Does ESG Performance Improve the Quantity and Quality of Innovation? The Mediating Role of Internal Control Effectiveness and Analyst Coverage

Abstract

This study tests the performance of environmental, social, and governance (ESG) on corporate innovation and the mediating effect of internal control effectiveness and analyst coverage on this relationship, by using data on China’s A-share listed companies from 2009 to 2020. The results show that ESG performance significantly improves innovation quantity, measured by the number of authorized invention patents, and innovation quality, measured by the number of patent citations. The internal and external mechanism tests show that the quantity and quality improvement effect of ESG performance on corporate innovation is dependent on effective internal controls and adequate analyst coverage. The results of the heterogeneity analysis show that improvement in the quality of enterprise-driven innovation is primarily due to knowledge spillover into the domestic market. The additional analysis suggests that the promoting effect is more obvious when the chief executive officer (CEO) serves as a board chairman, the corporation belongs to non-state-owned and large-scale enterprises, the industry in which the market competition is higher, and the corporation is located in a general city. This study provides a foundation for developing a better ESG valuation theory to assist management and investors in making better decisions to improve business performance and investment returns.

1. Introduction

The environmental, social, and governance (ESG) rating is a comprehensive assessment report for enterprises based on sustainability. Unlike financial reports, which only provide decisions for investors, ESG has a wider range of stakeholders, including investors and enterprise owners [1,2]. It involves employees, suppliers, the government, and a wide range of other stakeholders to communicate the company’s performance in various aspects, such as the sustainable development concept, corporate social responsibility, and corporate performance. According to the International Sustainability Investment Coalition, ESG rating indicators are key to corporate investment decisions in most foreign countries. The United States, European Union, and Asian economies, including China, are developing and improving regulations in the ESG sector to define sustainable investment products more transparently and rigorously, guiding investors while promoting quality economic development. Unlike the mandatory requirements for ESG ratings in Europe and the United States, China’s ESG system is still in its early stages, with no mandatory pressure on companies to participate in ESG ratings [3]. This provides a better entry point to investigate the effectiveness of sustainable development in China’s unique context. However, policy incentives create many ‘patent bubbles’ and ‘innovation illusions’ in China, where intellectual property rights are not yet perfect [4]. Technological innovation is at the heart of sustainable economic growth in the context of high-quality economic development in China [5]. How will the integration of ESG concepts into corporate values and strategy development promote innovation?

Previous discussions on ESG performance and corporate innovation concentrated on short-term financial performance and firm value. While these studies confirm a positive relationship among ESG, financial performance [6,7,8,9], and firm value [10,11,12], they also examine the role of their impact mechanisms from the perspective of stakeholders [2,13,14]. In recent years, some scholars have started to apply these discussions to corporate innovation activities [3,15,16], discovering that ESG performance can positively impact corporate innovation activities, but only through internal effects, such as alleviating financing constraints and agency costs [17,18]. The focus of internal and external governance mechanisms is different; the two are interdependent and complementary. A one-sided emphasis is unlikely to achieve the desired governance effect [19]. Relying solely on internal governance ignores the critical role of external oversight in interpreting information and promoting regulation; however, relying on external oversight alone does not bring the subjective role of internal governance into play [20]. Therefore, according to the Theory of Second Best, this article discusses the mediating role of internal control effectiveness and analyst coverage in the relationship between ESG performance and corporate innovation.

Furthermore, current research on corporate innovation is more likely to categorize corporate innovation behavior based on innovation content and intensity than analyze it from a comprehensive perspective of innovation quantity and quality. In reality, enterprises engage in research and development (R&D) innovation to maintain a competitive advantage through technological innovation and engage in strategic innovation behaviors to maximize the direct profitability of innovation [4,21]. It is necessary to eliminate strategic innovation interference as much as possible from the research to investigate whether ESG performance can effectively influence firms’ real innovation output and, thus, achieve sustainable development. This study analyzes the impact of ESG performance on firms’ real innovation by excluding strategic innovation and discussing innovation quality in an attempt to fill this knowledge gap.

This study examines the effect of ESG performance on corporate technological innovation and its internal mechanisms by using the number of granted invention patents and the number of citations of these patents to represent innovation quantity and quality, respectively. First, it examines the effect of ESG performance on firms’ technological innovation. After excluding strategic innovation, it finds that ESG performance significantly improves firms’ innovation quantity and quality, with clear quantitative and qualitative improvements. This implies that ESG performance increases firms’ willingness to innovate to maintain a competitive advantage and achieve sustainable development through improved technological innovation. Second, this study investigates the mediating role of ESG performance in increasing the quantity and quality of corporate technological innovation through a combination of internal governance and external monitoring mechanisms. The internal control design of enterprises based on the ESG concept enables management to reasonably control innovation risks, mitigate hidden risks caused by agency problems, and improve innovation efficiency. Meanwhile, better ESG performance draws the attention of more analysts, and professional information interpretation and continuous tracking by analysts serve as an external monitoring function, alleviating information asymmetry between enterprises and investors and ensuring innovation investment. Lastly, this study discusses the impact of ESG performance on a firm’s technological innovation at different levels, from individual to industry. Domestic enterprises mainly determine the impact of ESG performance on the quality of corporate innovation. The promotion effect of innovation quantity and quality improvement is better in non-state-owned and large-scale enterprises. Inspired by ESG scores, CEOs with dual functions pay more attention to long-term development and R&D innovation. In terms of innovation quality, industries with a low level of competition have a relatively high level of innovation quality, and enterprises in general cities are more motivated by ESG performance.

The main contributions of this study are as follows: first, the study expands on the mediating role of ESG on innovation in terms of internal and external governance. Most existing ESG effects on innovation are investigated in terms of the internal mechanisms of financing constraints [17,18], ignoring the role of ESG as other factors of firm ratings, particularly external monitoring mechanisms. Therefore, this study analyzes a combination of two paths: the effect of internal control effectiveness on management decisions and the effect of analysts’ external monitoring on investors’ decisions. Second, the connotations of technological innovation indicators are assessed. Most existing technological innovation indicators are measured using patent applications, which are not granted a certain percentage of strategic innovation [21]. In contrast, this study uses authorized invention patents to measure enterprise technological innovation, which has been thoroughly reviewed by relevant institutions and can maximize the elimination of strategic innovation while directly reflecting true innovation that contributes to economic growth. Thus, this study distinguishes the quantity and quality of corporate innovation by distinguishing between the number of granted invention patents and the number of granted invention patent citations, and the identification is more precise. Third, it deepens the multilevel impacts of heterogeneity. Contrary to previous studies that only considered the impact of the nature of the enterprise [22], this study explores the individual’s overall progressive moderating effects of differences in domestic and foreign patent citations [23,24], management duality characteristics [25], degree of industry competition [26,27], and level of urban economic development [28].

2. Literature Review and Research Hypothesis

2.1. Literature Review

2.1.1. ESG’s Concept of Sustainability

The importance of sustainability in global businesses continues to increase [29]. The challenges and opportunities related to the social transition to sustainable development are increasingly embedded in the transformation and development of enterprises [30]. They are embodied in the management system and information disclosure of enterprises [31]. Since its introduction by the United Nations in 2004 [32], the public has gradually expanded the criteria for judging corporate sustainability from corporate social responsibility (CSR) to ESG [33]. Pressure from official regulations, investors, and stakeholders to disclose ESG performance has influenced companies’ attitudes toward sustainability. From a financial perspective, ESG has been widely discussed as a socially responsible investment (SRI) [34]. However, the discussion of its impact on internal operations from the sustainable development (SD) perspective has only emerged in recent years [35], with a limited focus on understanding customer needs and integrating them with corporate innovation to generate value [36]. Traditional innovations can no longer meet the requirements of social development. Technological innovation and business model innovation based on sustainable development provides a new perspective for enterprise innovation [37,38,39,40]. In innovation theory, business model innovation is an important tool for companies to realize the value of technological innovation [41]. Furthermore, technological innovation is the key to ESG performance in driving sustainable development [15,18].

2.1.2. Corporate Innovation Behavior

Corporate innovation activities are a strategy to retain a certain level of competitiveness through technological innovation and may also be an action to meet the needs of the government and relevant regulatory bodies [4,21]. Corporate innovation, as a strategic behavior, is characterized by high risk and positive externalities. Specifically, the high riskiness of innovation is manifested in the uncertainty in terms of technology and returns [42], which directly limits the intensity and the motivation of firms to innovate. The positive externality of innovation manifests as the existence of a certain degree of spillover effects [43]. Knowledge spillover is an important feature of innovation. It refers to the ability of enterprises to learn, absorb, and imitate advanced knowledge and technology to improve enterprise production efficiency [44,45]. Knowledge spillovers depend on the size of the absorbing capacity of the absorbing party, the absorbing capacity of the spillover party, and the level of knowledge stock [46,47]. Furthermore, in addition to R&D investment, the knowledge gap between enterprises influences their knowledge-absorption capacity. The depth and breadth of externally acquired knowledge impact innovation at the source of the value creation chain [48].

2.2. Research Hypothesis

2.2.1. Impact of ESG on Enterprise Innovation

ESG rating is a comprehensive evaluation of corporate sustainability performance in three dimensions: environmental, social, and governance, which is integrated into the management system and information disclosure and gradually embedded into management decision-making and investor decision-making behavior [49]. From a management perspective, ESG ratings reflect a company’s value [50]. A higher rating indicates that the overall strategy and operations of the company are less risky, resulting in better financial performance [6]. From an investor perspective, investors focus on and obtain information about a company’s long-term value [51]. ESG performance promotes the development of related areas, such as eco-environmental protection and low-carbon transitions, while stimulating innovation activities at the heart of companies’ efforts to achieve sustainable development [11,18,52]. Therefore, this study investigates the impact of ESG performance on corporate innovation and develops an analysis according to three dimensions.

First, environmental information disclosure is essential for promoting enterprise innovation, reducing pollution and carbon emissions, and achieving the dual carbon strategic goal [53]. Conversely, environmental information disclosure can play a regulatory role for companies, enhance management’s capability for innovation activities, reduce agency costs, and thus, reduce investor risk [54]. In contrast, environmental information disclosure can improve the credibility of enterprises, reduce information asymmetry between stakeholders, such as the government, social investors, and enterprises, and provide important support for enterprises’ innovation activities [55]. Second, social responsibility affects a company’s strategy and business model; when companies make CSR activities part of their strategy, they must innovate to stay ahead of their competitors [56]. Conversely, being socially responsible helps companies share resources with external stakeholders, thus facilitating access to innovation [52]. In contrast, socially responsible companies are more likely to attract and retain better employees, thus motivating them and promoting innovative technologies [57]. Lastly, corporate governance is considered an effective way to address agency problems between managers and shareholders and may influence corporate innovation in several ways [58]. Conversely, corporate governance can align managers’ short-term incentives and career concerns with shareholders’ long-term interests, promoting innovation investment and increased productivity [59]. In contrast, strong or repressive governance practices may lead to overly strict supervision by managers and result in short-sighted management behavior [60]. Management tends to choose short-term, conservative, and safe projects with low investments to satisfy investors, resulting in insufficient investment in innovative R&D projects [61].

In addition, ESG distinguishes itself from compulsory institutional arrangements that produce a push-back mechanism for corporate innovation, leading to the problem that companies present in terms of quantity and quality of innovation, emphasizing quantity over quality [62]. Companies’ participation in ESG ratings is voluntary and can significantly alleviate the problem of slipping in the quality of innovation. That is, a company’s ESG performance promotes an increase in the quantity of corporate innovation while improving the quality of corporate innovation, showing the effect of increasing quantity while improving quality. On the basis of the above analysis, Hypothesis 1 is proposed.

Hypothesis 1 (H1).

ESG performance can improve the quantity and quality of innovation.

2.2.2. The Mediating Role of Internal Control Effectiveness

Basic Norms for Enterprise Internal Control in China indicate that internal control is a management process that ensures the safety of corporate assets, improves operational efficiency and effectiveness, and promotes the realization of corporate development strategies [63]. It plays an effective role in internal supervision and risk-management mechanisms. The ESG concept emphasizes the long-term development ability of an enterprise. This is consistent with the sustainable development goal of the enterprise’s internal control. The design of enterprise internal control based on the ESG concept is conducive to reducing the financing and operating risks of the enterprise and promoting the improvement of internal control ability [64]. In recent times, the output and quality of enterprises have no longer been the focus of enterprise competition and development but are more reflected in the internal management and strategic risk of enterprises. Therefore, enterprises must adopt the internal control mode of integrating sustainable development and ESG to improve the effectiveness and long-term development of their business activities.

The development of innovation in enterprises inevitably faces risks. Effective internal control can provide timely feedback to manage the risks arising from the innovation process and develop management countermeasures to cope with and prevent these risks [65]. First, effective internal control ensures that the innovation process is conducted in a normal and orderly manner, thus providing a conducive environment [66]. Second, effective internal control enables management to effectively identify risks and opportunities in the external environment and internal resources, reduce management’s short-sightedness toward innovation, and prevent long-term investment in innovation projects to target short-term performance [67]. Third, effective internal control reduces management’s incentive to capture private interests through innovation investments and mitigates agency conflicts between management and investors by improving the relevant monitoring and governance mechanisms [68]. The reduction in opportunistic behavior among executives improves capital use and management efficiency [69].

Since ESG performance has an indirect influence on corporate innovation without direct corporate governance [70], the intermediary factor of internal control is needed to influence the output activities of innovation: the design of corporate internal control based on the ESG concept, prompting management to reasonably control innovation risks, mitigating the hidden dangers caused by agency problems, and improving innovation efficiency, thus promoting an increase in the quantity of corporate innovation and improving innovation quality. On the basis of the above analysis, Hypothesis 2 is proposed.

Hypothesis 2 (H2).

Internal control effectiveness mediates the relationship between ESG performance and corporate innovation.

2.2.3. The Mediating Role Analyst Coverage

As third-party intermediaries in the external market, analysts play a further role in the transmission and dissection of information by obtaining financial and nonfinancial information about enterprises and making independent analyses and predictions about enterprises based on their long-term development. Analysts have unique advantages in monitoring listed companies compared to traditional corporate governance, and they are also recognized by investors for their objectivity, professionalism, and prudence [71]. In April 2002, the China Securities Regulatory Commission issued a Guidance on Investor Relationship Management for Listed Companies, which included ESG information consideration requirements for the first time in the communication content of investor relationship management. In recent years, ESG has become a popular investment concept and corporate evaluation standard in the capital market. Its content contains financial information about the company and integrates nonfinancial information, such as environmental, social, and corporate governance [72], which provides investors and financial analysts with incrementally useful information and thus attracts more attention from analysts.

Technological innovation has a certain professionalism, and shareholders’ lack of professional knowledge easily leads to information asymmetry between shareholders and management, increasing the cost of shareholder supervision [68]. Innovation technology has a certain degree of confidentiality, and, because of limited information and energy, external information users can easily lead to a certain degree of asymmetry between the internal and external information of the enterprise. This makes investors interpret the information with errors, thus affecting the value assessment of innovation and investment decisions [73]. Information asymmetry between enterprises and investors can be mitigated by the level of analyst coverage, which provides effective and timely information to inform users through professional analysis capabilities, ensures effective interpretation of information on the value of innovation projects, and promotes investment in innovation [74].

Better ESG performance attracts more attention from analysts, and analysts’ professional information interpretation and continuous tracking perform an external monitoring function that compensates for the lack of internal oversight. Therefore, the mediating factor of analyst coverage is required to influence innovation output activities. Professional information interpretation and continuous follow-up by analysts perform an external monitoring function and alleviate information asymmetry between companies and investors, thus better utilizing the market’s role in allocating credit funds and ensuring innovative inputs [75,76]. On the basis of the above analysis, this study proposes Hypothesis 3.

Hypothesis 3 (H3).

Analyst coverage mediates the relationship between ESG performance and corporate innovation.

The theoretical model of this research is shown in Figure 1. These tests include the H1 baseline test, the H2 and H3 mediating effect tests, and the heterogeneity tests.

3. Research Design

3.1. Sample Selection and Data Source

The article uses the data of manufacturing listed companies in China A-shares from 2009 to 2020 for the study. In order to avoid the interference of existing data by other factors, the sample is treated as follows: (1) companies that suffered losses for two consecutive years, net assets below stock par value (ST), and losses for three consecutive years subject to delisting risk warning (*ST) are excluded; (2) financial companies are excluded; (3) companies with incomplete data are excluded. A final sample of 18,058 observations was obtained. Since the main variables in this study are continuous variables, the relevant variables are winsorized at the 1% and 99% levels in order to exclude the influence of outliers. The ESG performance data in this study use the rating results provided by Sino-Securities Index Information Service, which is obtained from the Wind database, the innovation variables from the CSMAR database, the quantity of internal control quality from the DIB database, and the remaining control variables from the CSMAR database, supplemented by the Wind database for missing values.

3.2. Variable Definition

3.2.1. Corporate Innovation

In this study, we use granted invention patents to measure the quantity of innovation and citation of granted invention patents to measure the quality of innovation [77,78,79].

The reasons for this are mainly as follows: first, in terms of the quantity of innovation, the number of patents is an important indicator of innovation performance, which can objectively reflect the progress of innovation [80]. However, existing studies have found that patent applications contain a large number of strategic innovations; in order to meet the demands of stakeholders, enterprises meet the annual patent application quota by splitting a single patent application into multiple applications [81]. Granted patents, on the other hand, undergo substantive examination by the relevant authorities, eliminating as much strategic innovation as possible. Therefore, we focus on granted patents. Secondly, in terms of innovation quality, invention patents are highly heterogeneous, and their quantity is not a better measure of the quality of innovation [82,83]. In contrast, patent citations are direct feedback of the market value of patents, which can adequately represent the commercial value and technological impact of innovations [84,85], thus better capturing the dynamic changes of innovation quality [86]. Therefore, the article selects the number of citations of granted patents to measure quality.

3.2.2. ESG Performance

The ESG performance data in this study use the rating results provided by Sino-Securities Index Information Service. The ESG rating system of Sino-Securities Index Information Service is divided into nine levels, and a nine-point scale is used to assign a score to the ESG performance of enterprises, with higher scores representing better ESG performance. At the same time, the average value of each quarterly score is taken to measure the annual ESG performance. Unlike the ESG ratings of other institutions, the main reasons are as follows: (1) the data of the ESG evaluation system of Sino-Securities Index Information Service started early and the data coverage is more complete; (2) the index has the characteristics of high growth, high profitability, and low leverage; (3) the index focuses on ESG factors and deploys high growth industries.

3.2.3. Mediation Variables

The internal control effectiveness (ICE) of listed companies is measured using the Diebold China listed companies internal control index [87], and a higher value of this index indicates stronger internal control effectiveness of the company. Analyst coverage (Coverage) is measured by using the number of analysts who follow the company in a natural year plus one to take the natural logarithm [88,89].

3.2.4. Control Variables

The study uses enterprise size (Size), asset liability (Lev), profitability (Roa), growth rate of operating income (Growth), number of board members (Board), proportion of independent directors (Indep), CEO duality (Dual), state-owned enterprises (Soe), enterprise age (ListAge), shareholding ratio of the largest shareholder (Top1), and institutional shareholding ratio (INST) [18,90,91]. In addition, the study controls the year fixed effects and industry fixed effects. The industry classification is subdivided by the three-digit code of the latest version of the 2012 Guidelines for the Classification of Chinese Listed Companies. See

Table 1. Variable and definition.

Variable	Symbol	Description
Innovation quantity	lnPatent	Natural logarithm of the sum of the number of invention authorized patents and 1
Innovation quality	lnCitation	Natural logarithm of the sum of the number of citations of authorized invention patents and 1
	Dom_lnCit	Natural logarithm of the sum of the number of domestic citations of authorized invention patents and 1
	Fore_lnCit	Natural logarithm of the sum of the number of foreign citations of authorized invention patents and 1
ESG performance	ESG	According to the ESG rating of China Securities, the value is 1–9 from low to high
Internal control effectiveness	ICE	Dibo internal control index
Analyst coverage	Coverage	Logarithm of the total number of securities analysts issuing analysis and forecast reports
Corporate size	Size	Natural logarithm of total assets
Asset–liability ratio	Lev	Ratio of total liabilities to total assets
Return on assets	Roa	rate of return on total assets
Corporate growth	Growth	Growth rate of operating income
Board size	Board	Natural logarithm of the number of directors
Board independence	Indep	Independent directors divided by the number of directors
CEO duality	Dual	CEO concurrently serves as the chairman and general manager, it is 1; otherwise, it is 0
Equity nature	Soe	In case of state-owned holding, the value of Soe is 1; otherwise, the value is 0
Corporate years	ListAge	Establishment years of listed companies
Equity concentration	Top1	Shareholding ratio of the largest shareholder
Institutional share proportion	INST	Number of shares held by institutional investors divided by total number of shares

for specific variable definitions.

3.3. Model Setting

In this study, regression model (1) was used to test the impact of ESG performance on corporate innovation. The specific regression model is as follows:

$$Innovatio{n}_{i,t}={\beta }_0+{\beta }_{1}ES{G}_{i,t}+{\beta }_{2}Contro{l}_{i,t}+{\beta }_{3}In{d}_{i,t}+{\beta }_{4}Yea{r}_{i,t}+{ϵ}_{i,t},$$

(1)

where$Innovatio{n}_{i,t}$ denotes the dependent variable, including the quantity of innovation (lnPatent) and the quality of innovation (lnCitation). Since patents are count data, the study alleviates the estimation bias that may be caused by the right-skewed distribution of patent data by taking patents in a logarithmic form to make them more closely to a normal distribution. $ES{G}_{i,t}$ denotes the independent variable, which represents the ESG rating received by firm i in year t. $Contro{l}_{i,t}$ denotes the control variables in this study. In addition, to control for some unobservable factors that do not vary over time, industry fixed effects and time fixed effects are further controlled for in this study. To control for estimation bias due to possible heteroskedasticity and time-series correlation of the error terms, the study uses clustering robust standard errors at the firm level.

In order to further verify the pathways through which ESG performance affects corporate innovation, the study develops model (2) and model (3) to verify the mediating role of internal control effectiveness (ICE) and analyst coverage (Coverage), respectively, according to theoretical analysis.

$$Media{n}_{i,t}={\beta }_0+{\beta }_{1}ES{G}_{i,t}+{\beta }_{2}Contro{l}_{i,t}+{\beta }_{3}In{d}_{i,t}+{\beta }_{4}Yea{r}_{i,t}+{ϵ}_{i,t}.$$

(2)

$$Innovatio{n}_{i,t}={\gamma }_0+{\gamma }_{1}ES{G}_{i,t}+{\gamma }_{2}Media{n}_{i,t}+{\gamma }_{3}Contro{l}_{i,t}+{\gamma }_{4}In{d}_{i,t}+{\gamma }_{5}Yea{r}_{i,t}+{ϵ}_{i,t}.$$

(3)

4. Empirical Analysis

4.1. Descriptive Statistics

Table 2. Results of descriptive statistics of the main variables.

Variable	N	Mean	SD	Min	Max
lnPatent	18,057	1.205	1.338	0	8.430
lnCitation	18,057	1.501	1.775	0	10.11
Dom_lnCit	18,057	0.433	0.946	0	8.969
Fore_lnCit	18,057	1.458	1.740	0	9.729
ESG	18,057	6.344	1.022	2	9
Size	18,057	21.99	1.147	19.87	25.42
Lev	18,057	0.403	0.194	0.0530	0.865
Roa	18,057	0.0430	0.0620	−0.203	0.220
Growth	18,057	0.160	0.362	−0.476	2.258
Board	18,057	2.125	0.190	1.609	2.639
Indep	18,057	0.374	0.0530	0.333	0.571
Dual	18,057	0.294	0.455	0	1
Soe	18,057	0.300	0.458	0	1
ListAge	18,057	2.078	0.748	0.693	3.258
Top1	18,057	0.338	0.140	0.0900	0.716
INST	18,057	0.372	0.230	0	0.860

reports that the mean value of ESG performance is 6.344 with a variance of 1.022. This indicates that the average rating of manufacturing firms is relatively high, but there is a large intra-firm variation. Meanwhile, the mean values of corporate innovation quantity (lnPatent) and corporate innovation quality (lnCitation) are 1.205 and 1.501, respectively, with variances of 1.338 and 1.775, respectively. This indicates that the inter-firm differences in technological innovation are much greater than the differences between ESG performance, and the inter-firm differences in innovation quality performance are the greatest. In addition, after further observing innovation quality by distinguishing total patent citations into foreign citations and domestic citations, it shows the difference in innovation quality between firms mainly comes from the difference in foreign citations of patents when considering the knowledge spillover of firms’ innovation. The results of descriptive statistics for the remaining control variables are generally consistent with existing studies.

4.2. Baseline Regression Results

First, the study examines the impact of ESG performance on corporate innovation. Columns (1) and (3) of

Table 3. ESG performance and corporate innovation regression results.

	(1)	(2)	(3)	(4)
	lnPatent	lnPatent	lnCitation	lnCitation
ESG	0.2841 ***	0.0858 ***	0.3121 ***	0.0863 ***
	(0.020)	(0.014)	(0.024)	(0.018)
Growth		−0.0369		−0.0137
		(0.026)		(0.035)
Roa		1.3204 ***		1.7392 ***
		(0.219)		(0.287)
Lev		−0.2008 **		−0.3155 ***
		(0.093)		(0.121)
Top1		−0.2930 **		−0.4326 **
		(0.132)		(0.168)
Indep		0.6012 *		0.4381
		(0.359)		(0.459)
Board		0.2537 **		0.2711 *
		(0.111)		(0.140)
INST		0.2159 ***		0.2628 ***
		(0.067)		(0.087)
ListAge		−0.0886 ***		−0.1400 ***
		(0.025)		(0.032)
Size		0.5195 ***		0.6031 ***
		(0.024)		(0.029)
Soe		0.1278 **		0.1401 **
		(0.050)		(0.064)
Dual		0.0715 **		0.1036 ***
		(0.031)		(0.040)
_cons	−0.5977 ***	−11.3514 ***	−0.4789 ***	−12.7238 ***
	(0.122)	(0.584)	(0.145)	(0.699)
N	18,057	18,057	18,057	18,057
Ind	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
r2_a	0.2789	0.4393	0.3110	0.4300

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

report the regression results obtained without adding control variables and controlling only for industry fixed effects and year fixed effects. The regression results showed that the coefficient values of the core explanatory variables were 0.2841 and 0.3121, respectively, which were significantly positive at the level of 1%. Columns (2) and (4) report the coefficient values of the core explanatory variables are 0.0858 and 0.0863, respectively, while the significance level and sign of the regression coefficient values remain unchanged. This indicates that the quantification of ESG performance can significantly improve the substantive innovation of manufacturing enterprises, as well as increase the quantity of invention authorized patents and citations of invention authorized patents, with obvious characteristics of quantity and quality improvement. Specifically, ESG performance increases in terms of the number of patents by 8.58% and citations of patents by 8.63%. The results illustrate that companies conducting high-quality innovation activities are influenced by both their own development and the external environment, thus verifying Hypothesis 1 (ESG performance can promote the rise in the quantity and quality of corporate innovation).

4.3. Mediating Effect Tests

4.3.1. Mediating Effect Test of Internal Control Effectiveness

Effective implementation of internal control can ensure the quality of enterprise financial reporting disclosures and reduce the likelihood of corporate irregularities, which in turn improves corporate performance and meets the funding needs of innovative activities. Meanwhile, ESG performance is closely related to the quality of internal control. Companies with good ESG performance often have better internal control systems. In this study, we use the DIB internal control index to measure the effectiveness of internal control of listed companies [87]. The regression results are shown in

Table 4. Results of the mediating effect of internal control effectiveness and analyst coverage.

	(1)	(2)	(3)	(4)	(5)	(6)
	ICE	lnPatent	lnCitation	Coverage	lnPatent	lnCitation
ESG	0.0796 ***	0.0848 ***	0.0845 ***	0.0543 ***	0.0901 ***	0.0898 ***
	(0.011)	(0.014)	(0.018)	(0.010)	(0.017)	(0.022)
ICE		0.0135 *	0.0227 **
		(0.008)	(0.011)
Coverage					0.1030 ***	0.1421 ***
					(0.020)	(0.026)
Growth	0.0429 *	−0.0375	−0.0147	0.0024	−0.0632 **	−0.0362
	(0.026)	(0.026)	(0.035)	(0.021)	(0.032)	(0.042)
Roa	3.8660 ***	1.2681 ***	1.6515 ***	5.8575 ***	1.1113 ***	1.5055 ***
	(0.272)	(0.224)	(0.293)	(0.195)	(0.306)	(0.403)
Lev	−0.2216 ***	−0.1978 **	−0.3105 **	−0.2425 ***	−0.1249	−0.1925
	(0.076)	(0.093)	(0.121)	(0.069)	(0.118)	(0.155)
Top1	0.0549	−0.2937 **	−0.4338 ***	−0.6451 ***	−0.2632 *	−0.3704 *
	(0.061)	(0.132)	(0.168)	(0.085)	(0.158)	(0.201)
Indep	0.3006 *	0.5972 *	0.4313	0.2768	0.5324	0.3006
	(0.178)	(0.359)	(0.459)	(0.231)	(0.424)	(0.542)
Board	0.0135	0.2535 **	0.2708 *	0.0471	0.2729 **	0.3183 *
	(0.059)	(0.111)	(0.140)	(0.067)	(0.133)	(0.167)
INST	−0.0564	0.2167 ***	0.2641 ***	0.6920 ***	0.1248	0.1049
	(0.043)	(0.067)	(0.087)	(0.049)	(0.078)	(0.103)
ListAge	−0.0911 ***	−0.0874 ***	−0.1380 ***	−0.1786 ***	−0.0846 ***	−0.1246 ***
	(0.016)	(0.025)	(0.032)	(0.020)	(0.032)	(0.040)
Size	0.0520 ***	0.5188 ***	0.6019 ***	0.3754 ***	0.5313 ***	0.6118 ***
	(0.012)	(0.024)	(0.029)	(0.014)	(0.032)	(0.038)
Soe	0.0517 **	0.1271 **	0.1389 **	−0.1691 ***	0.1576 **	0.1671 **
	(0.025)	(0.050)	(0.064)	(0.033)	(0.061)	(0.078)
Dual	0.0297 *	0.0711 **	0.1029 **	0.0932 ***	0.0854 **	0.1184 **
	(0.017)	(0.031)	(0.040)	(0.022)	(0.038)	(0.047)
_cons	4.6199 ***	−11.4140 ***	−12.8287 ***	−6.7467 ***	−11.7950 ***	−13.1847 ***
	(0.271)	(0.584)	(0.700)	(0.303)	(0.739)	(0.883)
N	18,057	18,057	18,057	13,110	13,110	13,110
Ind	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.0999	0.4394	0.4301	0.3977	0.4425	0.4312

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

. Column (1) indicates that the ESG coefficient is 0.0796, which is significantly positive at the 1% level, indicating that the existence of a significant relationship between ESG and internal control quality is verified. Columns (2) and (3) show that the coefficient values of ESG performance and internal control effectiveness are both positive and significantly positive at the 1% level after the inclusion of mediating variables. In the further test, the Sobel value of innovation quantity is 0.01079, which is significant at the level of 1%. The Sobel value of innovation quality is 0.01335, which is significant at the level of 1%. Among them, the intermediary effect of innovation quantity accounts for 23.28% of the total effect. This shows that the quality of internal control effectiveness plays a mediating role in the influence of ESG performance on corporate innovation. In other words, ESG performance can promote the quantity and quality of innovation through the effectiveness of internal control, thus verifying Hypothesis 2.

4.3.2. Mediating Effect Test of Analyst Coverage

The analyst coverage behavior plays a positive role in reducing information asymmetry [24,92]. On the one hand, analysts provide effective and timely information for information users through their professional analysis ability, so as to ensure the effective interpretation of the value information of innovation projects and promote the investment in innovation. On the other hand, the continuous tracking of innovation activities by analysts can play a role of external supervision and make up for the deficiency of internal supervision. The research defines analyst coverage as the number of analysts who follow the firm in a natural year plus one, taking the natural logarithm [88]. The regression results are shown in Table 4. Column (4) indicates that, in the regression of ESG performance on analyst coverage, the ESG coefficient is 0.0543, which is significantly positive at the 1% level, which indicates a significant positive relationship between ESG and analyst coverage. Columns (5) and (6) show that the coefficient values of ESG performance and analyst coverage are both positive and significantly positive at the 1% level after the inclusion of intermediate variables. In a further test, the Sobel value for the explanatory variable being the number of innovations is 0.01312, which is significant at the 1% level. The Sobel value for the explanatory variable being the quality of innovation is 0.015758, which is significant at the 1% level. Among them, the mediating effect of quantity of innovation accounts for 6.559% of the total effect. The mediating effect of innovation quality accounts for 4.93% of the total effect. This indicates that analyst coverage has a mediating role in the effect of ESG performance on corporate innovation. That is, better ESG performance of firms can promote the quantity and quality of corporate innovation through analyst coverage, thus verifying Hypothesis 3.

4.4. Heterogeneity Test

We revealed the impact of ESG performance on corporate innovation and explored how ESG ratings promote the quantity and quality of corporate innovation from both internal and external governance perspectives. In this section, the study analyzes the heterogeneity of these results at the domestic and foreign patent citations, firm characteristics, industrial competition degree, and city level, as well as explores what factors affect corporate innovation activities in order to make the findings of the study more relevant.

4.4.1. Test of Domestic and Foreign Patent Citations

From the perspective of knowledge spillover, knowledge spillover is characterized by obvious geographical distance due to the convenience of inter-enterprise, as well as R&D personnel communication. Therefore, in this study, the number of patent citations is divided into domestic citations (Dom_lnCit) and foreign citations (Fore_lnCit) for a comparison based on the geographical attributes of patent citations. Columns (1) and (3) of

Table 5. Test of domestic and foreign patent citations.

	(1)	(2)	(3)	(4)
	Dom_lnCit	Dom_lnCit	Fore_lnCit	Fore_lnCit
ESG	0.1522 ***	0.0420 ***	0.3069 ***	0.0862 ***
	(0.017)	(0.011)	(0.024)	(0.017)
Growth		−0.0150		−0.0108
		(0.022)		(0.034)
Roa		0.9257 ***		1.6467 ***
		(0.195)		(0.280)
Lev		−0.1801 **		−0.2945 **
		(0.072)		(0.118)
Top1		−0.2767 **		−0.3951 **
		(0.114)		(0.165)
Indep		0.3412		0.4642
		(0.294)		(0.451)
Board		0.0485		0.2837 **
		(0.098)		(0.138)
INST		0.1664 ***		0.2473 ***
		(0.054)		(0.084)
ListAge		−0.0480 **		−0.1384 ***
		(0.021)		(0.032)
Size		0.3106 ***		0.5866 ***
		(0.025)		(0.028)
Soe		−0.0010		0.1506 **
		(0.044)		(0.062)
Dual		0.0635 **		0.1022 ***
		(0.025)		(0.039)
_cons	−0.5322 ***	−6.7448 ***	−0.4887 ***	−12.4575 ***
	(0.097)	(0.624)	(0.142)	(0.683)
N	18,057	18,057	18,057	18,057
Ind	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
r2_a	0.1660	0.2735	0.3112	0.4291

Note: ** and *** denote significance at the levels of 5% and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

are the regression results obtained after controlling for industry fixed effects and year fixed effects, and the coefficient values of the core explanatory variables are significantly positive at the 1% level. Columns (2) and (4) show the regression results after adding a series of control variables, showing that the coefficient values and significance levels of the core explanatory variables remain the same. Furthermore, the comparison of the coefficient values shows that ESG performance has a greater contribution to domestic patent citations, about twice as much as foreign patent citations. The increase in the number of patent citations represents wider market recognition and a higher quality of innovation. Thus, innovation drive released by ESG performance allows companies to improve the quality of their innovation output and it gains higher recognition in the domestic market.

4.4.2. Test of Enterprise Characteristics

Enterprise Ownership Test

Corporate ownership differences may yield different results in terms of the effect of ESG performance on firms’ innovative activities [23]. Specifically, state-owned enterprises, which have stable sources of capital but are subject to stronger regulation, have an advantage in the capital elements needed for innovation. However, non-SOEs are more motivated to innovate in order to meet the needs of stakeholders and for the survival and growth of the firm. For this reason, the study divides the sample into two groups: state-owned enterprises and non-state-owned enterprises, based on the ownership attributes of the enterprises. The results are shown in columns (1)–(4) of

Table 6. Tests of the ownership and size of firms.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	lnPatent		lnCitation		lnPatent		lnCitation
	Soes	Non_Soes	Soes	Non_Soes	Large	Non_Large	Large	Non_Large
ESG	0.0646 ***	0.0903 ***	0.0579 **	0.0910 ***	0.1005 ***	0.0317 **	0.1046 ***	0.0228
	(0.022)	(0.016)	(0.029)	(0.021)	(0.020)	(0.014)	(0.025)	(0.019)
Growth	−0.0966 **	−0.0179	−0.0403	−0.0092	−0.0963 **	0.0360	−0.0663	0.0462
	(0.046)	(0.031)	(0.060)	(0.041)	(0.038)	(0.029)	(0.049)	(0.042)
Roa	1.2613 ***	1.3540 ***	1.6318 ***	1.8052 ***	2.3964 ***	0.5983 ***	2.9497 ***	0.8187 ***
	(0.338)	(0.264)	(0.424)	(0.354)	(0.365)	(0.192)	(0.466)	(0.274)
Lev	−0.1835	−0.2054 *	−0.1954	−0.3625 **	−0.1520	−0.2976 ***	−0.3225	−0.4212 ***
	(0.145)	(0.110)	(0.187)	(0.144)	(0.158)	(0.088)	(0.200)	(0.123)
Top1	−0.1827	−0.3469 **	−0.3003	−0.4991 **	−0.2852	−0.3638 ***	−0.4177	−0.5491 ***
	(0.210)	(0.153)	(0.265)	(0.197)	(0.205)	(0.116)	(0.257)	(0.161)
Indep	1.0975 *	0.3877	1.2239 *	0.1266	0.7050	−0.0050	0.5229	−0.1970
	(0.580)	(0.417)	(0.727)	(0.535)	(0.496)	(0.329)	(0.633)	(0.459)
Board	0.2891	0.2261 *	0.3896 *	0.2153	0.3283 **	0.1099	0.3473 *	0.1009
	(0.185)	(0.129)	(0.231)	(0.162)	(0.166)	(0.096)	(0.202)	(0.137)
INST	0.1011	0.2624 ***	0.1159	0.3240 ***	0.1897 *	0.1756 ***	0.2130	0.2316 **
	(0.103)	(0.082)	(0.132)	(0.106)	(0.106)	(0.066)	(0.133)	(0.093)
ListAge	−0.0875 **	−0.0890 ***	−0.1248 **	−0.1459 ***	−0.1092 ***	−0.0766 ***	−0.1464 ***	−0.1370 ***
	(0.039)	(0.031)	(0.051)	(0.039)	(0.042)	(0.023)	(0.052)	(0.032)
Size	0.5205 ***	0.5197 ***	0.5872 ***	0.6097 ***	0.6522 ***	0.3947 ***	0.7474 ***	0.4887 ***
	(0.044)	(0.027)	(0.052)	(0.033)	(0.046)	(0.026)	(0.055)	(0.036)
Dual	0.1064	0.1324 **	0.1360	0.1411 **	0.0738	0.0338	0.1084 *	0.0574
	(0.124)	(0.053)	(0.150)	(0.068)	(0.053)	(0.029)	(0.065)	(0.040)
_cons	11.4324 ***	11.2475 ***	12.7551 ***	12.6186 ***	14.7202 ***	−7.7308 ***	16.3771 ***	−9.1585 ***
	(1.028)	(0.650)	(1.216)	(0.786)	(1.050)	(0.575)	(1.233)	(0.792)
N	5301	12,756	5301	12,756	8940	9117	8940	9117
Ind	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.4341	0.4440	0.4418	0.4284	0.4836	0.2793	0.4911	0.2972

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

. First, the effect of better ESG performance on enterprises’ innovation and quality exists regardless of whether they are state-owned or not. Secondly, the effect of better ESG performance on increasing the quality and quantity of innovation is much better in non-SOEs than in SOEs. This is explained by the fact that SOEs have easier access to funding for R&D activities due to the backing of the state and do not rely as much on their own resources related to ESG scores as non-SOEs. Therefore, the ESG performance of non-SOEs has a greater incentive effect on firms’ innovation activities.

Enterprise Size Test

Innovation activities have the characteristics of long cycle and high risk [22]. Therefore, differences in firm size may produce heterogeneous results in the contribution of ESG performance to corporate innovation. In this study, according to the size attributes of firms, samples with firm sizes above the mean are defined as large-scale firms, and those with firm sizes below the mean are defined as non-large-scale firms. The results are shown in columns (5)–(8) of Table 6. First, regardless of firm size, the promotion effect of better ESG performance on the quantity of corporate innovation is significant. However, in terms of innovation quality, only large-scale firms have improved. Second, from the coefficient values of the regression results, the coefficient values of large-scale firms are three times higher than those of non-large-scale firms in terms of the quantity of innovation. This suggests that the impact of better ESG performance on non-large-scale firms is small and limited. The explanation for this is that large-scale firms generally have relatively high social credibility due to their larger size, and their access to resources in terms of innovation through ESG performance allows them to effectively increase their ability to take innovation risks, which in turn increases their technological innovation.

Management Duality Test

In this study, the sample is distinguished into two groups of corporate managers with dual positions and dual non-combined positions, based on the attribute of whether the general manager and the chairman of the company are dual or not [93].

Table 7. Test of management duality.

	(1)	(2)	(3)	(4)
	lnPatent		lnCitation
	Dual	Non_Dual	Dual	Non_Dual
ESG	0.0962 ***	0.0724 ***	0.0909 **	0.0739 ***
	(0.028)	(0.016)	(0.035)	(0.020)
Growth	0.0716	−0.0735 **	0.1072 *	−0.0520
	(0.048)	(0.031)	(0.065)	(0.041)
Roa	1.0866 **	1.3817 ***	1.7188 **	1.7114 ***
	(0.511)	(0.229)	(0.684)	(0.304)
Lev	−0.7111 ***	0.0371	−0.9518 ***	−0.0160
	(0.198)	(0.098)	(0.258)	(0.128)
Top1	−0.2849	−0.3076 **	−0.3727	−0.4709 ***
	(0.284)	(0.140)	(0.355)	(0.182)
Indep	0.5015	0.6517	0.1848	0.5889
	(0.636)	(0.396)	(0.823)	(0.510)
Board	0.1667	0.2652 **	0.1847	0.2979 **
	(0.217)	(0.119)	(0.270)	(0.152)
INST	0.2777 *	0.1981 ***	0.4539 **	0.2156 **
	(0.161)	(0.072)	(0.205)	(0.094)
ListAge	−0.1266 **	−0.0796 ***	−0.1643 **	−0.1326 ***
	(0.061)	(0.027)	(0.078)	(0.035)
Size	0.5971 ***	0.4665 ***	0.6935 ***	0.5393 ***
	(0.041)	(0.029)	(0.050)	(0.034)
Dual	0.0825	0.0782 **	0.1612	0.1018 **
	(0.101)	(0.032)	(0.120)	(0.041)
_cons	−12.4774 ***	−10.2618 ***	−14.0085 ***	−11.5102 ***
	(0.929)	(0.745)	(1.147)	(0.878)
N	5412	12,645	5412	12,645
Ind	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
r2_a	0.4989	0.3957	0.4715	0.3995

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

indicates that the effect of ESG scores on the incremental quality improvement of corporate innovation exists regardless of whether the corporate managers’ dual positions are concurrent or not. However, the effect is better in the group where corporate managers have dual jobs than in the group where dual jobs are not combined. This is explained by the fact that when a corporate manager has both the chairman and general manager positions, they will be more focused on the long-term interests of the company and will be more determined to conduct R&D innovation with the incentive of ESG scores. At the same time, the dual role also avoids the principal–agent problem, and the managers are more motivated to invest in R&D activities for the long-term development of the company. Therefore, duality of COB and CEO is conducive to ESG performance having an innovative incentive effect.

4.4.3. Test of the Degree of Competition in the Industry

The study further examines the impact of ESG performance on corporate innovation under different competitive market environments. A higher HHI index indicates a lower degree of market competition, and vice versa [26,27]. We calculate the HHI index using firms’ main business income to measure the degree of product market competition. The sample is divided into two groups according to the median size of the index: high-cmp and low-cmp. The results are shown in columns (1)–(4) of

Table 8. Test of industrial competition degree and city level.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	lnPatent		lnCitation		lnPatent		lnCitation
	High_Cmp	Low_Cmp	High_Cmp	Low_Cmp	Imp_City	Gen_City	Imp_City	Gen_City
ESG	0.0851 ***	0.0861 ***	0.0795 ***	0.0914 ***	0.0387	0.1070 ***	0.0184	0.1158 ***
	(0.020)	(0.018)	(0.025)	(0.023)	(0.024)	(0.017)	(0.030)	(0.021)
Growth	0.0008	−0.0753 **	0.0141	−0.0468	−0.0480	−0.0320	−0.0101	−0.0153
	(0.037)	(0.035)	(0.049)	(0.047)	(0.043)	(0.033)	(0.056)	(0.044)
Roa	0.6739 **	1.8375 ***	0.9395 **	2.3876 ***	1.5868 ***	1.1914 ***	2.1249 ***	1.5249 ***
	(0.312)	(0.282)	(0.406)	(0.375)	(0.368)	(0.272)	(0.496)	(0.350)
Lev	−0.3125 **	−0.1062	−0.4285 **	−0.2101	−0.2480	−0.2209 *	−0.4090 **	−0.3384 **
	(0.129)	(0.120)	(0.170)	(0.157)	(0.154)	(0.114)	(0.201)	(0.149)
Top1	−0.0813	−0.4479 ***	−0.2387	−0.5675 ***	−0.5040 **	−0.1536	−0.7553 ***	−0.2335
	(0.180)	(0.171)	(0.233)	(0.219)	(0.221)	(0.160)	(0.280)	(0.205)
Indep	0.6448	0.4881	0.7174	0.0866	0.4351	0.6315	0.2713	0.4862
	(0.525)	(0.446)	(0.677)	(0.570)	(0.608)	(0.434)	(0.758)	(0.562)
Board	0.3285 **	0.1807	0.4208 **	0.1276	0.1584	0.3105 **	0.1055	0.3748 **
	(0.159)	(0.142)	(0.198)	(0.179)	(0.208)	(0.128)	(0.255)	(0.163)
INST	0.1681 *	0.2675 ***	0.2511 **	0.2862 **	0.4103 ***	0.1162	0.5771 ***	0.1074
	(0.094)	(0.087)	(0.124)	(0.112)	(0.116)	(0.080)	(0.150)	(0.104)
ListAge	−0.1072 ***	−0.0801 **	−0.1746 ***	−0.1196 ***	−0.0478	−0.1133 ***	−0.1062 **	−0.1633 ***
	(0.036)	(0.033)	(0.045)	(0.042)	(0.041)	(0.032)	(0.054)	(0.041)
Size	0.4662 ***	0.5707 ***	0.5372 ***	0.6665 ***	0.5727 ***	0.4981 ***	0.6718 ***	0.5765 ***
	(0.034)	(0.030)	(0.041)	(0.037)	(0.035)	(0.032)	(0.043)	(0.038)
Soe	0.1454 **	0.1064	0.1949 **	0.0829	0.0392	0.1897 ***	0.0250	0.2232 ***
	(0.065)	(0.067)	(0.084)	(0.083)	(0.086)	(0.060)	(0.112)	(0.076)
Dual	0.0567	0.0853 **	0.0902 *	0.1160 **	0.1060 **	0.0561	0.1395 **	0.0886 *
	(0.042)	(0.041)	(0.054)	(0.052)	(0.052)	(0.038)	(0.068)	(0.048)
_cons	−10.5085 ***	−12.1320 ***	−11.8257 ***	−13.5545 ***	−11.8964 ***	−11.1729 ***	−13.2558 ***	−12.6197 ***
	(0.806)	(0.748)	(0.969)	(0.895)	(0.978)	(0.706)	(1.159)	(0.853)
N	8259	9795	8259	9795	6287	11,770	6287	11,770
Ind	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.4028	0.4644	0.3806	0.4658	0.4629	0.4327	0.4644	0.4182

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

. First, the effect of ESG scores on the incremental quality of corporate innovation exists regardless of the degree of competition in the industry in which the firm is located. Second, ESG scores show significant differences in the quantity and quality of corporate innovation by level of industry competition. Regardless of the degree of industry competition, no significant difference is found in the quantity of ESG scores on corporate innovation, but the quality of innovation is relatively higher for firms in industries with lower levels of industry competition. The explanation for this is that, compared to industries with low competition, industries with high competition may engage in a large number of low-quality innovation activities due to factors such as rapid market capture, attracting stakeholders’ attention, and recouping capital. Since the sample used in this study is the number of granted invention patents, the invention patents that failed to pass the grant are not included in the sample, which may be the reason for the small number of innovations in industries with a high degree of competition. Therefore, firms in industries with a high degree of competition face more severe corporate survival problems, and there is a deficiency in the incentive effect of corporate ESG performance on the quality of innovation.

4.4.4. Test of City Level

Differences in city economic development levels and institutional environments may result in disparate outcomes of corporate innovation activities [28]. At the same time, the concentration of cities’ economic development levels and institutional environments is reflected in the classification of cities’ hierarchies. Planned cities, sub-provincial cities, and provincial capitals all have obvious differences from other cities in terms of economic development level and institutional environment. Accordingly, the study defines planned cities, sub-provincial cities, and provincial capitals as the key city group (Imp_city) and the remaining prefecture-level cities as the general city group (Gen_city). The results are shown in columns (5)–(8) of Table 8, where the effect of better ESG performance on the quantity and quality of corporate innovation is only present in the general city group. This may be because the level of intellectual property protection and marketization in general cities are weaker than in key cities. In the absence of a good market environment, the performance of ESG can properly alleviate the protection and transformation of patent output caused by intellectual property protection and marketization. Thus, it is helpful for enterprises to obtain more reports of innovation output in time and as much as possible; generally, enterprises in general urban groups are more motivated to innovate through ESG performance.

5. Robustness Test

5.1. Lag Effect of Explained Variable

Considering the proxy variable for innovation used in this study is patent data, there is a long period of R&D time for the innovative subjects promoted by ESG performance to submit patent applications. For this reason, the study lags both the number of invention patents and the number of invention patents cited, the proxy variables of innovation quantity and innovation quality, by one period to verify the regression results of the study. The regression results are shown in columns (1) and (2) of

Table 9. Tests for lag effect of explanatory variable, instrumental variables method, and replacing the measurement method of the explained variable.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Lag Effect		Step 1	Step 2		Substitute Variable
	${\mathbf{lnPatent}}_{\mathbf{t}-1}$	${\mathbf{lnCitation}}_{\mathbf{t}-1}$	ESG	lnPatent	lnCitation	lnPat_Prac	lnCit_Prac	Know_Widthth
ESG	0.1092 ***	0.1187 ***		0.1105 ***	0.1156 ***	0.0731 ***	0.0702 ***	0.0047 **
	(0.016)	(0.020)		(0.012)	(0.016)	(0.015)	(0.015)	(0.002)
L.ESG			0.7364 ***
			(0.007)
Growth	−0.2350 ***	−0.2426 ***	−0.0252	−0.0372	−0.0053	−0.0242	0.0047	−0.0010
	(0.030)	(0.039)	(0.018)	(0.026)	(0.034)	(0.028)	(0.032)	(0.005)
Roa	1.0930 ***	1.4452 ***	1.2319 ***	1.3476 ***	1.6820 ***	1.2555 ***	1.2258 ***	0.1407 ***
	(0.245)	(0.320)	(0.115)	(0.164)	(0.216)	(0.272)	(0.281)	(0.039)
Lev	−0.1838 *	−0.2967 **	−0.2265 ***	−0.1531 ***	−0.2599 ***	0.2235 **	0.0951	−0.0239 *
	(0.108)	(0.141)	(0.040)	(0.058)	(0.077)	(0.108)	(0.111)	(0.014)
Top1	−0.3514 **	−0.5353 ***	0.1067 **	−0.2983 ***	−0.4038 ***	−0.0719	−0.1531	−0.0345 *
	(0.151)	(0.194)	(0.048)	(0.069)	(0.091)	(0.139)	(0.143)	(0.018)
Indep	0.5587	0.4050	0.0562	0.5906 ***	0.4172 *	−0.3819	−0.3498	−0.0804 *
	(0.403)	(0.516)	(0.127)	(0.190)	(0.250)	(0.357)	(0.356)	(0.047)
Board	0.2396 *	0.2417	0.0643	0.2252 ***	0.2314 ***	−0.0961	−0.0485	0.0127
	(0.127)	(0.161)	(0.040)	(0.057)	(0.075)	(0.106)	(0.109)	(0.015)
INST	0.1988 **	0.2360 **	0.0604 **	0.1757 ***	0.1837 ***	0.0547	0.1203	0.0032
	(0.078)	(0.101)	(0.029)	(0.044)	(0.058)	(0.075)	(0.076)	(0.011)
ListAge	−0.1030 ***	−0.1798 ***	0.0152	−0.1118 ***	−0.1654 ***	−0.0266	−0.0656 **	−0.0072 *
	(0.033)	(0.042)	(0.011)	(0.017)	(0.022)	(0.028)	(0.028)	(0.004)
Size	0.5758 ***	0.6792 ***	0.0876 ***	0.5130 ***	0.5856 ***	0.5561 ***	0.5284 ***	0.0343 ***
	(0.027)	(0.033)	(0.008)	(0.011)	(0.014)	(0.025)	(0.027)	(0.003)
Soe	0.1300 **	0.1507 **	0.1366 ***	0.1362 ***	0.1499 ***	−0.0073	0.0111	−0.0065
	(0.055)	(0.071)	(0.017)	(0.023)	(0.031)	(0.049)	(0.051)	(0.007)
Dual	0.0578	0.0876 *	−0.0288 **	0.0688 ***	0.0966 ***	0.0025	0.0204	0.0088 *
	(0.036)	(0.046)	(0.013)	(0.019)	(0.026)	(0.034)	(0.036)	(0.005)
_cons	−12.4865 ***	−14.1930 ***	−0.4958 ***	−12.9494 ***	−14.4449 ***	−9.6676 ***	−9.8414 ***	−0.5630 ***
	(0.641)	(0.782)	(0.159)	(0.244)	(0.321)	(0.558)	(0.595)	(0.059)
N	14,915	14,915	14,915	14,915	14,915	14,732	14,732	18,057
Ind	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.4186	0.4050	0.5956	0.4558	0.4493	0.4395	0.5586	0.1489
			Anderson-Rubin Wald test 50.06 (p = 0.000)

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

. After controlling for industry fixed effects and year fixed effects, the corporate innovation promotion effect due to ESG performance is significantly present.

5.2. Instrumental Variable Method

The research chooses ESG performance lagging by one period as the instrumental variable, and robustness is tested by the two-stage least squares method [94]. First, a firm’s ESG performance in the current period is influenced by that firm’s ESG performance in the previous period, satisfying the endogeneity requirement of the instrumental variable. Second, the firm’s ESG performance in the previous period is not directly related to the firm’s innovation behavior in the current period, satisfying the exogeneity requirement of the instrumental variable. Column (3) of Table 9 reports the regression results of the first stage, and the value of the weak instrumental variable of the Anderson–Rubin Wald test in the first stage is 50.06, which is much larger than 10, strictly rejecting the original hypothesis and empirically testing the assumption that the relevance of the instrumental variable holds. Columns (4) and (5) report the regression results for innovation quantity and innovation quality, respectively. The coefficient values of ESG are 0.1105 and 0.1156, which are significantly positive at the 1% level, respectively, indicating that the conclusion that ESG performance promotes the quantity and quality of corporate innovation is robust.

5.3. Replacing the Measurement Method of the Explained Variable

In the robustness test, the explanatory variables are replaced in this study.

First, the quantity of utility model patents (lnPat_prac) and the number of citations (lnCit_prac) are used as proxy variables for innovation quantity and innovation quality. Columns (6) and (7) of Table 9 indicate that the coefficient values of the core explanatory variables are significantly positive after controlling for industry fixed effects and year fixed effects, i.e., the promotion effect of ESG performance on corporate innovation is significantly present.

Second, the knowledge breadth method is used to measure patent quality. In the previous section, the study referred to the majority of the literature that uses patent citation as a proxy variable for patent quality. In the robustness test, the study measures the quality of patents through the lens of differences in IPC classification numbers [95]. Different patent classification numbers represent different technological fields involved in patents, and the Herfindahl index of each patent’s IPC classification number at the large group level is measured in this study. The formula is as follows: $know\_width=1-{{\displaystyle \sum }}^{\text{}}{\mathsf{\alpha }}_{\mathrm{i}}^2$, where ${\alpha }_i^2$ is the proportion of each large group i in the patent classification number. This formula shows a greater variety of classification numbers of a patent at the large group level; a greater variability among the classification numbers is correlated with a higher quality of the patent. In this study, the knowledge width of each invention patent is summed up according to the mean value to obtain the patent quality data of each enterprise each year. The regression results are shown in column (8) of Table 9, where the coefficient values of the core explanatory variables are significantly positive after controlling for industry fixed effects and year fixed effects. This indicates that the conclusion that ESG performance reflects the quality of corporate innovation is robust.

5.4. Change Sample Statistical Year

In order to exclude possible estimation bias due to specific years, the study continuously subdivides the regression years into four groups in the original time dimension of 2009–2020. As shown in

Table 10. Test for statistical year of change sample.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	2010–2019		2011–2018		2012–2017		2013–2016
	lnPatent	lnCitation	lnPatent	lnCitation	lnPatent	lnCitation	lnPatent	lnCitation
ESG	0.1049 ***	0.1099 ***	0.1171 ***	0.1294 ***	0.1195 ***	0.1312 ***	0.1263 ***	0.1371 ***
	(0.016)	(0.020)	(0.018)	(0.024)	(0.021)	(0.027)	(0.023)	(0.030)
Growth	−0.0313	−0.0019	−0.0267	0.0164	−0.0496	−0.0156	−0.0342	0.0057
	(0.029)	(0.039)	(0.032)	(0.042)	(0.036)	(0.048)	(0.044)	(0.058)
Roa	1.5539 ***	1.9675 ***	1.9157 ***	2.3419 ***	2.3221 ***	2.9629 ***	2.6935 ***	3.4795 ***
	(0.251)	(0.329)	(0.292)	(0.383)	(0.355)	(0.472)	(0.403)	(0.540)
Lev	−0.2419 **	−0.3762 ***	−0.2551 **	−0.4031 ***	−0.2559 **	−0.4458 ***	−0.2024	−0.3885 **
	(0.103)	(0.134)	(0.114)	(0.149)	(0.128)	(0.169)	(0.142)	(0.189)
Top1	−0.3385 **	−0.4717 **	−0.3889 **	−0.5500 ***	−0.3413 *	−0.5357 **	−0.3206 *	−0.4333 *
	(0.146)	(0.186)	(0.159)	(0.202)	(0.174)	(0.225)	(0.192)	(0.248)
Indep	0.5830	0.3985	0.5839	0.3286	0.5575	0.2442	0.5773	0.2819
	(0.393)	(0.497)	(0.422)	(0.534)	(0.470)	(0.598)	(0.524)	(0.668)
Board	0.2630 **	0.2750 *	0.2993 **	0.3157 *	0.3329 **	0.3565 *	0.3301 **	0.3641 *
	(0.122)	(0.151)	(0.131)	(0.165)	(0.146)	(0.186)	(0.161)	(0.205)
INST	0.2478 ***	0.2965 ***	0.2727 ***	0.3399 ***	0.2919 ***	0.3621 ***	0.2364 **	0.2756 *
	(0.077)	(0.099)	(0.086)	(0.112)	(0.098)	(0.128)	(0.111)	(0.145)
ListAge	−0.0959 ***	−0.1528 ***	−0.1010 ***	−0.1590 ***	−0.0993 ***	−0.1642 ***	−0.1053 **	−0.1705 ***
	(0.029)	(0.036)	(0.032)	(0.040)	(0.037)	(0.048)	(0.042)	(0.054)
Size	0.5806 ***	0.6784 ***	0.6155 ***	0.7324 ***	0.6338 ***	0.7762 ***	0.6333 ***	0.7779 ***
	(0.026)	(0.032)	(0.028)	(0.034)	(0.030)	(0.037)	(0.032)	(0.039)
Soe	0.1487 ***	0.1641 **	0.1583 ***	0.1743 **	0.1473 **	0.1701 **	0.1522 **	0.1698 *
	(0.055)	(0.070)	(0.060)	(0.077)	(0.065)	(0.084)	(0.070)	(0.090)
Dual	0.0679 *	0.0952 **	0.0585	0.0799	0.0750 *	0.0916	0.0770	0.0922
	(0.035)	(0.045)	(0.039)	(0.050)	(0.044)	(0.057)	(0.049)	(0.064)
_cons	−12.6453 ***	−14.2995 ***	−13.4181 ***	−15.4545 ***	−13.8274 ***	−16.3121 ***	−13.7862 ***	−16.3176 ***
	(0.621)	(0.748)	(0.653)	(0.785)	(0.708)	(0.853)	(0.759)	(0.928)
N	15,065	15,065	12,122	12,122	8942	8942	5874	5874
Ind	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.4228	0.4109	0.4089	0.3851	0.4064	0.3657	0.4084	0.3585

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

, columns (1) and (2) are the group of regression years 2010–2019, columns (3) and (4) are the group of regression years 2011–2018, columns (5) and (6) are the group of regression years 2012–2019, and columns (7) and (8) are the group of regression years 2013–2016; this shows that, for the different regression years, the effect of ESG performance on the promotion of corporate innovation remains significantly present.

5.5. Other Robustness Tests

Furthermore, to ensure the robustness of the findings in this study, additional robustness tests were conducted. Columns (1) and (2) of

Table 11. Test of change regression models and control joint fixed effects.

	(1)	(2)	(3)	(4)	(5)	(6)
	Mixed OLS		Negative Binomial		Combined Fixation
	lnPatent	lnCitation	lnPatent	lnCitation	lnPatent	lnCitation
ESG	0.0858 ***	0.0863 ***	0.0536 ***	0.0509 ***	0.0842 ***	0.0839 ***
	(0.008)	(0.011)	(0.007)	(0.007)	(0.014)	(0.018)
Growth	−0.0369	−0.0137	−0.0129	0.0008	−0.0409	−0.0204
	(0.024)	(0.032)	(0.017)	(0.019)	(0.027)	(0.035)
Roa	1.3204 ***	1.7392 ***	1.5510 ***	1.6536 ***	1.2706 ***	1.7108 ***
	(0.140)	(0.187)	(0.129)	(0.148)	(0.225)	(0.293)
Lev	−0.2008 ***	−0.3155 ***	−0.1951 ***	−0.2218 ***	−0.1860 **	−0.2828 **
	(0.050)	(0.068)	(0.044)	(0.049)	(0.094)	(0.123)
Top1	−0.2930 ***	−0.4326 ***	−0.2238 ***	−0.2922 ***	−0.3015 **	−0.4583 ***
	(0.065)	(0.085)	(0.049)	(0.054)	(0.132)	(0.169)
Indep	0.6012 ***	0.4381 *	0.0593	−0.1360	0.5792	0.3889
	(0.185)	(0.245)	(0.130)	(0.148)	(0.359)	(0.457)
Board	0.2537 ***	0.2711 ***	0.1500 ***	0.1168 ***	0.2434 **	0.2585 *
	(0.055)	(0.072)	(0.038)	(0.043)	(0.111)	(0.139)
INST	0.2159 ***	0.2628 ***	0.2027 ***	0.1856 ***	0.2230 ***	0.2685 ***
	(0.039)	(0.053)	(0.033)	(0.036)	(0.067)	(0.087)
ListAge	−0.0886 ***	−0.1400 ***	−0.0566 ***	−0.0818 ***	−0.0991 ***	−0.1547 ***
	(0.013)	(0.017)	(0.011)	(0.012)	(0.025)	(0.032)
Size	0.5195 ***	0.6031 ***	0.3727 ***	0.3649 ***	0.5232 ***	0.6082 ***
	(0.011)	(0.014)	(0.007)	(0.008)	(0.024)	(0.029)
Soe	0.1278 ***	0.1401 ***	0.0784 ***	0.0742 ***	0.1336 ***	0.1450 **
	(0.022)	(0.029)	(0.016)	(0.017)	(0.050)	(0.064)
Dual	0.0715 ***	0.1036 ***	0.0367 ***	0.0475 ***	0.0702 **	0.0993 **
	(0.017)	(0.023)	(0.014)	(0.016)	(0.031)	(0.040)
_cons	−11.6208 ***	−12.3958 ***	−9.0204 ***	−7.9062 ***	−11.3749 ***	−12.7503 ***
	(0.254)	(0.329)	(0.158)	(0.186)	(0.584)	(0.696)
N	18,058	18,058	18,058	18,058	18,049	18,049
Ind	Yes	Yes	Yes	Yes	No	No
Year	Yes	Yes	Yes	Yes	No	No
Ind&Year	No	No	No	No	Yes	Yes
r2_a	0.4393	0.4300	0.2044	0.1849	0.4488	0.4418

Note: *, **, and *** denote significance at the levels of 10%, 5%, and 1%, respectively. The standard error of clustering robustness is shown in parentheses. The same applies to later tables.

use a mixed OLS regression model. The study uses a mixed OLS regression model for causality identification, and the regression results show that the values of the core explanatory variable coefficients are significantly positive at the 1% level. Columns (3) and (4) use a negative binomial regression model. In the above regression, the study adopted a logarithmic treatment for the number of patents granted [96], but since patent data are counted data, which is a discrete distribution, referring to the previous literature, the study adopts a negative binomial regression model for robustness testing, and the regression results show that the values of the coefficients of the core explanatory variables are significantly positive at the 1% level. Columns (5) and (6) control joint fixed effects. The study further controls for industry–year fixed effects on top of industry fixed effects and year fixed effects. The regression results show that the core explanatory variables’ coefficient values are significantly positive at the 1% level. All the above results indicate that the findings of the study are robust.

6. Discussion

Traditional valuation theories focus on measuring financial indicators while ignoring the long-term impact of ESG elements on enterprise value, and they are incapable of measuring corporate value comprehensively and scientifically. Constructing a more comprehensive ESG valuation theory can assist management and investors in making better decisions to improve business performance and investment returns. Therefore, this study explored the impact of ESG performance on the quantity and quality of corporate innovation from the perspective of corporate innovation and provides a scientific basis for companies to improve corporate governance to achieve sustainable development.

First, the study found that ESG performance significantly improves firms’ innovation quantity and quality, with clear quantitative and qualitative improvements. Unlike existing studies that only used patent applications to measure corporate innovation [11,18], this study used authorized invention patents that have been substantively reviewed by relevant institutions to measure corporate innovation, which can maximize the elimination of strategic innovation and directly reflect true innovation that contributes to economic growth [4]. Furthermore, the article not only used the number of citations as a proxy variable for innovation quality but also examined the quality of patents from the perspective of patent IPC structure in the robustness test [97], which made the conclusion of innovation quality more convincing. Compared to other studies that used invention patents as a proxy variable for innovation quality, citations are an important indicator of the market value of patents and are more appropriately used as a proxy variable for innovation quality in firm-level analysis [84,85].

Second, unlike existing ESG effects on innovation that were mostly tested from the internal mechanism of financing constraints [17,18], this study performed the analysis through a combination of two paths: the effect of internal control effectiveness on management decisions and analysts’ external monitoring on investors’ decisions. The Theory of Second Best considers that the combination of internal and external factors maximizes the effectiveness of governance. When internal controls are of high quality, a good internal environment, risk assessment, and the design and implementation of control activities can reduce the risk of agency problems. Meanwhile, a well-established external monitoring and feedback mechanism can reduce information asymmetry and align the goals of agents, shareholders, and investors.

Lastly, this study explores the impact of multilevel heterogeneity from the individual to the whole. Corporate innovation is influenced by both internal and external factors [22], which are mainly reflected in differences in patent citations [23,24], corporate property rights, corporate size, management characteristics [25], industry competitiveness [26,27], and city class [28].

7. Conclusions

7.1. Research Conclusions

The ESG rating is a comprehensive assessment report for enterprises based on sustainability, which is different from financial reports that are only available to investors for decision making, and it has a wider range of stakeholders. Good sustainability strategies not only build a good corporate image, but also bring a positive impact on corporate innovation. To this end, the study tests the ESG performance on corporate innovation and the mediating effect of internal control effectiveness and analyst coverage on this relationship, by using data on China’s A-share listed companies from 2009 to 2020. The results show that ESG performance significantly improves innovation quantity, measured by the number of authorized invention patents, and innovation quality, measured by the number of patent citations, with clear features of quantitative and qualitative improvement. The internal and external mechanism tests show that the quantity and quality improvement effect of ESG performance on corporate innovation is dependent on effective internal controls and adequate analyst coverage. The results of the heterogeneity analysis show that improvement in the quality of enterprise-driven innovation is primarily due to knowledge spillover into the domestic market. The additional analysis suggests that the promoting effect is more obvious when the chief executive officer (CEO) serves as a board chairman, the corporation belongs to non-state-owned and large-scale enterprises, the industry in which the market competition is higher, and the corporation is located in a general city.

7.2. Insights and Recommendations

ESG focuses on several aspects: environment (E), social (S), and corporate governance (G). In terms of connotations, G is mainly concerned with how well the company is doing. It is undeniable that, if the E and S aspects are not adequately considered, the business model will not last long. Therefore, the development of many industries is not long-lived because they do not have a deep understanding of E and S issues. This gives rise to the following question: How should stakeholders use the ESG concept to create incremental value?

For management, it is important to look at the sustainability of technological and business model innovation from an ESG perspective. In the past, when developing an innovation project, companies only needed to study the feasibility of their business model, i.e., whether it could make a profit. However, in recent times, companies need multidimensional considerations, i.e., whether it will cause environmental and social problems while making money. This is a long-term solution for the sustainable development of corporate innovation [98], thus more clearly reflecting the incremental value created by management’s capabilities to manage of all elements of ESG.

For investors, it is important to pay attention to corporate information in terms of financial reporting performance and in terms of ESG reports, where companies are comprehensively evaluated [99]. In April 2002, the China Securities Regulatory Commission issued Guidelines on Investor Relations Management for Listed Companies (2022), which included, for the first time, the requirement to consider ESG information in the communication content of investor relations management. ESG includes financial information and nonfinancial information, such as environmental, social, and corporate governance information [72], which is an important basis for investors to evaluate companies and make decisions.

For other sectors, Chinese ESG rating agencies should continue to strengthen two-way communication with mature international ESG rating agencies to promote the establishment of an ESG evaluation system that is in line with international standards and national conditions and better reflects the real situation of corporate ESG. Furthermore, ESG data are mainly provided by statistical agency vendors and greatly affect the value of the data considering the different ESG standards used [17]. Regulators should gradually improve the ESG rating and information disclosure systems to effectively guide analysts, investors, and other stakeholders in providing more efficient financial services for sustainable areas such as corporate innovation. A fair and reasonable reward and punishment mechanism should be set up for ESG rating results to encourage enterprises to disclose ESG information and strive to improve their ratings to provide a real and effective decision-making basis for stakeholders.

7.3. Limitations and Future Research

First, the study in the article focused on Chinese manufacturing listed companies, and the selected industries were not fully representative of other industries. Since the focus of this study was on the Chinese manufacturing sector, future research could extend the analysis to Chinese companies in other industries, and even overseas. Second, the ESG performance in this paper was measured by selecting the composite rating data of ESG companies. ESG contains three dimensional scores of environmental, social, and corporate, and future research can explore the impact of the three dimensions separately on innovation. Lastly, this study focused on the issue of innovation quantity and innovation quality without directly responding to the economic value of innovation. Future research could discuss the economic value of innovation with consideration of the current patent contracts signed between firms.