Due to the complexity and uncertainty of innovation activities, firms are unable to obtain all the information and knowledge needed for innovation from internal sources alone and are forced to realize innovation through wider external cooperation (Escribano et al., 2009). As a result, a network of firms’ technological innovation cooperation is formed through the continuous exchange of information and resources between firms, with each firm in the network in a different position and with different resources under its control. Being at the center of the network helps firms to access the information and resources they need for their innovation activities, which in turn helps them to innovate.

The main variables that measure the location of a firm’s network are centrality and structural holes. Analyzing business networks entails looking at and identifying networks made up of companies connected via alliances, partnerships, and cooperative efforts. In this study, the term of centrality refers to valuing a company’s significance in the larger network and its ability to exert control over essential resources required for the network’s operation. (Wasserman and Faust, 1994). The higher the centrality, the more the firm is at the core of the network, the more resources and information the firm has access to, and the stronger the firm’s ability to learn, absorb and transform new information and knowledge, which is conducive to promoting technological innovation (Tsai, 2001). Secondly, firms with higher centrality possess a greater number of information channels and sources, enabling them to effectively mitigate information asymmetry between the firm and stakeholders through the comparison of diverse information. This conducive scenario further enhances the prospects of technological innovation within the firm. Finally, enterprises with high centrality can establish links with more enterprises, which is more helpful for enterprises to select appropriate partners and promote innovation through mutual learning and cooperation (Larcker et al., 2013). Structural holes refer to voids in the network framework that arise between nodes possessing information and resources, yet lacking direct connections or interrupted relationships. These gaps serve as a vital determinant of a firm’s network positioning. Firms located in structural holes are more conducive to technological innovation because they play a central role in information exchange and have the advantage of access to and control over information (Burt, 1997). First, firms occupying structural holes have non-redundant and heterogeneous linkages and access to differentiated information domains that can be filtered and integrated for innovation (Kwon et al., 2020). Additionally, firms occupying structural holes have a greater propensity to discern the qualifications of potential trading partners and collaborators, as well as swiftly grasp opportunities or threats. This ability aids in circumventing innovation failures and augmenting the likelihood of innovation success (Uzzi, 1997). Finally, maintaining ties with many firms is costly, and firms that eliminate redundant ties can devote their limited managerial effort to the most valuable ties, which is beneficial for the efficiency of collaborative innovation (Gnyawali and Madhavan, 2001; Soda et al., 2004).

Favorable ESG performance has the potential to enhance a company’s network position. On one hand, it can boost the company’s reputation, which, to some extent, signifies its standing and influence within the network. The higher the status and influence, the more important the company’s position in the network. Alternatively, favorable ESG performance communicates to the market and stakeholders the company’s dedication to the principles of sustainable development. This highlights the company’s focus on environmental conservation, social accountability, and corporate governance, resulting in heightened societal recognition. The higher the social recognition, the more the company is at the core of the network, and the more power the company has to control resources in the network (Wang et al., 2014). Powell et al. (2005) and others found that core companies with obvious advantages control the corporate network, and their network control power is brought about by their network position. Therefore, good ESG performance can help improve the firm’s position, influence and control over resources, i.e., increase the firm’s network position.

Institutional investors, with their expertise and information advantages, have gradually become a major investor in the capital market. Compared with individual investors, institutional investors have more initiative to participate in corporate governance and are more able to influence firms’ technological innovation decisions and actions and reduce short-sighted behavior (Jiang and Yuan, 2018). Institutional investors are able to gather more information about the firm and increase the transparency of corporate information, thereby alleviating corporate financing constraints to some extent (Cornett et al., 2007). Simultaneously, institutional investors, in their capacity as shareholders, can furnish ample financial support for firms’ decisions and endeavors related to technological innovation.

Good ESG performance helps increase institutional shareholdings. Companies with good ESG performance tend to be larger, faster growing and more profitable companies, which are more preferred by institutional investors (Lin et al., 2014); optimal ESG performance augments the transparency of corporate disclosures and diminishes information asymmetry with the market, making them more appealing to institutional investors (Leuz et al., 2009; Bushee et al., 2014; McCahery et al., 2016); ESG performance can have a significant impact on company performance, which is also reflected in the stock market, making it easier to attract institutional investors to hold company shares.

Drawing upon the analysis presented above, the subsequent hypothesis is formulated:

H2a: Favorable ESG performance has the potential to enhance firms’ technological innovation by augmenting their network positioning.

H2b: Favorable ESG performance can stimulate firms’ technological innovation by augmenting institutional shareholding.

Due to the high investment, risk, and long return cycle associated with innovation activities, companies facing high financing constraints may experience greater financial pressure and reduce their innovation inputs, ultimately inhibiting their technological innovation. Good ESG performance can alleviate financing constraints and promote technological innovation. In the first instance, as a form of non-financial information that is disseminated to external entities, ESG has the potential to mitigate information disparities between firms and markets. This can subsequently contribute to a decrease in external financing costs and assuage financing constraints experienced by firms, ultimately exerting a favorable impact on technological innovation (Dhaliwal et al., 2012; Zhang and Lucey, 2022). Secondly, good ESG performance can demonstrate corporate social responsibility, enhance corporate reputation, and build a positive corporate image. This can help to secure financial support from stakeholders, alleviate financing constraints, and promote technological innovation within the enterprise. Consequently, strong ESG performance assists companies in alleviating their financing constraints and securing external financial backing, thereby exerting a positive influence on corporate technological innovation (Zeng, 2018).

Favorable ESG performance is likely to lead to an escalation in the labor costs incurred by a company. First, in the social responsibility aspect of ESG, human resource management and development and employee health and safety are important manifestations of corporate social responsibility. Second, in the corporate governance aspect of ESG, wages, dividends and benefits are important components of corporate governance performance. For companies, good ESG performance must be achieved by continuously improving employee benefits and promoting human resource development, which will inevitably lead to rising labor costs. Rising labor costs, in turn, generate factor substitution effects, human capital enhancement effects, and innovation compensation effects that encourage firms to innovate technologically (Acemoglu, 2010; Antonelli and Quatraro, 2014). First, when labor costs rise, firms will choose to use capital and technology instead of labor, generating a factor substitution effect that promotes firms’ technological innovation. Second, rising labour costs lead to higher employee compensation. This in turn has a dual effect: it attracts more innovative employees to the firm and encourages the firm to improve talent training, thereby increasing the innovation efficiency of employees. (Hutchens, 1989). Finally, rising labor costs cause firms to increase production costs and reduce profits, which to some extent stimulates firms’ demand for new technologies and compensates for lower profits caused by rising labor costs through technological innovation (Li J. et al., 2020). Therefore, good ESG performance leads to higher labor costs and thus promotes technological innovation in firms.

Drawing upon the analysis presented above, the subsequent hypothesis is formulated:

H3a: Good ESG performance can promote firms’ technological innovation by alleviating financing constraints.

H3b: Good ESG performance leads to an increase in labor costs and thus promotes corporate technological innovation.

The specific mechanism of action is shown in Figure 1.

In line with the methodology adopted by Hsu et al. (2014), this study employs the count of patents granted to firms as a proxy for measuring the technological innovation output and evaluation of the level of innovation exhibited by enterprises. Because the number of patents granted is highly right-skewed, and if the number of patents granted is 0, it will cause missing values, so this study adopts the number of patents granted plus 1, and then takes the logarithm as the measure of enterprise technological innovation.

At present, scholars primarily rely on third-party rating agencies to assess the ESG performance of corporations. However, this study utilizes the Bloomberg ESG score as the primary explanatory variable to assess ESG performance in corporations. The Bloomberg ESG score comprises three crucial dimensions, namely, environmental, social, and corporate governance indicators, making it a comprehensive measure. A higher score denotes better fulfillment of relevant responsibilities by the company. This indicator provides a comprehensive and objective reflection of a company’s fulfillment of its environmental, social and corporate governance responsibilities.

Enterprise network location is measured by degree centrality and structural holes (SH), the measure of centrality is degree centrality (Centrality), this study draws on the research of Freeman (1978), Burt (1992), degree centrality and structural holes of the calculation formula is as follows: C e n t r a l i t y i = 1 n − 1 ∑ i ≠ j θ i j (2) where θ i j denotes whether or not there is a chain director relationship between node firms i and j. If there is a chain director relationship, it is assigned a value of 1, and if there is no chain director relationship, it is assigned a value of 0. n denotes the number of firms in the chain director network. S H i = 1 − ∑ j P i j + ∑ q P i q P q j 2 (3) where q ≠ i , j ; P i j is the strength of the direct relationship from firm i to firm j; P i q P q j denotes the strength of the indirect relationship from firm i to firm j through firm q; and ∑ q P i q P q j denotes the sum of all the indirect relationships between firm i and firm j. The larger the difference, the richer the structural hole.

Following Fang and Na (2020) and Li and Lu (2015), this study uses institutional investors’ holdings in listed firms to denote institutional shareholding.

Following Hadlock and Pierce (2010), this study constructs and measures the SA index using firm age (Age) and firm size (Size). Financing variables with endogenous characteristics are excluded to avoid endogenous interference and to better measure the degree of financing constraints of firms. Based on the study of Longsheng and Hui (2022), this study uses the logarithm of the absolute value of the SA index (Fc) to measure firms’ financing constraints. The formula for calculating the SA index is as follows Hadlock and Pierce (2010): S A = − 0.737 × S i z e + 0.043 × S i z e 2 − 0.04 × A g e (4)

Labor cost refers to the expenses paid by enterprises (or all other types of units and institutions) for employing social labor, and its statistical accounting scope is theoretically larger than wages. As the definition and statistical scope of other forms of labor cost are not clear, this study makes reference to the investigation conducted by Xia and Dong (2014) and employs the average remuneration of employees as a metric to gauge labor costs.

Size: The nature of R&D necessitates substantial, long-term investments. Large enterprises, with their inherent scale advantages, tend to enjoy more stable cash flows and possess greater access to financing options. Consequently, they are better equipped to allocate resources towards innovative projects. For the purpose of this study, we adopt the natural logarithm of total assets as a metric to capture the essence of size.

Age: Mature enterprises often accumulate rich experiences in innovation, leading to reduced costs and a more proactive approach towards R&D. Conversely, some argue that as companies age, they may become complacent about past successes and lose sight of the importance of ongoing innovation. Therefore, controlling for the age of the enterprise is crucial. For our analysis, we consider the number of years since the company’s establishment as a metric of age.

Growth: Growth represents a company’s potential for future expansion and its ability to identify and capitalize on investment opportunities. From a value perspective, growth signifies a company’s capacity to generate outsized returns and its potential worth as a going concern. In this study, we employ sales revenue growth rate as a proxy for growth.

Financial Leverage (Lve): The equipment, personnel, and results of R&D activities possess unique characteristics that render their conversion costs substantial. Given the substantial nature of R&D investments and the associated risks, creditors often impose stringent conditions on the funding they provide to protect their interests. Consequently, lower financial leverage is more favorable for technological innovation investment. High levels of debt can have a negative impact on innovation investment decisions. For this study, we use total liabilities/total assets as a proxy for financial leverage.

Cash Flow (CF): Sustaining R&D activities requires significant cash flow investments. Firms without sufficient cash flow may struggle to maintain their R&D efforts. Therefore, adequate net cash flow is a crucial factor for firms to engage in R&D activities. We measure cash flow using the natural logarithm of net cash flow from operating activities.

Fixed Asset Share (Fasset): A higher fixed asset ratio indicates less liquid assets and potentially lower investments in innovation. This, in turn, may limit funds available for innovative R&D activities. To capture this aspect, we use net fixed assets/total assets as a metric for the fixed asset share.

Shareholding Concentration (Top1): A higher shareholding concentration ratio signifies that the primary shareholder holds a disproportionately large stake in the enterprise. This concentration of power not only enhances their voice in management and operational decisions but also grants them a greater say in strategic choices. In their pursuit of maximizing their own interests, these shareholders may be inclined to avoid high-risk, high-reward endeavors like innovation and research and development due to their long investment cycles and inherent uncertainties. Such a mindset often leads to more conservative investment strategies. This study employs the proportion of shares held by the largest shareholder as a metric to assess shareholding concentration.

Return on Total Assets (Roa): Return on Assets (ROA) serves as a proxy for the profitability of an enterprise. A higher ROA indicates that the enterprise is generating sufficient profits, thereby providing low-cost internal financing for innovative activities. This, in turn, not only supports the company’s internal financing needs but also sends positive signals to external financiers, making it easier for the enterprise to secure external financing for its innovative pursuits. In this study, we use net profit/total assets as a measure of Return on Total Assets.

In the benchmark regression, this study adopts the panel fixed effects model for regression. The chosen explanatory variable in this study is the number of patents granted. Due to the nature of patenting processes, it is possible to observe a significant number of zero values. To address this issue, we employ the Tobit model as an appropriate statistical framework. This study adopts the Tobit model to test whether the regression results of the above fixed effects model are robust. The results from analyzing column (1) of Table 3 show that a positive correlation between sound ESG performance and firms’ technological innovation still exists and is statistically significant. This indicates that the key findings and conclusions derived from this study are durable and dependable, offering convincing proof to affirm that effective ESG performance can propel technological innovation within companies.

By utilizing the total count of patents granted by firms to measure technological innovation, this study aims to capture the breadth and diversity of technological advancements achieved by different firms in the regression. Patents can be categorized into three types: invention patents, utility model patents, and design patents. Among these, invention patents serve as a measure of the innovation quality of enterprises, providing a more comprehensive reflection of their substantive technological innovation. Consequently, this study utilizes the count of invention patents granted by enterprises, rather than the overall count of patents, as the explanatory variable to assess the resilience of the aforementioned regression findings. Building on the studies by Tan et al. (2014) and Tong et al. (2014), this study utilizes a natural logarithmic transformation of the total number of patents awarded to companies, with the addition of 1. The findings presented in column (2) of Table 3 demonstrate that positive ESG performance continues to play a vital role in driving firms’ technological innovation, which reinforces the robustness of the central conclusions drawn in this study.

In this study, the Huazheng ESG rating (Hzesg) for the sample companies is employed as an alternative core explanatory variable in robustness testing, replacing the Bloomberg ESG rating. Here, Huazheng ratings are assigned values: 9 levels from C to AAA are assigned values from 1 to 9. The results of the regression analysis, as reported in column (3) of Table 3, demonstrate that the positive relationship between good ESG performance and companies’ technological innovation remains statistically significant. These findings provide convincing evidence in support of the primary conclusions of the study, hence attesting to the robustness and dependability of the study’s principal outcomes.

Due to the severe impact of the new crown epidemic in 2020 and 2021, the production and business environment of firms may have changed dramatically, and ignoring this factor may affect the robustness of the regression results. In this study, based on the study of Cao et al. (2023), we exclude the impact of the epidemic factor, limit the sample interval to 2011–2019, and re-run the regression, and the results are shown in column (4) of Table 3, which shows that the role of good ESG performance in promoting technological innovation of enterprises is still significant, suggesting that the core conclusions of this study are robust.

Since the economic development level of the four first-tier cities (Beijing, Shanghai, Guangzhou, and Shenzhen) is significantly different from that of other cities, and the economic development level of the region where firms are located significantly influences their business environment, this study adopts the study by Yu et al. (2018) and re-runs the regression after removing the sample of enterprises in the four first-tier cities, and the results are shown in column (5) of Table 3. The results show that the role of good ESG performance in promoting technological innovation of enterprises remains significant, indicating that the core findings of this study are robust.

To reduce the likelihood of omitted variables or unobservable effects in the model, this study implements the methods outlined in Tang et al. (2020), including controlling for “time-industry” joint fixed effects at a higher order, controlling for time-varying industry-level unobservable variables, and re-estimating the regressions. The regression results, as shown in column (1) of Table 4, provide further evidence supporting the enduring importance of favorable ESG performance in fostering technological innovation within firms. These findings reinforce the strength and validity of the fundamental conclusions drawn in this study.

The regression outcomes from the benchmark analysis indicate a positive relationship between a company’s ESG performance and its level of technological innovation. Specifically, as a company’s ESG performance improves, there is a corresponding increase in its level of technological innovation. However, it is also possible that the issue of endogeneity resulting from bidirectional causality emerges due to the fact that firms exhibiting high levels of innovation tend to be more proactive in enhancing their ESG performance. In order to mitigate the two-way causality problem, this study draws on the study of Longsheng and Hui (2022) and selects the one-period lagged term (L_Esg) and the two-period lagged term (L2_Esg) of the explanatory variable (Esg) as the explanatory variables, and re-runs the regression. The regression results, as shown in columns (2)–(3) of Table 4, show that the coefficients of L_Esg and L2_Esg are still significantly positive, indicating that the key findings of the study remain robust.

This study employs the mean (Esg_mean) of the ESG performance of other listed companies within the province of the focal enterprise as an instrumental variable. The use of 2SLS for instrumental variable regression allows us to account for potential endogeneity issues in our analysis. Notably, the instrumental variable is chosen based on the premise that other listed companies in the same provincial location as our sample enterprises share a similar institutional environment. This correlation between instrumental variables and the institutional environment within the same province ensures that the chosen instrumental variable is exogenous to the technological innovation of our sample firms. Crucially, the ESG performance of these other firms within the same province is assumed to have no direct impact on the technological innovation of our sample firms, further reinforcing the exogenous nature of our instrumental variable. To address this relationship, the instrumental variable approach is employed to address potential endogeneity issues and establish a causal relationship between a firm’s ESG performance and its financial performance in this study. The two-stage least squares (2SLS) approach is employed to conduct instrumental variable regression. Table 4, column (4) presents the results of the first stage regression, the coefficient of the instrumental variable (Esg_mean) is significantly positive at 1% level, indicating that the selected instrumental variable is highly positively correlated with ESG performance. Column (5) presents the results of the second stage regression based on Cragg-Donald Wald F = 335.335, {S-Y10% critical value: 16.38}; Kleibergen-Paap rk LM = 78.578, p-value less than 0.001 and Kleibergen-Paap rk Wald F = 117.543, {The results of statistics such as S-Y10% critical value: 16.38} indicate that the weak instrumental variables test is passed and there is no over-identification. In the second stage, the coefficient of ESG performance is significantly positive at the 1% level, suggesting that despite addressing endogeneity concerns, positive ESG performance remains a driving factor of firms’ technological innovation. Furthermore, the key findings of the study remain robust.

There are a large number of studies confirming the positive impact of corporate network location on firm technological innovation (Ahuja, 2000; Woods et al., 2019). To test whether corporate network location is one of the mechanisms through which ESG promotes corporate technological innovation, this study takes corporate network location (Centrality, SH) as an explanatory variable and conducts regression analyses on ESG performance (Esg) and corporate network location, and the outcomes are presented in columns (1)–(2) of Table 5, where the regression coefficients for the relationship between ESG advantage (Esg) and Centrality, as well as SH, are observed to be significantly positive at a significance level of 1%. This suggests a substantial enhancement in corporate technological innovation resulting from the ESG advantage. Similarly, the coefficients for the firm’s network position are also significantly positive, implying that ESG advantage enhances the firm’s central position and control in the network, thus promoting technological innovation. Hence, hypothesis H2a is accepted.

A large number of studies have confirmed the positive impact of institutional shareholding on firms’ technological innovation (Manso, 2011; Aghion et al., 2013). To verify that institutional shareholding is one of the mechanisms through which ESG promotes firms’ technological innovation, this study takes institutional shareholding (INST) as an explanatory variable and conducts regression analyses on ESG performance (Esg) and institutional shareholding, and the results are shown in column (3) of Table 5. The coefficient of Esg is significantly positive, suggesting that ESG performance can effectively increase firms’ institutional shareholding and thus promote firms’ technological innovation. This is consistent with the findings of Bai et al. (2022). Hence, hypothesis H2b is accepted.

A large number of studies have been conducted to confirm the negative impact of financing constraints on firms’ R&D (Brown et al., 2009; Mancusi and Vezzulli, 2010). To examine the role of financing constraints as a mediating factor in the relationship between ESG performance and firms’ technological innovation, this study employs financing constraints as an independent variable and performs regression analyses on both ESG performance (Esg) and financing constraints (Fc). The findings, presented in column (4) of Table 5, provide evidence supporting the presence of such a mechanism. The estimated coefficient of Esg is significantly negative, indicating that ESG advantage obviously alleviates the problem of firms’ financing constraints and further promotes their technological innovation development. This is consistent with the findings of He et al. (2023). Hence, hypothesis H3a is accepted.

The research validates the positive influence of labor costs on firms’ technological innovation (Acemoglu, 2010; Wei et al., 2020). To verify that financing constraints is one of the mechanisms through which ESG promotes firms’ technological innovation, this study takes labor cost (Wage) as an explanatory variable and conducts a regression analysis between ESG performance (Esg) and labor cost, and the outcomes are presented in column (5) of Table 5, where a statistically significant positive regression coefficient for the relationship between labor cost and Esg on Wage is observed at a significance level of 5%. This implies that firms’ labor costs play a significant role as a mechanism through which ESG performance influences an important aspect of firms’ technological innovation. Hence, hypothesis H3b is accepted.