The Impact of Digital Inclusive Finance on SME Innovation

Abstract

This study analyzes data from companies listed on the Small and Medium Enterprise Board and the Growth Enterprise Market between 2011 and 2021 to assess the impact of digital inclusive finance on SME innovation. It also explores how this effect varies across different industries. The results indicate a significant positive correlation between the digital inclusive finance index and the SMEs’ innovation output at the 1% significance level. Furthermore, improving the coverage breadth and usage depth of digital inclusive finance is more effective in promoting SME technological innovation than enhancing their digitalization level. Additionally, the heterogeneity analysis underscores its significant impact on driving innovation within the secondary sector. Mechanism testing reveals that digital inclusive finance enhances the innovation ecosystem by reducing financing barriers and improving access to capital. The study recommends accelerating the development of a digital inclusive financial system and optimizing the allocation of financial resources. Furthermore, it emphasizes the need to support digital transformation, foster talent development, and implement targeted industrial policies. These efforts are crucial for unlocking the full innovation potential of SMEs and driving high-quality development.

1. Introduction

The Third Plenary Session of the 20th CPC Central Committee emphasized the deep integration of the digital economy and the real economy, underscoring the strategic importance of the digital economy in China’s contemporary economic development. This policy direction also provides robust support for digital inclusive finance as a key driver of SME innovation. The session highlighted the necessity of establishing institutional mechanisms that foster high-quality economic growth and comprehensive innovation. This reflects the urgency of creating a fairer, more efficient financial service environment to support SME innovation amid China’s economic transformation.

Traditionally, SMEs have relied heavily on bank credit for financing. However, factors such as information asymmetry, risk uncertainty, adverse selection, and credit discrimination (Oyegbade et al., 2023) [1] have led traditional financial institutions to impose constraints on SMEs. These constraints significantly limit the capacity for innovation and growth among SMEs. In recent years, digital inclusive finance, powered by technologies like big data, artificial intelligence, and blockchain (Tang et al., 2020) [2], has helped mitigate these barriers by reducing information asymmetry and transaction costs, thereby improving financial resource allocation efficiency. Digital inclusive finance enhances the accessibility, precision, and availability of financial services (Wahlstrøm and Becker, 2023) [3], providing SMEs with lower-cost, higher-efficiency financial solutions. This not only lowers the financial entry barriers for SMEs but also creates a more conducive external environment for innovation.

Despite the significant potential of digital inclusive finance, the specific mechanisms and effects through which it drives SME innovation remain insufficiently explored. While existing studies highlight the role of digital finance in improving financial access and reducing transaction costs, its impact on fostering sustained innovation in SMEs remains uncertain. Therefore, this study aims to fill this gap by examining the empowering effects of digital inclusive finance on SME innovation from three perspectives: theoretical mechanisms, impact assessment, and policy optimization. By focusing on financing constraints, this research offers concrete policy recommendations to guide policymakers in unlocking the full innovation potential of SMEs and promoting their high-quality development.

2. Review of Literature

While considerable research has explored digital inclusive finance in the context of enterprise technology and corporate governance, systematic studies on the innovation mechanisms of SMEs remain limited. Existing literature mainly focuses on the role of digital inclusive finance in promoting SME innovation and growth, but a deeper exploration of its underlying mechanisms and empirical validation is still needed. Research on digital inclusive finance has primarily focused on how digital technologies facilitate the development of inclusive financial systems. Digital inclusive finance represents the integration and evolution of internet technologies within the financial sector, fostering leapfrog progress and inclusive growth in financial services (Assimakopoulos et al., 2025) [4]. The development of digital inclusive finance in China should be guided by new development philosophies to promote high-quality economic growth (Zou et al., 2014) [5]. In terms of aligning digital inclusive finance with the new development paradigm and high-quality enterprise development, the integration of digital technologies and financial services is considered beneficial for advancing inclusive finance alongside SMEs. This integration helps reduce financing costs, expand financial access, deepen financial penetration, and further broaden the SMEs’ funding channels, thereby significantly improving the efficiency of resource allocation and their innovation capacity (Du et al., 2024) [6]. The inclusive nature of digital finance also plays a vital role in extending financial services, optimizing financial structures, and enhancing financial stability. Its unique advantages contribute to expanding the SMEs’ access to finance, improving funding efficiency, and optimizing the distribution of financial resources. The existing literature on this topic primarily focuses on the following four aspects:

Currently, there are primarily two methods for measuring the innovation of SMEs. The first method assesses an enterprise’s early-stage innovation investment in R&D. Common indicators include total R&D investment and its proportion relative to annual revenue (Booltink, 2018) [7]. The second approach examines the firms’ innovation output during advanced R&D stages. Commonly used indicators for this assessment include the proportion of annual revenue generated by new products, the number of patent applications or grants, and similar metrics (Teirlinck et al., 2022) [8].

In China, the disclosure of R&D expenditure is relatively brief and lacks sufficient standardization. Patent count is widely recognized as a key measure of a firm’s innovation achievements. Moreover, the patent application and approval process are transparent, and data collection is relatively straightforward. Thus, many researchers in China and worldwide consider invention patents as the primary metric for evaluating SME innovation. Patent quantity currently serves as a crucial benchmark for evaluating corporate innovation capacity (Ponta et al., 2021; Ma and Yu, 2021; Lu et al., 2022) [9,10,11].

At the level of factors influencing SME innovation, government policy support and the optimization of the institutional environment play a critical role in fostering enterprise innovation. According to signaling theory, government subsidies can significantly boost the firms’ investments in innovation and their output of patents. This incentive mechanism is particularly pronounced in high-tech companies, those with robust internal control systems, and those operating in a favorable legal environment (Sun, 2021) [12]. This study primarily focuses on the incentive mechanisms of high-tech enterprises and proposes strategies to improve these mechanisms. Furthermore, as economic policy uncertainty grows, the risk of bankruptcy increases, leading to a decrease in the firms’ investments in research and development (R&D) (Liu et al., 2022) [13]. Simultaneously, effective intellectual property protection laws, bankruptcy laws, and other legal frameworks can safeguard the investors’ rights and encourage their investment, thereby helping firms secure financial support for R&D activities. Thus, a sound legal and regulatory system is vital for promoting enterprise innovation (Zhao et al., 2022) [14]. Additionally, a competitive environment and a strong financial system foster a high-quality external financing climate for businesses. Financial institutions play a key role in providing enterprises with credit solutions essential for R&D and innovation, thereby supporting the sustained growth of the real economy (Yao and Yang, 2022) [15]. Furthermore, relaxing market access restrictions by banks in other regions can expand credit availability, reduce reliance on loan guarantees, and create more financing opportunities for firms, ultimately enhancing their innovation capabilities (Franquesa and Vera, 2021) [16]. Innovative financial models, such as fintech and digital inclusive finance, have alleviated financing pressures on enterprises by introducing advanced financing mechanisms and optimizing capital allocation, thereby enhancing investment efficiency and returns (Gu et al., 2023) [17]. These models are instrumental in enhancing financial processes and driving corporate innovation.

From a mechanistic perspective, digital inclusive finance promotes SME innovation by mitigating financial inefficiencies that obstruct resource distribution. Li et al. (2022) [18] highlighted that financial constraints often impede the SMEs’ access to funding, affecting their innovation capacity. Digital inclusive finance effectively mitigates these inefficiencies, fostering sustainable corporate growth. Grounded in endogenous finance theory, Zhang et al. (2023) [19] examined how inclusive finance alleviates SME financing constraints. Zheng et al. (2023) [20] developed a theoretical framework of “fintech-financing constraints”, analyzing their empirical impact across multiple dimensions, including dynamic effects, heterogeneity, and macro-micro mechanisms, in fostering corporate innovation. Additionally, based on endogenous growth theory, Li et al. (2024) [21] examined how digital inclusive finance impacts total factor productivity, highlighting critical structural factors that drive innovation. Employing data from firms registered on the New Third Board, Zhang et al. (2023) [19] determined that digital inclusive finance fosters SME innovation through reduced financing expenses. Li et al. (2021) [22] confirmed its contribution to strengthening corporate financial autonomy and resolving funding imbalances.

At the determinant level, Agwu (2021) [23] extensively assessed digital inclusive finance development, formulating an index with three key dimensions: coverage breadth, usage level, and digital support services. Their research analyzed how digital payment solutions fill voids in conventional financial systems across underdeveloped areas, offering significant advantages to SMEs, and highlighted the mechanisms through which digital inclusive finance fosters enterprise innovation. Yu et al. (2020) [24] outlined a framework illustrating how digital inclusive finance stimulates SME innovation through three main pathways: government policies, financial structures, and technological advancements. Lee et al. (2023) [25] conducted a detailed analysis of the mediating role played by financial structure optimization and corporate information transparency in linking digital inclusive finance development to enterprise value creation. Their results underscored the diverse influences of digital inclusive finance on SME innovation. Further, Ma et al. (2023) [26] investigated the interplay between digital inclusive finance, funding limitations, urban prosperity, and corporate green innovation. Employing an instrumental variable regression approach, their study affirmed digital inclusive finance as a key driver of green tech innovation in enterprises, highlighting urban wealth and funding restrictions as pivotal mediators.

However, existing research still presents certain limitations. On the one hand, most studies have primarily focused on the direct effects of digital inclusive finance on enterprise innovation, with fewer exploring the underlying mechanisms or indirect pathways in detail. On the other hand, although some literature examines the independent effects of different dimensions of digital inclusive finance, the relative strength of these effects remains unclear, and there is a lack of research addressing the underlying mechanisms from an industry heterogeneity perspective. Therefore, it is necessary to further clarify the differential impacts of various dimensions of digital inclusive finance on innovation and to conduct in-depth analysis of their specific mechanisms in light of industry characteristics.

In summary, prior studies have significantly deepened insights into digital inclusive finance and its influence on entrepreneurship, financial needs, and economic growth. This research offers key contributions. First, it introduces an effect model to examine how digital inclusive finance drives SME innovation, conducting a thorough theoretical analysis to establish a solid foundation for comprehending its impact on innovation. Second, focusing on SME innovation, this study develops a comprehensive evaluation index and empirically assesses the distinct impacts of digital inclusive finance. Third, this study explores strategies for enhancing SME innovation via various dimensions of digital inclusive finance, offering practical policy suggestions suited to current economic conditions.

3. Theoretical Mechanism and Model

Current studies suggest that advancing digital inclusive finance enhances financial resource availability and efficiency, particularly through broader coverage, deeper utilization, and greater digital adoption. Various financial services alleviate SME financing obstacles, mitigate financial limitations, and stimulate business innovation (Liang, 2018) [27]. This research integrates effective modeling and action mechanisms to explore how digital inclusive finance drives SME innovation, alongside assessing its targeted promotional effects.

3.1. Impact Model of Digital Inclusive Finance on SME Innovation

From a theoretical perspective, this study draws on the Information Asymmetry Theory and Schumpeterian Innovation Theory to explain the mechanisms through which digital financial inclusion promotes SME innovation. In traditional financial systems, SMEs often suffer from incomplete or opaque credit information, making it difficult for financial institutions to accurately assess their risk. This leads to credit discrimination and financing constraints, which suppress firms’ willingness and ability to engage in innovation. Digital financial inclusion leverages technologies such as big data and algorithmic risk assessment to reduce information asymmetry, enhance transparency, and improve the efficiency of financial resource allocation.

To describe SMEs’ decision-making behavior under changing financing conditions, this paper incorporates the von Neumann–Morgenstern utility function and assumes that SMEs are risk-neutral agents. When digital financial inclusion improves the ability of financial institutions to observe and evaluate a firm’s effort level, the probability of successful financing increases. This, in turn, motivates firms to invest in innovative projects, as the expected return becomes more achievable under reduced financing friction. Thus, digital financial inclusion not only eases access to funding but also strengthens the incentive for SMEs to innovate. This process forms a logical chain of “financial empowerment—credit alleviation—innovation activation”, providing a solid theoretical foundation for the subsequent modeling and empirical analysis.

When SMEs undertake innovative projects, financial support is crucial. The von Neumann–Morgenstern utility function assumes that firms are risk neutral, i.e., $U^{\prime }>0,U^{″}=0$, within the analytical framework.

At the first stage, the enterprise secures funding K, while at the second stage, the successful investment generates a return G. The probability of success $\alpha \left(E\right)$ ($0\le E\le 1$; $0\le \alpha \left(E\right)={\alpha }_0+{\alpha }^{\prime }E\le 1$) depends on both objective factors and the enterprise’s effort level (Jiang and Yi, 2022) [28]. The investment cost is given by $\beta E^2/2,\left(\beta >0\right)$, and the probability of investment failure is $1-\alpha \left(E\right)$.

Innovative projects undertaken by SMEs are closely linked to financial support. In this context, the opportunity cost of lending to the investor is l. Due to information asymmetry, investors cannot fully assess enterprise risk and can only obtain partial risk information, denoted as l_i, for enterprise i, where $0\le l_i\le 1$. Based on this risk information, financial institutions assess the likelihood of investment success, denoted as P(l_i).

Initially, SME A seeks a loan from financial institution B₁, using assets worth M as collateral (discount rate $\delta$, where $0\le \delta \le 1$), and pay transaction cost C₀ and loan interest $r_1\cdot K$. At the same time, SME A can also choose to finance from informal financial institution B₂, and the loan interest it needs to pay is $r_2\cdot K$. In addition, formal financial institution B₁ needs to bear the pre-loan examination cost C₁. If the review is passed, the loan is granted; if the review is not passed, the loan is denied. If the loan is rejected, the guaranteed return equals $r_3\cdot K$.

At the second stage, SME A decides on defaulting or repaying. Defaulting will incur losses in areas such as reputation, represented by Q. If the enterprise repays on time, the likelihood of repayment is $\pi$.

In the first stage, SME A determines its effort level e based on the prevailing conditions. In the second stage, financial institution B₁ chooses the best decision $p_1$ from the available set p after assessing the firm’s effort level e.

Assuming that the returns for SME A are $u_1\left(e\right)$, the returns and decision-making process of formal financial institution B₁ can be expressed as:

$$u_2\left(e,p_1\right)$$

(1)

During the execution of this innovation project, given an enterprise’s effort level e, the financial institution determines a distinct optimal solution $u_1\left(e,O_2\left(p_1\right)\right)$. Meanwhile, the SME can anticipate the financial institution’s potential action plan for each e and adjust its optimal response $e^{\ast }$ accordingly to maximize its benefits. Consequently, the SME’s decision to seek funding from informal financial institutions can be expressed as follows:

$$u_1\left(e,O_2\left(p_1\right)\right)$$

(2)

Then $u_1\left(e^{\ast },O_2\left(p_1\right)\right)$ is the optimal solution of Equation (2).

3.2. Dynamics of Investment and Financing in Conventional Financial Markets

Informal financial institutions often impose higher interest rates, denoted as $r_2>r_1$. Consequently, SMEs typically seek financing from formal financial institutions, which offer lower financing costs. This study examines the financing dynamics between SMEs and formal financial institutions, where the expected profit for SMEs is $V_{A_1}$, and the expected profit for formal financial institutions is $V_{B_1}$.

If SMEs cannot obtain financing from formal financial institution B₁, then

$$V_{A_1}=\alpha \left(E\right) \left[G-K\left(1+r_2\right)\right]+\left[1-\alpha \left(E\right)\right] \left[-K\left(1+r_2\right)\right]-C_0$$

(3)

In this case, the expected return of financial institution B₁ is as follows:

$$V_{b_1}=K\left(1+r_3\right)$$

(4)

If SMEs successfully obtain financing from formal financial institution B₁, then

$$V_{A_2}=\pi \alpha \left(E\right) \left[G-K\left(1+r_1\right)\right]+\left(1-\pi \right)\left[1-\alpha \left(E\right)\right] \left[-M-Q\right]-C_0$$

(5)

In this case, the expected return of financial institution B₁ is as follows:

$$V_{A_2}=\pi \alpha \left(E\right) \left[G-K\left(1+r_1\right)\right]+\left(1-\pi \right)\left[1-\alpha \left(E\right)\right] \left[-M-Q\right]-C_0$$

(6)

In general, the expected income of SMEs is as follows:

$$V_{A_i}=\left[1-P\left(l_i\right)\right] \left\{\alpha \left(E\right) \left[G-K\left(1+r_2\right)\right]+\left[1-\alpha \left(E\right)\right] \left[-K\left(1+r_2\right)\right]\right\}+P\left(l_i\right) \left\{\pi \alpha \left(E\right) \left[G-K\left(1+r_1\right)\right]+\left(1-\pi \right) \left[1-\alpha \left(E\right)\right] \left(-M-Q\right) \right\}-C_0-{\displaystyle \frac{1}{2}}\beta E^2$$

(7)

$$V_{B_i}=\left[1-P\left(l_i\right)\right] K\left(1+r_3\right)+P\left(l_i\right) \left\{\pi \alpha \left(E\right) K\left(1+r_1-l\right)+\left(1-\pi \right) \left[1-\alpha \left(E\right)\right] \left(\delta M-K\right)-C_1\right\}$$

(8)

After SMEs secure financing for an innovative project, they can maximize their own benefits E by adjusting their subjective effort level. The constraints are as follows:

$$V_{A_i}\ge 0, G\ge K\left(1+r_2\right), G\ge K\left(1+r_1\right), \left[1-P\left(l_i\right)\right] V_{A_1}+P\left(l_i\right) V_{A_2}\ge {\displaystyle \frac{\beta E^2}{2}}$$

(9)

The optimal condition of effort level can be expressed as follows:

$${\displaystyle \frac{\partial V_{A_i}}{\partial E}}=\left[1-P\left(l_i\right)\right] {\alpha }^{\prime }G+P\left(l_i\right) \left\{\pi {\alpha }^{\prime } \left[G-K\left(1+r_1\right)\right]-\left(1-\pi \right) {\alpha }^{\prime }\left(-M-Q\right)\right\}-\beta E=0$$

(10)

Therefore, the subjective optimal effort level of enterprises can be expressed as follows:

$$E^{\ast }={\displaystyle \frac{\left[1-P\left(l_i\right)\right]{\alpha }^{\prime }G+P\left(l_i\right) \left\{\pi {\alpha }^{\prime } \left[G-K\left(1+r_1\right)\right] \right\}-\left(1-\pi \right) {\alpha }^{\prime }\left(-M-Q\right)}{\beta }}$$

(11)

Since each loan of formal financial institutions needs to meet the condition of non-negative expected returns, it is expressed as follows:

$$V_{B_i}\ge 0, K\ge L_0={\displaystyle \frac{C_1-P\left(l_i\right) \left(1-\alpha \right) \left(1-\pi \right)\delta M}{\left[1-P\left(l_i\right) \left(1+r_3\right)\right]+P\left(l_i\right) \left[\pi \left(r_1-l\right)-1+\alpha +\pi \right]}}$$

(12)

where L₀ represents the minimum loan size that formal institutions are willing to accept. If the loan amount is below L₀, financial institutions will refuse to provide loans to SMEs.

$${\displaystyle \frac{\partial L_0}{\partial M}}=\left\{\begin{array}{l}>0, 0<P\left(l_i\right)<{\displaystyle \frac{-1-r_3}{\pi \alpha \left(r_1-l\right)-2+\alpha +\pi -r_3}}, \mathrm{greate} \mathrm{risks}\\ \le 0,{\displaystyle \frac{-1-r_3}{\pi \alpha \left(r_1-l\right)-2+\alpha +\pi -r_3}}\le P\left(l_i\right)<1, \mathrm{small} \mathrm{risks}\end{array}\right.$$

(13)

In the context of incomplete market information, formal financial institutions face challenges in accurately assessing the true effort level of enterprises. They can only evaluate the risk associated with SMEs and make financing decisions based on the value of collateral. This information asymmetry makes it difficult to effectively address the financing challenges in traditional financial markets and hampers the ability to achieve an accurate balance in loan decisions between formal financial institutions and SMEs. Therefore, a thorough analysis of the effect model is essential to uncover the underlying mechanisms and influences.

3.3. Mechanism Effect

This paper defines the extent of digitalization in digital inclusive finance as a measure of the formal financial institutions’ digital advancement $\gamma \left(\gamma >1\right)$. As digitalization progresses, formal financial institutions gain increased access to information i. Their pre-loan assessment cost $C_1/\gamma$ steadily declines, nearing zero over time. In this scenario, the profit increment of formal financial institutions can be expressed as follows:

$$\Delta \prod =C_1-{\displaystyle \frac{C_1}{\gamma }}$$

(14)

Formal financial institutions evaluate the success likelihood P(l_i) of SME innovation projects. Given that banks approve investments based on a critical probability $p^{\ast }$, the corresponding probability distribution function is $f\left(p^{\ast }\right)$, while its density function is as follows:

$$\dot{p}={\displaystyle \frac{{\displaystyle {\int }_0^{p^{\ast }}p\cdot f\left(p\right) dp}}{{\displaystyle {\int }_0^{p^{\ast }}f\left(p\right) dp}}}$$

(15)

Given that is a linear function of digitization level $\gamma$, its derivative with respect to $\gamma$ is as follows:

$${\displaystyle \frac{\partial \dot{p}}{\partial \gamma }}={\displaystyle \frac{1}{{\gamma }^2}}\left[{\displaystyle \frac{p{C}_1}{\pi \alpha \left(E\right)K\left(1+r_1-l\right)+\left(1-\pi \right) \left[1-\alpha \left(E\right)\right] \left(\delta M-K\right)-C_1/\gamma -K\left(1+r_3\right)}}\right]\times \left[{\displaystyle \frac{f\left(p^{\ast }\right)}{F^2\left(p^{\ast }\right)}}\right]\times \left[p^{\ast }\cdot F\left(p^{\ast }\right)-{\displaystyle {\int }_0^{p^{\ast }}pf\left(p\right)dp}\right]$$

(16)

On account of $p^{\ast }F\left(p^{\ast }\right)={\displaystyle {\int }_0^{p^{\ast }}f\left(p^{\ast }\right) d{p}^{\ast }}$,

$${\displaystyle {\int }_0^{p^{\ast }}f\left(p^{\ast }\right) d{p}^{\ast }}\ge {\displaystyle {\int }_0^{p^{\ast }}pf\left(p\right) dp}$$

(17)

Hence, $p^{\ast }F\left(p^{\ast }\right)\ge {\displaystyle {\int }_0^{p^{\ast }}pf\left(p\right) dp}$,

$$p^{\ast }F\left(p^{\ast }\right)-{\displaystyle {\int }_0^{p^{\ast }}pf\left(p\right) dp}\ge 0$$

(18)

For $p\in [0,p^{\ast }]$ with $F\left(p\right)\ge 0$, it follows that ${\displaystyle \frac{\partial \dot{p}}{\partial \gamma }}\ge 0$.

3.3.1. Analysis of Direct Mechanism

According to Equation (18), as digital inclusive finance expands, formal financial institutions become increasingly likely to offer loans to SMEs. This underscores the essential function of digital inclusive finance in easing SMEs’ financial barriers. The rise in digitalization enhances the SMEs’ access to credit. Higher digitalization levels $\gamma$ enhance information symmetry between financial institutions and SMEs, i.e., $\partial l_i/\partial \gamma \ge 0$. Based on the derivation of $\partial \dot{p}/\partial \gamma \ge 0$, it can be concluded that a higher level of digitalization exerts a positive impact on SMEs.

Accordingly, as bank digitalization advances, financial institutions can assess the success probability $P\left(E\right)$ of innovative projects based on the subjective effort level E of SMEs and make financing decisions accordingly, guiding financing decisions, represented as follows:

$$P\left(E^{\ast }\right)={\displaystyle \frac{\beta E-{\alpha }^{\prime }G}{\pi {\alpha }^{\prime }\left[G-K\left(1+r_1\right)\right]-\left(1-\pi \right){\alpha }^{\prime }\left(-M-Q\right)-{\alpha }^{\prime }G}}$$

(19)

The derivation of the subjective effort level E of SMEs can be obtained as follows:

$${\displaystyle \frac{\partial P}{\partial E}}={\displaystyle \frac{\beta }{\pi {\alpha }^{\prime }\left[G-K\left(1+r_1\right)\right]-\left(1-\pi \right){\alpha }^{\prime }\left(-M-Q\right)-{\alpha }^{\prime }G}}>0$$

(20)

With the progression of digital inclusive finance, the SMEs’ efforts become clearer to financial institutions, reducing adverse selection issues stemming from information asymmetry. Moreover, financial institutions rely less on SMEs’ collateral in financing decisions, thereby reducing borrowing costs.

Employing comparative static analysis (Equation (21)) and the implicit function derivative method (Equation (22)), the link between effort level E and digitization $\gamma$ is further examined.

$${\displaystyle \frac{\partial E}{\partial \gamma }}={\displaystyle \frac{\partial V_{B_i}/\partial \gamma }{\partial V_{B_i}/\partial E}}$$

(21)

$${\displaystyle \frac{\partial V_{B_i}}{\partial \gamma }}={\displaystyle \frac{P\left(E\right)C_1}{{\gamma }^2}}>0$$

(22)

$${\displaystyle \frac{\partial V_{B_i}}{\partial E}}=P^{\prime }\left(E\right) \left\{\pi \alpha \left(E\right)K\left(1+r_1-l\right)+\left(1-\pi \right) \left[1-\alpha \left(E\right)\right] \left(\lambda M-K\right)-{\displaystyle \frac{C_1}{\gamma }}-K\left(1+r_3\right)\right\}+P\left(E\right) \left[\pi {\alpha }^{\prime }K\left(1+r_1-l\right)\right]-{\alpha }^{\prime }\left(1-\pi \right) \left(\delta M-K\right)>0$$

(23)

3.3.2. Analysis of Indirect Mechanism

Therefore, with $\partial E/\partial \gamma >0$, the level of subjective effort exerted by the degree of digitalization is positively associated with SMEs. This indicates that enhanced digitalization makes the efforts of enterprises more easily recognized by formal financial institutions, significantly increasing the likelihood of securing financing. Increased digitalization provides formal financial institutions with more comprehensive information, enabling them to assess the efforts of firms more accurately. As enterprises invest greater efforts, the probability of success for their innovative projects rises accordingly. Given these advantages, formal financial institutions are more inclined to provide financing to enterprises exhibiting higher effort levels, increasing loan approval likelihood.

The effect model indicates that digital information technology enhances traditional financial institutions’ credit processes, mitigating information asymmetry in conventional credit markets. Moreover, digital inclusive finance reduces SMEs’ financing expenses, encouraging higher investment and improving financial institutions’ profitability. As loan yields increase, formal financial institutions exhibit a higher willingness to provide credit support to SMEs.

However, notable variations exist in the profitability of innovation projects across industries. This results in significant variations in how digital inclusive finance influences SME innovation. For instance, in technology-intensive sectors (e.g., information technology and manufacturing), digital inclusive finance is instrumental in driving innovation by overcoming barriers to financing and reducing barriers to technological research and development. In contrast, in labor-intensive industries, where capital demand for innovation is relatively low, the enabling effect of digital inclusive finance remains limited.

Thus, digital inclusive finance is crucial in reducing information asymmetry for SMEs, eliminating loan bias from formal financial institutions, and lowering financing hurdles. Additionally, it exhibits varying impacts across different sectors. Based on these perspectives, the study presents the following hypotheses:

Hypothesis 1.

Digital inclusive finance significantly promotes innovation investment among SMEs, and its different dimensions have distinct effects on innovation.

Hypothesis 2.

Digital inclusive finance supports SMEs’ innovation by reducing financial constraints.

Hypothesis 3.

The effects of digital inclusive finance on SMEs’ innovation investment differ significantly across industry types.

4. Research Methodology

4.1. Data Processing and Interpretation

Firms listed on China’s SME Board and Growth Enterprise Board are generally characterized by small to medium size, high growth potential, and strong technological innovation capabilities. These enterprises typically face more pronounced financing constraints while exhibiting strong demand for innovation, making them ideal subjects for this study. Therefore, this study examines domestic firms registered on the SME Board and Growth Enterprise Board between 2011 and 2021. The data processing procedures include the following (The data processing and analysis software used in this study is Stata 18.0, and the operating system is Windows 11. The device is a Windows 11 laptop made in China):

(1)

Exclusion of financial and real estate enterprises, as their distinct operational models may introduce biases in digital inclusive finance research.

(2)

Removal of enterprises with significant missing data for key variables or those delisted during the study period.

(3)

Elimination of missing values and outliers in financial data to enhance reliability.

(4)

After screening and data refinement, 7700 valid records remain. Key data sources include the Peking University Digital Inclusive Finance Index (2011–2021), the CSMAR database.

Specifically, the variables are sourced as follows:

(1)

Firm innovation output (innov), financing constraint index (SA), firm size (size), firm age (age), financial leverage (lev), growth potential (growth), CEO–Chair duality (dual), proportion of independent directors (indratio), fixed asset ratio (fas), and managerial shareholding ratio (msh) were all obtained from the CSMAR database;

(2)

The digital inclusive finance index (index), along with its sub-dimensions—coverage breadth (cover), usage depth (depth), and level of digitalization (digital)—were sourced from the Peking University Digital Inclusive Finance Index (2011–2021).

4.2. Variable Description

1.

Explanatory variable: This study uses enterprise innovation output (innov) as a measure of innovation capability. While the existing literature commonly assesses corporate innovation from both input and output perspectives, factors such as earnings management and outsourcing may compromise the reliability of innovation input data. Additionally, issues such as statistical errors and missing data often arise in relevant databases. Thus, this study utilizes the natural log of total patents, including inventions, utility models, and designs (plus 1), as a metric for evaluating enterprise innovation output and capability.

2.

This study’s primary explanatory variable is the Digital Inclusive Finance Index, published by Peking University’s Institute of Digital Finance, evaluating digital inclusive finance progress from 2011 to 2021. This index comprises three sub-indices: coverage breadth (cover), usage depth (depth), and degree of digitalization (digital). To ensure dimensional consistency across variables, each sub-index was standardized by dividing it by 100 and used as the final variable in the analysis.

3.

This study employs the financial constraints index (SA) as an intermediate variable, measuring the extent of financial limitations faced by enterprises. Since the index is expressed as a negative value, its absolute value is utilized for analysis, with a higher absolute value signifying more significant financing constraint.

The calculation formula is as follows:

$$SA=-0.737 size+0.043 siz{e}^2-0.04 age$$

(24)

where size and age represent:

(1)

size, measured as the natural log of total assets at year-end.

(2)

age, calculated as the current year minus the establishment year, plus one.

4.

This study includes control variables to evaluate SMEs’ characteristics: enterprise age (age), size (size), and financial leverage ratio (lev), defined as total liabilities divided by total assets at year-end. Additionally, enterprise growth ability (growth) is quantified by the growth rate of operating income, while chairman-CEO integration (dual) is represented as a dummy variable, where 1 indicates role integration and 0 indicates separation. The independent director ratio (indratio) is determined by the number of independent directors relative to the total board members. The fixed assets share (fas) reflects the percentage of total assets comprising fixed assets and depreciation. Lastly, the management shareholding ratio (msh) denotes management’s ownership percentage in total shares.

4.3. Metrological Model Design

This study conducts the LM and Hausman tests, revealing that the fixed effect model aligns well with the dataset’s characteristics and analytical needs. As a result, the paper takes into account both time effects and industry effects, constructing a two-way fixed effect model.

$$inno{v}_{i,t}=\alpha +{\beta }_{1}inde{x}_{i,t}+{\beta }_{2}control+year+ind+{\epsilon }_{i,t}$$

(25)

Here, i represents an enterprise, t denotes time, and $inno{v}_{i,t}$ captures the innovation output of the ith firm in year t; The variable $inde{x}_{i,t}$ quantifies digital inclusive finance development for the ith firm in year t; while control includes the control variables, and c${\epsilon }_{i,t}$ denotes the random disturbance term. Moreover, year reflects time-fixed effects, while ind represents industry-fixed effects, addressing industry heterogeneity and time-varying unobserved factors.

To examine how digital inclusive finance supports SME innovation and growth by easing financing constraints, the following model is introduced:

$$S{A}_{i,t}={\gamma }_0+{\gamma }_{1}inde{x}_{i,t}+{\gamma }_{2}control+year+ind+{\epsilon }_{i,t}$$

(26)

where $S{A}_{i,t}$ serves as the mediating variable, representing the financing constraints experienced by the ith firm in year t.

4.4. Empirical Examination

4.4.1. Summary Statistics

Table 1. Summary statistics of key variables.

Variable	Observations	Mean	SD	Min	Max
innov	7700	0.880	0.612	0.000	2.183
index	7700	2.558	1.072	0.180	4.590
cover	7700	2.373	1.046	0.020	4.330
depth	7700	2.636	1.118	0.070	5.110
digital	7700	3.025	1.240	0.080	4.620
sa	7700	3.721	0.236	2.974	4.751
size	7700	8.058	0.967	5.878	12.206
age	7700	23.369	4.471	15.000	43.000
lev	7700	0.355	0.186	0.051	0.888
growth	7700	0.189	0.474	−0.949	5.046
dual	7700	0.349	0.477	0.000	1.000
indratio	7700	37.673	5.432	25.000	66.670
fas	7700	0.100	0.086	0.002	0.616
msh	7700	20.685	20.611	0.000	89.725

′s descriptive analysis indicates that the average enterprise innovation output (innov) is 0.880, ranging from 0 to 2.183, demonstrating considerable differences in innovation capabilities among sampled firms. The digital inclusive finance index (index) averages 2.558 with a standard deviation of 1.072, highlighting differences in development among firms.

For control variables, the mean enterprise age (age) is 23.369 years, with an SD of 4.471, suggesting that most firms are well-established, with their ages concentrated within a narrow range. The minimum enterprise growth capacity (growth) is −0.949, signifying operational challenges and revenue fluctuations for certain firms. Additionally, the mean management shareholding ratio (msh) is 20.685, with an SD of 20.611 and a maximum nearing 90%, implying variation in ownership structures and governance models.

Overall, the distribution of the sample data is statistically reasonable, ensuring its suitability for subsequent empirical analysis.

4.4.2. Baseline Regression Result

Table 2. Digital inclusive finance and SME innovation: benchmark regression estimates.

	(1)	(2)	(3)	(4)	(5)
index	0.095 *** 1	0.083 ***
	(0.026)	(0.026)
cover			0.062 ***
			(0.024)
depth				0.063 ***
				(0.019)
digital					0.080 *
					(0.043)
_cons	0.309 ***	−0.004	0.003	−0.001	−0.011
	(0.072)	(0.114)	(0.114)	(0.114)	(0.114)
Control	No	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes
N	7700	7700	7700	7700	7700
R2	0.094	0.111	0.111	0.111	0.110

¹ t statistics in parentheses * p < 0.1, *** p < 0.01.

′s analysis shows that the index remains significant at the 1% level, even after incorporating control variables (as seen in Model 1), with all coefficients displaying positive effects. Models (2–5) examine three key facets of digital inclusive finance separately: —cover, depth, and digital. Findings suggest that cover and depth play the most crucial roles in driving corporate innovation, whereas digital exerts a comparatively weaker positive effect. However, it remains significant at the 10% level. Furthermore, expanding coverage and enhancing usage prove more influential in fostering innovation than merely increasing digitalization levels. Therefore, Hypothesis 1 is supported.

4.4.3. Robustness Test

This paper will conduct robustness tests using three methods: variable substitution, sample selection, and changes in estimation techniques.

• The substitution variable replaces the dependent variable with corporate profitability (roa), a crucial metric evaluating a firm’s net profit generation capability. roa captures how digital inclusive finance supports enterprise growth in terms of economic performance. Digital inclusive finance eases financing constraints, lowers transaction costs, and encourages innovation investment, leading to better resource allocation, efficient asset use, and higher profitability. Unlike corporate innovation output as an indirect factor, roa directly assesses digital inclusive finance’s impact on corporate profitability. This approach broadens the research scope from an innovation-centric view to an economic benefits perspective, offering robust empirical support for policy and practical improvements. The regression results using this substitution variable are presented in

  Table 3. Digital inclusive finance and SME innovation: robustness test.

  	(1)	(2)	(3)	(4)
  index	0.007 *** 1
  	(0.002)
  cover		0.007 ***
  		(0.002)
  depth			0.005 ***
  			(0.002)
  digital				−0.014 ***
  				(0.004)
  _cons	−0.096 ***	−0.096 ***	−0.096 ***	−0.089 ***
  	(0.010)	(0.010)	(0.010)	(0.010)
  Control	Yes	Yes	Yes	Yes
  Industry	Yes	Yes	Yes	Yes
  Year	Yes	Yes	Yes	Yes
  N	7000	7000	7000	7000
  R2	0.257	0.257	0.257	0.258

  ¹ t statistics in parentheses *** p < 0.01.

  . The content in parentheses in Table 3 represents standard errors, which measure the precision of the regression coefficients. The values in parentheses in Table 3 are all below 1, indicating high precision and robust results.

Table 3 illustrates that digital inclusive finance, along with its individual components, plays a crucial role in shaping corporate profitability (roa). Cover and depth exhibit positive effects, with coefficients of 0.007 and 0.005, both statistically significant at the 1% level. However, the coefficient of the digitalization level (digital) is −0.014 and statistically significant at the 1% level, indicating that higher levels of digitalization may exert a negative impact on firms’ short-term profitability. A possible explanation is that the digital transformation process often involves substantial upfront capital investments, such as the construction of IT infrastructure, development and maintenance of information systems, and training of employees in digital skills. These initial expenditures significantly increase operating and fixed costs in the short term, thereby compressing profit margins and reducing roa. Although digital transformation can enhance the firms’ productivity, innovation capacity, and market competitiveness in the long run, it is difficult to achieve a balance between investment returns and cost inputs in the short term. This negative effect reflects the time lag and inherent tension between short-term financial pressure and long-term strategic benefits in the implementation of digital strategies. Therefore, it is crucial to focus on how to effectively alleviate the financial burden faced by enterprises during the early stage of digital transformation, helping them overcome transitional cost bottlenecks and fully realize the long-term development potential of digital inclusive finance.

2.

Wind-down treatment (

Table 4. Robustness test: core explanatory variables lagged by one period.

	(1)	(2)
lindex	0.263 *** 1	0.280 ***
	(0.076)	(0.076)
	0.791 ***	0.852 ***
_cons	(0.171)	(0.224)
	0.263 ***	0.280 ***
Control	No	Yes
Industry	Yes	Yes
Year	Yes	Yes
N	7000	7000
R2	0.049	0.057

¹ t statistics in parentheses *** p < 0.01.

): To address potential reverse causality, where SME innovation might impact digital inclusive finance, the study introduces a one-period lag in the key explanatory variables. Even with this adjustment, digital inclusive finance continues to have a distinct and significant impact on SME innovation. The values in parentheses in Table 4 represent standard errors, all of which are less than 1, indicating that the experimental error is small.

3.

Alternative estimation method: This study employs enterprise innovation output (innov) as a proxy for innovation capability.

Table 5. Robustness test: Tobit model test.

	(1)	(2)	(3)	(4)
index	0.119 *** 1
	(0.036)
cover		0.091 ***
		(0.033)
depth			0.087 ***
			(0.025)
digital				0.116 *
				(0.059)
_cons	−0.274 *	−0.264 *	−0.269 *	−0.283 *
	(0.160)	(0.160)	(0.160)	(0.161)
Control	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	7700	7700	7700	7700
R2	0.053	0.053	0.053	0.052

¹ t statistics in parentheses * p < 0.1, *** p < 0.01.

results validate the stability of the baseline regression analysis. The values in parentheses in Table 5 represent standard errors, all of which are less than 1, indicating that the experimental error is small.

4.4.4. Mediating Effect Test

Table 6. Mediation effect test: Financing constraints.

	(1)	(2)	(3)	(4)	(5)
index	0.1161 *** 1
	(143.71)
innov		−0.0094 ***
		(−3.95)
cover			0.1175 ***
			(140.36)
depth				0.1034 ***
				(119.07)
digital					0.0927 ***
					(115.09)
_cons	2.3470 ***	2.0844 ***	2.3577 ***	2.3065 ***	2.2943 ***
	(224.64)	(105.39)	(221.51)	(194.51)	(190.61)
Control	Yes	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes
N	7700	7700	7700	7700	7700
R2	0.9201	0.7062	0.9173	0.8965	0.8931

¹ t statistics in parentheses *** p < 0.01.

′s analysis highlights how digital inclusive finance mitigates corporate financing limitations. Model (1) shows that digital inclusive finance effectively alleviates SME financing difficulties, with a 1% significance level. Model (2) demonstrates that financing constraints at the same significance level hinder SME innovation. Models (3–5) indicate that broader coverage, increased utilization depth, and digitalization of inclusive finance significantly lessen financing barriers. The results show that digital inclusive finance improves financing conditions by broadening service access, strengthening business–financial institution ties, and enhancing financial service efficiency. The values in parentheses in Table 6 represent standard errors, all of which are less than 1, indicating that the experimental error is small.

Mechanistically, expanding digital inclusive finance eases SME financing constraints via multiple channels. Broadening service coverage improves financial resource accessibility, enabling more enterprises to secure formal funding. The increased depth of use fosters long-term collaboration between firms and financial institutions, improving financial accessibility and convenience. Furthermore, advancements in digitalization optimize credit approval processes, minimize information asymmetry and decrease financing costs. Overall, digital inclusive finance is essential in broadening access to financial services, enhancing resource allocation efficiency, and optimizing the financing environment. Hypothesis 2 is confirmed.

4.4.5. Heterogeneity Test

Examining how digital inclusive finance influences SME innovation under varying intra-industry competition highlights the diverse profitability of innovation projects across sectors. The likelihood of financial institutions granting SME loans depends on their evaluation of project success, which varies accordingly.

Table 7. Heterogeneity analysis.

	(1)	(2)	(3)
	primary industry	secondary industry	tertiary industry
index	0.100	0.082 *** 1	0.073
	(0.511)	(0.029)	(0.062)
_cons	1.152	0.331 ***	1.406 ***
	(1.315)	(0.104)	(0.214)
Control	Yes	Yes	Yes
Industry	Yes	Yes	Yes
Year	Yes	Yes	Yes
N	73.000	6151.000	1476.000
R2	0.263	0.046	0.075

¹ t statistics in parentheses *** p < 0.01.

explores how digital inclusive finance differently impacts SME innovation. The values in parentheses in Table 7 represent standard errors, all of which are less than 1, indicating that the experimental error is small.

The digital inclusive finance index positively influences enterprise innovation in the secondary sector, showing a coefficient of 0.082 at a 1% significance level. However, it does not exhibit a notable effect on primary and tertiary industries. This finding highlights the sectoral differences in how digital inclusive finance impacts innovation capabilities. From a mechanistic standpoint, the secondary industry, which is mainly driven by manufacturing, is characterized by high capital and technological requirements, making enterprises within this sector more reliant on external financing for innovation. Conversely, the primary industry, reliant on natural resources and infrastructure, has a low demand for technological innovation, restricting digital inclusive finance’s capacity to foster innovation. In the tertiary sector, the service industry employs a flexible innovation approach with minimal technological investment needs, leading to more diverse financing options. These findings support Hypothesis 3.

5. Conclusions

This study investigates the impact of digital inclusive finance on SME innovation, using data from domestic firms listed on the SME and Growth Enterprise Boards (2011–2021), alongside records of digital inclusive finance. Through both theoretical and empirical approaches, the key findings are as follows:

Firstly, digital inclusive finance plays a significant role in boosting SME innovation. Among its various components, broader coverage and more extensive utilization have a greater effect on innovation than digitalization alone. Secondly, digital inclusive finance alleviates financial constraints for SMEs, providing essential support for innovation. By expanding coverage, increasing usage, and enhancing digitalization, it improves financing conditions and simplifies access to funds for innovative projects. Thirdly, the effect of digital inclusive finance on SME innovation varies across industries, with the most pronounced impact observed in the secondary sector, where it surpasses its influence on the primary sector.

Although this study systematically investigates the theoretical mechanisms and effects through which digital inclusive finance supports SME innovation, certain limitations should be acknowledged. First, the dataset covers the period from 2011 to 2021. While firm-level financial data for 2022 is available, the most recent digital inclusive finance index from Peking University extends only to 2021, limiting the timeliness of the analysis. Second, this study primarily uses panel regression models. Despite conducting multiple robustness tests, issues related to omitted variables and endogeneity remain, which may affect the validity of the results.

Future research could address these limitations in several ways. With the release of updated digital inclusive finance index data, subsequent studies could incorporate more recent datasets to enhance the practical relevance of the conclusions. Additionally, future studies could employ advanced causal inference methods, such as instrumental variables, difference-in-differences, or other natural experimental approaches, to strengthen causal identification and improve empirical robustness.

6. Recommendations

6.1. First, Enhance the Digital Inclusive Finance Service System for Precise Financial Resource Allocation

The government should accelerate the establishment of a more efficient and targeted digital inclusive finance system, significantly broadening financial service coverage across regions and industries. Specifically, policymakers should introduce tailored measures to enhance financial access in underserved sectors and geographic areas. Financial institutions must be encouraged to develop digital finance products and services customized for SMEs, including digital credit, online supply chain finance, and data-driven risk management solutions. Additionally, building a digital resource allocation platform based on big data and artificial intelligence is critical to precisely match financial resources with SME requirements. This strategy ensures equitable and efficient allocation of financial resources, thus promoting sustainable innovation among SMEs.

6.2. Second, Strengthen Support for SMEs’ Digital Transformation to Alleviate Initial Financial Pressures

Policymakers need to establish concrete mechanisms for cost alleviation, such as dedicated funding programs, tax incentives, and technology-upgrade subsidies, thereby effectively reducing SMEs’ initial financial burdens during digital transformation. Furthermore, specialized digital service providers should receive policy support to offer SMEs professional technical consulting, digital strategy design, and implementation assistance. These measures will significantly reduce SMEs’ trial-and-error costs and accelerate their transformation. Concurrently, SMEs should enhance their internal digital talent development through clear training and evaluation frameworks supported by government initiatives. This approach ensures that the processes and outcomes of SMEs’ digital transformations are measurable, traceable, and ultimately lead to substantial improvements in technological innovation capabilities.

6.3. Third, Implement Targeted Industry-Specific Support Strategies to Unlock SMES’ Innovation Potential

The government must formulate detailed support policies aligned with specific industry characteristics. For the primary industry, targeted financial resources should prioritize agricultural digital technology R&D and application, including digital agricultural infrastructure, training, and demonstration bases. For the secondary industry, governments should facilitate deeper collaboration between financial institutions and high-tech manufacturing enterprises, focusing on R&D financing, digital production line upgrades, and smart manufacturing technologies. These measures support the manufacturing enterprises’ transition from traditional production methods to intelligent manufacturing. For the tertiary industry, policy initiatives should emphasize fintech, e-commerce, and digital creative industries, encouraging innovative digital business models and industry integration. Such targeted measures foster a comprehensive innovation ecosystem, activating the innovation potential across industries and driving high-quality economic growth.