022; Lu and Dilanchiev, 2023) [14,15,16]. However, some of them find that digital inclusive finance may promote household overconsumption, increase household leverage, and reduce household investment efficiency (Tian, 2022; Guo et al., 2022) [17,18]. The lack of moderate regulation and high-quality digital technology can lead to serious risks. Although poor people can participate in financial markets through digital inclusive financial services, it is difficult for them to reap the benefits (Ozili, 2020) [19].

Based on the above literature, the main contributions of this article are as follows. First, we provide a more accurate empirical study of the impact of developing digital inclusive finance in China. Existing studies may lack reliable data on digital inclusive finance and are limited to theoretical studies. Meanwhile, some studies have errors in the use of data, because they often fail to effectively distinguish different enterprise scales (Tang et al., 2020; Zhang et al., 2024) [20,21]. The main service targets of digital inclusive finance are micro-, small-, and medium-sized enterprises (MSMEs) and low-income people; large enterprises and listed companies should not be the research objects of digital inclusive financial services. Using large enterprises and listed companies to discuss the impact of developing digital inclusive finance is not in line with the characteristics of digital inclusive finance. Second, on the demand side, digital inclusive finance improves the financial services situation and income of low-income groups. Demand is an important driving force for enterprise innovation (Frenkel et al., 2015) [22], but the existing literature rarely analyses the impact of digital inclusive finance on enterprise innovation through this mechanism. The research in this paper enriches the existing literature on the impact effect and mechanism of digital inclusive finance. Third, digital inclusive finance affects the production of enterprises and the lives of residents, and socio-economic development also affects the development of digital finance. Some literature still lacks discussion on endogeneity issues. This article attempts to use an instrumental variable method to minimize the impact of endogeneity as much as possible. The major works and results of more relevant studies and the main contribution of this paper are listed in the Appendix A

Table A1. Relevant studies.

Author/s, Year	Main Idea	Objects	Method
Bech et al., 2017 [2]	Digital finance improves the efficiency of financial transactions and payments	Payment systems	Theoretical analysis
Reshetnikova, 2021 [11]	DIF may provide more financing opportunities for MSMEs and individuals, but also lead to serious risks	MSMEs, individuals	Theoretical analysis
Pavlidis, 2021 [12]	Regulating digital finance is necessary to improve financial inclusion and business financing.	EU, cryptocurrency market	Theoretical analysis
Fuster, 2019 [25]	FinTech can improve the efficiency of loan applications to promote financing	U.S. mortgage market	Theoretical analysis
Candraningrat et al., 2021 [4]	FinTech helps small and medium-sized enterprises to get loans easier	Indonesia, MSMEs	Questionnaire investigation
Sadigov et al., 2020 [8]	Fintech can enhance e-commerce revenue, provide loan convenience for enterprises	Physical enterprises	OLS
Nureen et al., 2023 [10]	Green supply chain management has a positive effect on eco-technological innovation	Manufacturing sector	SEM
Emara, 2021 [13]	The impact of digitization on remittance inflows is non-linear	Remittances inflow	GMM
Zhang et al., 2024 [21]	DIF intensifies bank competition, alleviates financing constraints, and promotes enterprise innovation	Listed companies	Empirical analysis
Tang et al., 2020 [20]	DIF can alleviate the financing constraints of micro enterprises and promote innovation	Listed companies	IV
Liu et al., 2022 [9]	DIF improves green innovation	Listed firms	IV
Wan et al., 2020 [3]	Digital inclusive finance (DIF) alleviates financing constraints for enterprises and stimulates innovation	Listed companies, MSMEs	IV
Liu et al., 2021 [5]	DIF can promote economic growth	MSMEs’ entrepreneurship	VAR
Dewi and Wiksuana, 2023 [6]	Digital financial has a positive impact on the performance of enterprises through technological, organizational, and environmental (TOE) characteristics.	MSMEs	SEM
Feng and Zhang, 2021 [14]	Digital finance has a positive impact on consumption upgrading	Consumption	Theoretical analysis
Ozili, 2020 [19]	DIF enables poor people to enter financial markets, bringing them enormous risks	Individuals	Theoretical analysis
Li et al., 2020 [36]	DIF has promoted the regular consumption expenditure of Chinese households	households	Mediating model
Ren et al., 2019 [37]	DIF can reduce cultural consumption differences between urban and rural residents	Individuals	OLS
Huang, 2021 [38]	DIF can significantly increase household leverage ratio	households	OLS
Tian, 2022 [17]	DIF may promote household overconsumption and increase household leverage	Household consumption	OLS
Yang et al., 2022 [15]	DIF promotes rural residents’ income and rural household subsistence consumption	Rural consumption	Lasso algorithms
Lu and Dilanchiev, 2023 [16]	Financial deepening effectively reduces poverty	Household consumption expenditure	SUR estimators
Guo et al., 2022 [18]	Digital finance in China can promote household investment portfolio efficiency	consumption and overconsumption	IV
Lu et al., 2024 [27]	Digital finance improves investment efficiency for investors	Individual Investors	IV
Wang and Zhao, 2020 [42]	Digital finance smoothes consumption and reduces income inequality	Individuals	IV
Li and Zhang, 2023 [7]	Digital financial alleviates the liquidity constraints of rural credit and improved the level of agricultural technology	Residential energy consumption	IV
This paper	DIF promotes innovation by alleviating financing constraints and expanding consumer demand	MSMEs, consumption	IV

.

3. Theoretical Analysis

Digital inclusive finance is an inclusive financial service conducted in a digital manner, providing comprehensive financial services including but not limited to microcredit for the poor and MSMEs, thereby meeting needs of individuals and enterprises. Innovation is the main driving force of economic development, and whether digital inclusive finance can promote technological innovation in MSMEs is an important basis for testing the effectiveness of digital inclusive finance development. Based on the service targets of financial inclusivity, we argue that digital inclusive finance mainly promotes innovation in micro-, small- and medium-sized enterprises in two ways, as described in the following paragraphs.

First, digital inclusive finance can ease the financing constraints of MSMEs and enhance their innovative capacity. Enterprises often face strong financing constraints during the long and risky cycles of technological innovation, making it difficult for them to innovate effectively. In comparison to large enterprises, MSMEs are more susceptible to discrimination by financial institutions due to information asymmetry issues, such as insufficient qualifications and guarantees and high business risks. Digital inclusive finance presents more opportunities for MSME financing. Enterprise financing constraints are often caused by information asymmetry (Kaplan and Zingales, 1997) [23]. Digital inclusive finance can help manage information efficiently through digital technology, providing accurate and timely assessments for MSMEs and reducing information asymmetry. This can help alleviate financing constraints and promote enterprise innovation (Jiménez et al., 2013) [24]. Meanwhile, digital inclusive finance can overcome barriers of time and space, save resources through online operations, and improve service efficiency (Fuster, 2019) [25], thereby reducing financing costs for MSMEs. Furthermore, the digital finance sector provides an opportunity to expand the source of funds, making it easier for small-scale investors to participate in the financial market, enhancing the allocation efficiency of financial capital, and improving the possibility of MSMEs’ financing (Gomber et al., 2017; Lu et al., 2024) [26,27]. Since MSMEs rely heavily on external financing during the innovation process (Trinh et al., 2017) [28], the alleviation of financing constraints through digital inclusive finance can be a significant incentive for firms to innovate. Therefore, we propose the first hypothesis of this paper:

Hypothesis 1: 

Digital inclusive finance promotes firm innovation by alleviating the financing constraints of MSMEs.

Second, the development of digitally inclusive finance has increased the consumption levels of low-income groups and stimulated innovation investment and technological upgrading by firms. According to the demand-led innovation theory, new demand in the market is an important driving force for innovation (Mowery and Rosenberg, 1979; Fontana and Guerzoni, 2008) [29,30]. When there is an income gap, increasing the income and consumption capacity of the poor can foster innovation (Foellmi and Zweimüller, 2006) [31]. Also, heterogeneous consumer demand may be lucrative for firms and enable them to innovate successfully (Björk, 2014) [32]. For individuals, digitally inclusive financial services are mainly used to fulfil consumer and financial needs. Some literature points out that petty loans are used to fulfil consumer needs (Attanansio et al. 2011; Peattie and Collins 2009) [33,34]. Digital inclusive finance can lower the service threshold for the last mile of financial services, helping people with relatively weak financial foundations to smooth consumption and manage risk. For example, microcredit, microsavings, and microinsurance have become effective financial management tools. Payment methods such as e-payments and mobile payments through mobile phones may reduce transaction costs (Ozili, 2023) [35], bring convenience, and increase the consumption capacity of the poor, leading to the growth of recurring needs, cultural needs, and other diverse needs (Li et al., 2020; Ren et al., 2019) [36,37]. However, the poor’s access to more convenient financial services through digital inclusive finance increases their consumption capacity while also increasing their consumption leverage (Huang, 2021) [38]. Financial risks are huge for the poor, and digital financial inclusion enables their access to financial markets, while it may be difficult for them to benefit from it. It is necessary to test whether digital financial inclusion is effective in increasing consumption and incentivising firms to innovate. Therefore, we propose the second hypothesis of this paper:

Hypothesis 2: 

Digital inclusive finance stimulates enterprise innovation by promoting consumer demand.

4. Model and Variables

4.1. Data

The provincial and municipal digital financial inclusion indices used in this paper were obtained from the Peking University Digital Financial Inclusion Index 2011–2020 (Guo et al., 2020) [39]. Compared to statistics from other financial institutions or research organisations, these data include information from all provincial-level administrative regions and 337 prefectural-level cities in China, providing a wide range of coverage. In terms of time span, these data have temporal coherence and cover a long period of time. As to indicator classification, these data comprise three dimensions, namely the breadth of digital financial coverage, the depth of digital financial use, and the degree of digitisation of financial inclusion, which is a more comprehensive measure of the development of digital inclusive finance. Therefore, this indicator can effectively depict the development trend of digital financial inclusion in various regions of China. The data are sourced from Alipay; while this platform holds a significant position in digital financial inclusion in China, it does not represent the entire landscape. Therefore, there may be an underestimation of the level of development of digital financial inclusion.

The data on microenterprises provided in this paper come from the 2008–2014 National Innovation Survey Enterprise Database of the National Bureau of Statistics (NBS). This database provides detailed information on the innovation activities of industrial and related service enterprises, making it the most comprehensive database for researching the innovation activities of microenterprises in China. It is worth noting that the current articles on the impact of digital financial inclusion on the economic activities of microenterprises often fall into the misuse of data and fail to categorise enterprises properly according to their size. As digital financial inclusion mainly works for MSMEs, we classify all enterprises into large, medium, small, and micro enterprises according to the Provisions on the Classification Standards for Small- and Medium-sized Enterprises jointly issued by the Ministry of Industry and Information Technology (MIIT), the National Bureau of Statistics (NBS), the Development and Reform Commission (DRC), and the Ministry of Finance (MOF) in 2011. We removed the samples belonging to large-scale enterprises and retained only the samples of MSMEs in the data to provide a more accurate analysis of the impact of digital financial inclusion.

We merged the above two databases by city name and year and finally obtained the panel data from 2011 to 2014. The other provincial and municipal control variables used in this article are from the National Bureau of Statistics (NBS) of China and the China Urban Statistical Yearbook.

4.2. Empirical Strategy

In order to find out whether the development of digital inclusive finance can promote technological innovation in firms, this paper constructs the following regression model using two-way fixed effects:

$${Inno}_{ijkt}={\alpha }_i+{\beta }_1{index}_{jt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(1)

In Equation (1), i denotes the enterprise, j denotes the city where the enterprise is located, k denotes the province where the enterprise is located, and t denotes the year. The independent variable in this paper is the digital financial inclusion index of the city where the enterprise is located, which is used to measure the level of development of digital financial inclusion, and the robustness test is conducted using the digital financial inclusion index of the province where the enterprise is located. The independent variable in this paper represents the level of innovation of firm i in year t. We use the output value of new products to measure the level of the innovation output of firms. The number of enterprise patent applications is mostly used to measure the innovation capacity of enterprises. Chen et al. (2020) [40] found that the number of enterprise patents, including the number of invention patents, does not effectively promote enterprise productivity, the reason for which is easy to understand. The existence of various forms of financial funding subsidy policies for enterprise patent application and authorization at all levels of government in China may cause a bubble phenomenon in the number of enterprise patents as well as distort innovation behaviour (Zhang et al., 2016) [41]. In order to effectively test the innovation capacity of enterprises, we use the output value of new products of enterprises to measure the innovation capacity of enterprises. In addition, we also use enterprise innovation input and the number of enterprise patent applications to measure the level of innovation output of manufacturing enterprises as a robustness test. Two-way fixed effects of industry and time are used to avoid confounding by unobservable factors.

The control variables selected in this study include the characteristic variables at the city level (GDP per capita, government expenditure per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita, road density per capita) and characteristic variables at the firm level (the number of employees in the firm, the government innovation subsidy per capita, the firm age, the innovation fixed capital investment per capita, the squared variable of firm age, export capacity, the degree of competition faced by the firm in the industry). In particular, given that numerous studies have found a non-linear relationship between the age of firms and their own innovation activities, i.e., that start-ups tend to rely on innovation inputs to gain a competitive advantage in the market, whereas mature firms prefer to reduce production costs or expand their market share to maintain a competitive advantage in the market, we include the squared variable of firm age in the regression. The variables are defined in

Table 1. Summary statistics.

Variables	Obs	Mean	SD	Min	Max	Definition
DFI	283,320	130.79	40.04	17.02	199.40	Digital inclusive finance index
Newproductoutput	284,988	6.06	5.38	0.000	19.172	The output value of new products, taken as the logarithm
gov_exp	284,238	5406.21	1909.51	705.91	9152.64	Government expenditure per capita
pergdp	281,461	96,054.97	76,627.67	6183.99	481,692.38	GDP per capita
Internet	272,377	0.91	0.75	0.06	5.10	The ratio of internet-using households
Industrialstructure	271,384	0.90	0.41	0.11	3.76	Tertiary industry output divided by secondary industry output
perfixedassets	278,956	48,482.00	24,971.67	383.16	219,392.83	Fixed-asset investment per capita
roads_pop	281,207	23.11	10.89	2.90	156.46	Miles of road per 10,000 people
staff	284,988	3.28	1.20	0.69	11.22	The number of firm employees
per_goversubsidy	284,988	0.66	1.26	0.00	8.99	Per capita government innovation subsidies
perinnovationcapital	284,986	2.02	1.97	0.00	10.73	Per capita investment in innovation fixed assets
age	284,141	13.67	9.58	2.00	188.00	Firm’s age
newproductexport	284,988	0.09	0.24	0.00	1.00	Firm’s new product export ratio
hhi	284,238	0.30	0.03	0.24	0.91	The degree of competition faced by the firm in the industry
per_inno_spend	284,988	4.85	1.04	0.00	12.43	Innovation investment per capita
patent	284,988	0.98	1.21	0.00	9.29	The number of patents
fin_dependence	191,648	15.44	4.98	2.68	34.65	Industry financing constraints
consumption	236,148	9.88	0.27	8.81	10.43	City-level consumption expenditure
Hightechindustries	285,058	0.16	0.36	0.00	1.00	High-tech industry or not
Eastern	284,988	0.71	0.45	0.00	1.00	Eastern region or not
CitylevelCoverage	283,320	132.89	43.37	1.86	219.98	The breadth of coverage
CitylevelUsage	283,320	131.83	38.87	1.94	215.29	The depth of use
CitylevelDigitization	283,320	121.99	51.59	−26.86	581.23	The degree of digitisation

.

4.3. Endogeneity Issues

Digital inclusive finance can be seen as a kind of infrastructure construction, and the level of enterprise development is closely related to the level of local economy, which may affect the development of digital inclusive finance in the region and is, therefore, prone to the problem of reverse causality. In addition, although other variables that affect the innovative capacity of firms have been controlled at the provincial, city and firm levels, there may still be an omitted variables problem, leading to biased estimation results. The potential endogeneity problems mentioned above have been addressed in this paper by using the instrumental variables method. Specifically, with reference to the literature (Wang and Zhao, 2020) [42], we use the average of the digital financial inclusion index of other cities in the province where the firm is located as an instrumental variable for the independent variable digital financial inclusion index, because digital finance does not show a completely hyper-geographical characteristic, and the level of digital financial inclusion in one region is positively correlated with its neighbouring regions (Guo et al., 2017) [43], but the innovation capacity of firms in a region is not directly affected by the level of digital financial inclusion development in other regions. The two-stage least squares model constructed in this paper is as follows:

$${index}_{jt}={\alpha }_i+{\beta }_2{nearbyindex}_{jt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(2)

$${Inno}_{ijkt}={\alpha }_i+{\beta }_3{index}_{jt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(3)

4.4. Mechanism Analysis

In order to verify whether digital inclusive finance can alleviate financing constraints and promote corporate innovation, this paper measures the external financing dependence of the industry in which the firms are located. If digital inclusive finance can indeed promote technological innovation by alleviating the financing constraints of the firms, it should be possible to observe that firms located in industries with a higher degree of external financing dependence are promoted more significantly. The National Bureau of Statistics (NBS) published the composition of capital by industry from 2011 to 2013, including enterprises’ own capital, domestic loan funds, bonds, foreign investment, and other sources of money; we use the ratio of all capital other than enterprises’ own capital to the whole amount of capital to measure the degree of industry reliance on external financing.

In this paper, we multiply the industry’s external financing dependence degree with the digital financial inclusion development index of the region where the firms are located and construct the following econometric model:

$${Inno}_{ijkt}={\alpha }_i+{\beta }_4{index}_{jt}\times {dependence}_{mt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(4)

In Equation (4), ${dependence}_{mt}$ denotes the industry’s external financing dependence degree, and m denotes the industry. If ${\beta }_4$ is positive, it can indicate that with the development of digital inclusive finance, the technological innovation ability of enterprises in industries with higher external financing dependence is much more improved.

In order to verify whether digital financial inclusion can provide an incentive for firms to innovate through a demand-led mechanism, we found the following models:

$${consumption}_{kt}={\alpha }_i+{\beta }_5{index}_{jt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(5)

$${Inno}_{ijkt}={\alpha }_i+{\beta }_6{consumption}_{kt}+\lambda X+{\mu }_{firm}+{\delta }_{year}+{\epsilon }_{ijkt}$$

(6)

Among them, ${\beta }_5$ and ${\beta }_6$ are the estimated parameters that this paper mainly focuses on, Equation (5) measures the effect of digital inclusive finance on consumption, and Equation (6) measures the effect of the consumption level on the innovation ability of enterprises. When both estimated coefficients are positively significant, it indicates that the demand-led effect is an important mechanism for the development of digital inclusive finance to promote the innovation of manufacturing enterprises.

5. Results and Discussion

5.1. Baseline Results

Table 2. Baseline results.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	All	All	All	Medium	Small	Micro
DFI	0.0171 ***	0.0347 ***	0.0220 ***	0.0269 ***	0.0163 **	0.0292 **
	(0.0057)	(0.0063)	(0.0070)	(0.0076)	(0.0069)	(0.0118)
Constant	3.7593 ***	1.7621 **	1.4999	1.5496	1.4478	−1.9816
	(0.7046)	(0.8847)	(1.7785)	(1.9116)	(1.7497)	(2.7847)
City Controls	No	Yes	Yes	Yes	Yes	Yes
Firms Controls	No	No	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
R2	0.0323	0.0585	0.2266	0.2414	0.2015	0.2768
Observation	204,899	194,591	194,583	99,834	92,183	2566

Note: This table shows the results obtained using ordinary least squares regression with two-way fixed effects. For all regressions, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In columns (1)–(3), the sample is all MSMEs. In columns (4)–(6), the sample is medium-, small-, and micro-sized enterprises respectively. Column (1) shows the regression results without any control variables added. Column (2) shows the regression results with city control variables. Columns (3)–(6) show the regression results with both city and firm control variables. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** and ** indicate that the coefficients are statistically significant at the 1% and 5% levels, respectively.

presents the regression results for the impact of digital financial inclusion on the innovation of MSMEs. For all regressions, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the innovation capability of MSMEs, which is measured by the new product output value of MSMEs. In columns (1)–(3) of Table 1, the coefficients of the independent variable are positively significant and robust after gradually adding province control variables and firm control variables to the model, indicating that the development of digital inclusive finance can have a positive impact on the innovation in MSMEs, which verifies Hypothesis 1. Columns (4)–(6) of Table 2 classify enterprises into medium, small and micro enterprises according to their size, and the coefficients of the independent variable are significantly positive, suggesting that the development of digital inclusive finance plays a positive role in promoting technological innovation in enterprises regardless of different sizes. The micro enterprises are the most disadvantaged of all enterprises, and the result in Column (6) reveals that the development of digital inclusive finance has the strongest promoting effect on innovation by micro enterprises. At the same time, medium-sized enterprises are also strongly supported by technological innovation. Although medium-sized enterprises are larger than micro and small enterprises, they are in an awkward intermediate place between large enterprises and micro and small enterprises. Banks tend to favour large enterprises and policy subsidies are often for small enterprises which means that medium-sized enterprises often face inadequate financial services, and digital inclusive finance has effectively alleviated the plight of such enterprises.

5.2. Instrumental Variable Method

Since the impact of digital inclusive finance on the economy and society may have endogeneity problems because of reverse causality or omitted variables, we have used the instrumental variable method to conduct a robust test.

Table 3. Instrumental variable method.

	(1)	(2)	(3)	(4)
Variables	All	Medium	Small	Micro
DFI	0.0691 ***	0.0767 ***	0.0592 ***	0.0714 ***
	(0.0134)	(0.0150)	(0.0122)	(0.0260)
City Controls	Yes	Yes	Yes	Yes
Firms Controls	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
KP-LM statistic	54.189 ***	51.790 ***	54.798 ***	42.555 ***
KP-F statistic	79.615	82.990	76.278	43.861
R2	0.2066	0.2143	0.1862	0.2430
Observation	179,236	92,467	84,406	2363

Note: This table shows the results obtained using instrumental variable method. For all regressions, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In Column (1), the sample is all MSMEs. In columns (2)–(4), the sample is medium-, small-, and micro-sized enterprises respectively. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** indicates that the coefficients are statistically significant at the 1% level.

shows the results of the regression using the instrumental variable method. After taking into account the endogeneity, the coefficients of the independent variable are significantly positive, indicating that the facilitating effect of digital inclusive financial development on the technological innovation of MSMEs is robust. The instrumental variable passed the underidentification test and the weak identification test, which means that the instrumental variables are valid.

5.3. Other Robustness Tests

In order to test the robustness of the regression results, this paper further replaces the explanatory variables with enterprise innovation investment which can effectively promote enterprise productivity and the number of patents, which is a commonly used proxy variable to measure the innovation capability of Chinese firms. Meanwhile, this paper uses the provincial digital financial inclusion index to replace the municipal financial inclusion index to test the incentivising role of digital financial inclusion development at the level of micro enterprise innovation.

Table 4. Other robustness tests.

	(1)	(2)	(3)
Variables	Innovation Investment	Patents	New Product Output
DFI	0.0055 ***	0.0031 ***
	(0.0012)	(0.0011)
Provincial DFI			0.0250 ***
			(0.0066)
Constant	3.7172 ***	−0.6422 ***	−1.4806
	(0.3439)	(0.2072)	(2.2828)
City Controls	Yes	Yes	Yes
Firms Controls	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
R2	0.1117	0.2492	0.2293
Observation	194,583	194,583	194,583

Note: This table shows the robustness tests results. For all regressions, the sample is all MSMEs. In Column (1), the independent variable is the city’s digital financial inclusion index, and the dependent variable is the innovation investment of MSMEs. In Column (2), the independent variable is the city’s digital financial inclusion index, and the dependent variable is the number of patents in MSMEs. In Column (3), the independent variable is the provincial digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In Column (1) the sample is all MSMEs. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** indicates that the coefficients are statistically significant at the 1% level.

shows the results after replacing the variables; in Columns (1) and (2), the dependent variable is the innovation investment and the number of patents of MSMEs, respectively, and the regression coefficients are both significant positive. In Column (3), the independent variable is the provincial digital financial inclusion index, and the dependent variable is the new product output value of MSMEs, and the regression coefficient is significantly positive. The results in Table 4 indicate that digital financial inclusion development does play a role in promoting the innovation capacity of MSMEs. Digital financial inclusion should be strengthened in order to promote sustainable economic development and enhance the innovative capacity of enterprises and cities.

5.4. Underlying Mechanisms

5.4.1. Financing Constraints

In order to test the role of the financing constraint mechanism in the process of the level of digital inclusive finance development affecting firms’ innovation, this paper adds the interaction term between the financing constraint and the level of digital inclusive finance development in model (1) to form model (4). Specifically, we use the industry financing constraints to which the enterprise belongs to portray the financing constraints of the enterprise. If the industry is more dependent on external financing, then the financing constraints faced by enterprises are more severe.

Table 5. Financing constraints.

	(1)	(2)	(3)	(4)
Variables	All	Medium	Small	Micro
DFI	0.0091	0.0106	0.0087	0.0030
	(0.0079)	(0.0088)	(0.0076)	(0.0194)
Financing dependence	−0.0086	−0.0311 **	0.0049	0.0686
	(0.0103)	(0.0130)	(0.0150)	(0.1201)
DFI × Financing dependence	0.0003 ***	0.0004 ***	0.0001	0.0009
	(0.0001)	(0.0002)	(0.0001)	(0.0009)
Constant	2.2887	2.9231	1.7434	−1.6717
	(1.6808)	(1.8578)	(1.6299)	(3.2480)
City Controls	Yes	Yes	Yes	Yes
Firms Controls	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
R2	0.2210	0.2355	0.1964	0.2726
Observation	153,114	74,431	76,631	2052

Note: This table shows the results of financing constraint mechanism. All regressions are with an interaction term between the financing constraint and the level of digital inclusive finance development, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In Column (1), the sample is all MSMEs. In Column (2)–(4), the sample is medium-, small-, and micro-sized enterprises respectively. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** and ** indicate that the coefficients are statistically significant at the 1% and 5% levels, respectively.

shows the results of the tests on the financing constraint mechanism, and our main concern is the coefficient of the interaction term. Column (1) displays the regression result with the full sample, and the coefficient of the interaction term is significantly positive, indicating that the more severe the financing constraints faced by enterprises, the more effectively the development of digital inclusive finance can promote their innovation capabilities. This result verifies that digital inclusive finance can promote the enterprise’s innovation capacity by alleviating the enterprise’s financing constraints. The coefficient of the interaction term in Column (2) is significantly positive, suggesting that digital inclusive finance has a significant effect on alleviating financing constraints in medium-sized enterprises. Columns (3) and (4) show that the financing constraint alleviation effect of small and micro enterprises is positive but not significant enough. Compared to medium-sized enterprises, the support of digital inclusive finance for financing of small and micro enterprises is relatively weak and more attention should be paid to the financing needs of small and micro enterprises.

5.4.2. Demand-Led Mechanisms

We use residents’ consumption expenditure to measure the level of residents’ demand in this paper, so as to find out whether digital inclusive finance can promote residents’ consumption capacity and stimulate corporate innovation through the demand-led mechanism. We first replace the dependent variables in Equation (1) with the level of urban residents’ consumption expenditure to form model (5). Column (1) in

Table 6. Demand-led mechanisms.

	(1)	(2)
Variables	Consumption	New Product Output
DFI	0.0093 ***
	(0.0010)
Consumption		1.1866 **
		(0.5970)
Constant	8.6247 ***	−5.6948
	(0.1036)	(5.7346)
City Controls	Yes	Yes
Firms Controls	Yes	Yes
Industry FE	Yes	Yes
Year FE	Yes	Yes
R2	0.8244	0.0666
Observation	219,182	311,879

Note: This table shows the results of demand-led mechanism. In Column (1), the independent variable is the city’s digital financial inclusion index, and the dependent variable is the city-level consumption expenditure. In Column (2), the independent variable is the city-level consumption expenditure, and the dependent variable is the new product output value MSMEs. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** and ** indicate that the coefficients are statistically significant at the 1% and 5% levels, respectively.

shows that the coefficient of the independent variables is significantly positive, which indicates that digital inclusive finance development level increase does promote residents’ consumption capacity. Further, we replace the independent variables in Equation (1) with the level of consumption expenditure to form the model (6). Column (2) of Table 6 shows that the coefficient of consumption expenditure is significantly positive, indicating that the increase in the level of citizens’ consumption does stimulate enterprise innovation, which verifies the hypothesis that digital inclusive finance promotes enterprise innovation through the demand-led mechanism. Demand plays an important role in driving innovation, and the results of this article find that digital inclusive finance enhances the consumption demand of residents and then promotes innovation in MSMEs.

5.5. Heterogenous Effects

5.5.1. Heterogenous Effects of Firms’ Technology Levels

The high-tech industry is the main force of enterprise technological innovation and the promotion of the high-quality development of the economy; meanwhile, there is no lack of small- and medium-sized micro-enterprises in high-tech enterprises. In order to explore whether the development of digital inclusive finance improves the technological innovation ability of high-tech enterprises, we classify the technological level of enterprises in accordance with the classification standard for high-tech industries (manufacturing) issued by the National Bureau of Statistics in 2011. The National Bureau of Statistics (NBS) has classified the technology level of all manufacturing industries into six categories according to the level of R&D inputs, including pharmaceutical manufacturing, aerospace equipment manufacturing, electronic and communication equipment manufacturing, computer manufacturing, medical equipment manufacturing, and chemical manufacturing. Therefore, we construct a dummy variable for high-tech manufacturing industry, which takes the value of one if the industry to which the enterprise belongs is a high-tech manufacturing industry and zero if the industry is not a high-tech manufacturing industry. We first classify all MSMEs according to whether they belong to the high-tech level manufacturing industry, and then we test the effect of digital financial inclusion development on the innovation ability of enterprises in the subsample of high-tech level manufacturing enterprises and other level manufacturing enterprises, respectively, and further add the interaction term between the level of digital financial inclusion development and the dummy variable of the industry’s technological level into Equation (1) to test the difference between the two types of sub-samples.

As can be seen from the Columns (1) and (2) of

Table 7. Heterogenous effects of firms’ technology levels.

	(1)	(2)	(3)
Variables	High-Technology	Others	With Interaction Term
DFI	0.0200 **	0.0227 ***	0.0222 ***
	(0.0081)	(0.0072)	(0.0070)
DFI×			−0.0016 *
High-Tech Dummy			(0.0009)
Constant	1.3142	1.3589	1.4929
	(1.9322)	(1.8665)	(1.7784)
City Controls	Yes	Yes	Yes
Firms Controls	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
R2	0.2003	0.2316	0.2267
Observation	30,181	164,402	194,583

Note: This table shows the results of firms’ technology levels heterogenous effects. For all regressions, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In Column (1), the regression is with the high-tech subsample. In Column (2), the regression is with the non-high-tech subsample. In Column (3), the regression is with an interaction term between the level of digital financial inclusion development and the dummy variable of the industry’s technological level. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. ***, **, and * indicate that the coefficients are statistically significant at the 1%, 5%, and 10% levels, respectively.

, digital financial inclusion development significantly promotes technological innovation in both high-technology firms and other firms. Column (3) is the result after adding the interaction term between the firm’s technology level and the level of digital financial inclusion development, which shows that digital financial inclusion development promotes the innovation capacity of MSMEs, but the coefficient of the interaction term indicates that the high-tech enterprises are at a disadvantage relative to other firms. Firms with higher levels of technology will face higher risks and have greater responsibility and pressure to innovate than other firms; therefore, they may be less likely to obtain financing resources. MSMEs with low or medium levels of technology are more likely to be engaged in more traditional and mature production and are also relatively less exposed to risk and have relatively less difficulty in accessing finance support. While the development of digital inclusive finance has been effective in facilitating new product innovations by firms of all technical levels, it lacks targeted support for enterprises to innovate. MSMEs are the main force of innovation; in the future, digital inclusive finance should provide more convenience for innovative enterprises, especially high-tech enterprises.

5.5.2. Regional Heterogeneity

The development of digital inclusive finance in China has obvious spatial agglomeration and regional heterogeneity. The development of digital finance has eased the differences in financial resources between regions so that the promotion effect on enterprises in regions that are less economically developed should be more significant. For example, the central and western parts of the country, where economic development is lagging behind, are more likely to benefit.

In order to test the heterogeneity of the effect of digital inclusive finance in different regions, we divide the entire sample into two sub-samples, the eastern region and the central and western region, according to where the enterprises are located.

The results in the first column of

Table 8. Regional heterogeneity.

	(1)	(2)
Variables	Eastern Region	Central and Western Region
DFI	0.0088	0.0295 ***
	(0.0087)	(0.0113)
Constant	4.2840 *	−2.8170
	(2.3774)	(1.7712)
City Controls	Yes	Yes
Firms Controls	Yes	Yes
Industry FE	Yes	Yes
Year FE	Yes	Yes
R2	0.2475	0.1850
Observation	146,910	47,673

Note: This table shows the results of regional heterogeneity. For all regressions, the independent variable is the city’s digital financial inclusion index, and the dependent variable is the new product output value of MSMEs. In Column (1), the regression is with the eastern region subsample. In Column (2), the regression is with the central and western region subsample. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** and * indicate that the coefficients are statistically significant at the 1% and 10% levels, respectively.

show that the development of digital inclusive finance in the eastern region has a facilitating effect on the technological innovation of enterprises, but it is not significant. The results in the second column of Table 7 show that the effect of digital inclusive finance in the central and western regions is very significantly positive, which indicates that digital inclusive finance has fulfilled the purpose of alleviating the differences in regional financial resources and supporting the development of enterprises in regions that are less economically developed.

5.5.3. Different Dimensions of Digital Financial Inclusion

The digital financial inclusion index consists of three dimensions, namely, the breadth of digital financial coverage, the depth of digital financial use, and the degree of digitalisation of financial inclusion; each of these three dimensions contains a number of specific indicators. In this part, we further use these three indices to analyse the impact of the different dimensions of digital financial inclusion on enterprise innovation. Among them, the breadth of coverage indicator reflects the extent to which the supply of digital financial services can ensure that users receive appropriate services. Unlike traditional financial institutions, where the ability to reach users is reflected in the number of financial institution branches and financial service staff members, digital finance’s ability to reach customers is reflected in the number of e-accounts. The depth of digital financial usage measures digital financial inclusion from two perspectives: from the perspective of the types of financial services, including payment services, money fund services, insurance services, investment services, and credit services; and from the perspective of the usage, including both the total actual usage indicator (the number of people using these services per 10,000 Alipay users) and the activity indicator (the number of transactions per capita and the amount transactions per capita). The digitisation indicator measures the degree of digitisation of inclusive finance; specifically, a more convenient (e.g., high ratio of mobile payments to total payments), lower cost (e.g., low interest rates on consumer and micro-enterprise loans), and greater credit usage (e.g., high ratio of unsecured payments to total payments) of digital financial services reveal a higher degree of digitisation. Since the development of digital inclusion can be measured in different dimensions, we have further explored the heterogeneous impact of different dimensions of digital inclusion on firm innovation.

Table 9. Heterogeneity of different dimensions.

	(1)	(2)	(3)
Variables	Breadth	Depth	Digitisation
DFI	0.0081	0.0246 ***	−0.0074 **
	(0.0052)	(0.0047)	(0.0036)
City Controls	Yes	Yes	Yes
Firms Controls	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
R2	0.2253	0.2299	0.2253
Observation	194,583	194,583	194,583

Note: This table shows the results of digital financial inclusion’s different dimensions. In columns (1)–(3), the independent variable is the breadth, depth, and digitisation of digital financial inclusion, and the dependent variable is the innovation investment of MSMEs in all regressions. Control variables at the city level include GDP per capita, government expenditure per capita, internet infrastructure density, industrial structure, fixed asset investment per capita and road density per capita, the number of employees, and the government innovation subsidy per capita. Control variables at the firm level include age, the innovation fixed capital investment per capita, the squared variable of age, export capacity, and the degree of competition at the firm level. *** and ** indicate that the coefficients are statistically significant at the 1% and 5% levels, respectively.

presents the heterogeneity effect of different dimensions of digital financial inclusion on firm innovation. Columns (1)–(2) of Table 9 show that digital financial inclusion can have a facilitating effect on firms’ innovation in terms of breadth of coverage and depth of use, but the facilitating effect of depth of use is more significant. Compared to traditional financial services, digital inclusive finance can lower the cost and increase accessibility for low-income groups and MSMEs. Additionally, digital inclusive finance can facilitate financial innovation through digital technology, offer tailored services to meet diverse needs, and enhance the overall usage of financial services. Thus, promoting innovation in MSMEs through digital inclusive finance can be achieved by ensuring both breadth of coverage and depth of use. Column (3) of the table shows that the degree of digitisation may have a negative effect on business innovation, which may be due to the fact that the existing digital financial inclusion has lower costs and higher credit level. As mentioned in many studies, the use of digital finance requires a careful response to risk. While promoting the inclusive nature of digital financial inclusion, it is equally important to enhance credit risk management and prevent individuals and businesses from making uninformed investments. Otherwise, it would not only worsen their financial situation but also be detrimental to the innovative development of enterprises.

6. Conclusions and Recommendations

In October 2023, the Central Financial Work Conference proposed for the first time the strategy of a “strong financial country”. In recent years, China has gradually taken an advantageous position in the international arena in terms of the development of digital technology. Digital inclusive finance is an important form of digital finance and inclusive finance, and the results in this paper show that the development of digital inclusive finance in China has a significant incentive effect on enterprise innovation and plays an important role in promoting the high-quality development of the economy.

The results of this paper demonstrate that the original purpose of digital all-inclusive finance is basically realised. Unlike existing studies, this paper argues that digital inclusive finance works through two mechanisms: alleviating financing constraints and promoting consumption. On one hand, the development of digital inclusive finance eases the financing constraints of enterprises and enhances their innovative capacity. On the other hand, the development of digital inclusive finance expands consumer demand, which is an important incentive for innovation.

With reference to the present findings, we provide some practical guidance for the development of digital inclusive finance:

First, for firms with different levels of technology, digitally inclusive finance has played an important role in fostering innovation in both high-tech and other firms. However, high-tech firms have not shown a stronger facilitating effect relative to other firms. In the future, the precision of digital inclusive finance can be improved to increase support for high-tech firms with relatively higher risks and more serious financing problems, so that digital inclusive finance can better serve the innovative development of MSMEs.

Second, for different regions, we find significant differences in the impact of digital financial inclusion on MSME innovation. In the central and western regions, digital financial inclusion has a significant impact on firm innovation, but in the eastern region, digital financial inclusion does not significantly promote firm innovation. This difference may be due to the lower level of economic development in the central and western regions, which are more dependent on inclusive finance. Many more efforts should be made to develop digital and financial infrastructure in the central and western regions so that more people can gain access to financial services and the development of less economically developed regions can be promoted.

Third, for the different dimensions of digital financial inclusion, continuing to improve the breadth of coverage and depth of use is an effective way for digital financial inclusion to promote innovation in MSMEs. Digital inclusive finance should endeavour to provide more comprehensive financial services to more people, but financial risk is an inevitable problem. The government should help low-income groups and micro-, small-, and medium-sized enterprises (MSMEs) to improve their financial literacy, make reasonable investments, and prevent financial risks to make digital inclusive finance function well.

In China, the development of digital financial inclusion has not been occurring for long, but it has been somewhat effective. Digital finance and financial inclusion are important directions of financial development and necessary elements to improve the financial system. The findings of this paper present some feasible suggestions for the development of digital inclusive finance, which can help regulate digital inclusive financial services and enhance the ability of digital inclusive finance to promote economic development. Digital finance may entail a variety of risks, and further research is needed on how to prevent risks while enabling digitally inclusive finance to promote innovation and economic development.