The Impact of ESG Ratings on Corporate Sustainability: Evidence from Chinese Listed Firms

Abstract

As participants in sustainable development, corporations face the important and controversial issue of whether they can promote corporate sustainability through environmental, social, and governance (ESG) practices. To address this issue, we examine the relationship between ESG performance and corporate sustainability, measured by green total factor productivity (GTFP). Using a panel dataset of 17,559 firm-year observations from non-financial firms listed on the Shanghai and Shenzhen stock exchanges in China between 2011 and 2019, we employ fixed-effects regression models and two-stage least squares (2SLS) with instrumental variables to empirically test the impact of ESG ratings on GTFP, identify the underlying mechanisms, and examine potential heterogeneity across firms. The results show that higher ESG ratings are significantly associated with increased GTFP. Mediation analysis further reveals that this positive relationship operates through reduced financing constraints and enhanced green innovation. Notably, the mediating role of financing constraints is more pronounced for firms with greater reliance on external capital. Heterogeneity analysis indicates that ESG ratings exert stronger effects in eastern regions, pollution-intensive sectors, and state-owned enterprises. These findings provide empirical support for the role of ESG performance as an effective mechanism to advance corporate sustainability through ethics-driven financial access and innovation capability.

1. Introduction

The intensification of global environmental challenges has fundamentally reshaped the context in which firms operate, leading to a shift in how corporate behavior is assessed [1]. Companies are now expected not only to create economic value but also to contribute to broader social and environmental objectives [2]. In this context, the notion of a green economy—characterized by long-term resource efficiency and reduced environmental harm—has gained increasing international attention [3]. This shift has promoted the growth of green investment, referring to capital allocated to projects with clear environmental benefits. These developments have also supported the institutional transformation of smart cities, where digital infrastructure, legal frameworks, and sustainable planning work together to build low-carbon urban systems [4,5]. Such systemic changes have resulted in stricter regulatory requirements, greater stakeholder expectations, and more intense market scrutiny regarding firms’ environmental and social responsibilities. Consequently, companies face growing pressure to demonstrate their sustainability performance through objective and measurable indicators [6]. Among these, environmental, social, and governance (ESG) ratings, defined as an assessment of a company’s overall performance in managing environmental, social responsibility, and governance practices, have become a widely used evaluation tool to identify companies aligned with broader societal goals [7]. However, despite their growing influence on investment decisions and corporate disclosures, it remains uncertain whether ESG ratings enhance firms’ sustainable development outcomes.

High-quality ESG ratings provide stakeholders with comprehensive and consistent data on corporate nonfinancial information, aiding in the integration of corporate social responsibility outcomes into decisions from investment to policymaking, which reflects the deep intertwining of ethical considerations with business operations [8]. The drive to meet growing regulatory and societal expectations often leads companies to strive for alignment with the “best practice” standards set forth by ESG rating agencies [9]. This alignment has the potential to increase firms’ sustainability performance, particularly in aspects that are quantified, emphasized, and valued by widely recognized ratings, underscoring the ethical necessity for businesses to accurately reflect their efforts in sustainability.

Based on criticisms of the quality and potential bias of ESG data, investors often believe that the information disclosed through ESG ratings is inadequate for making informed decisions [10]. Poor evaluations may mislead investments, tarnish reputations, and discourage sustainability efforts. Even with the presence of standards that ensure high-quality ESG data, its use prompts concerns. Investors may overreact, positively or negatively, to ESG disclosures [11], which may be captured by financial markets, leading to short-term gains but long-term setbacks. Moreover, superficial compliance practices such as “greenwashing” for better ESG scores raise significant ethical concerns and invite questions about the authenticity and transparency of their sustainability initiatives [12]. Hence, whether ESG ratings function as effective instruments for promoting corporate sustainability remains an empirical question.

Total factor productivity (TFP) has long served as a core metric for evaluating firm-level efficiency, reflecting the portion of output not attributable to capital and labor inputs [13,14]. As global attention shifts toward corporate sustainability, scholars have increasingly examined how ESG performance influences TFP. The prevailing empirical evidence suggests that firms with stronger ESG engagement generally experience enhanced TFP, as ESG initiatives can reduce financing frictions, improve risk management, and stimulate innovation-driven efficiency gains [15,16]. For instance, Piserà et al. [17] analyze over 400 non-financial firms across 15 European countries and confirm a positive association between ESG scores and TFP, with the environmental pillar exerting the strongest influence. In China, similar findings emerge: Deng et al. [18] and Ding et al. [19] demonstrate that firms with higher ESG ratings benefit from lower borrowing costs and stronger investor confidence, which together ease capital constraints and support productivity-enhancing investments. These studies underscore the cross-national consistency of ESG’s productivity effects, extending beyond firm ethics to measurable economic outcomes. However, the relationship is not necessarily linear. Some scholars have identified an inverted U-shaped association, wherein moderate ESG engagement enhances productivity, but excessive ESG expenditure may divert resources from core operations, leading to diminishing returns [20,21]. Such findings highlight the importance of strategic balance in ESG implementation, and the potential trade-offs involved. Taken together, these studies highlight TFP’s central role in empirical work on high-quality growth and sustainability—almost all major investigations of ESG or green investment now examine TFP as an outcome (or mediator) of sustainable practices.

However, scholars increasingly caution that conventional TFP measures may misrepresent true sustainability performance. Traditional TFP accounts for only marketed inputs and desirable outputs, ignoring “bad” outputs like waste and emissions. This omission can bias productivity estimates [22]. Based on Organization for Economic Co-operation and Development (OECD) [23], standard TFP measures often fail to account for negative externalities such as pollution, while simultaneously incorporating firms’ pollution abatement costs, which leads to a distorted assessment of actual performance and technological progress. In other words, firms that emit a lot of pollutants can appear artificially productive if one neglects their environmental footprint. To address this, the literature has introduced green TFP (GTFP) metrics. GTFP extends the production function to include resource use and emissions alongside labor and capital. For example, Ye et al. [24] emphasize that GTFP integrates resource consumption and pollution emissions into the production function and thus places a greater emphasis on the sustainability of the economy.

Existing studies on GTFP have focused primarily on computational methods and influencing factors. GTFP calculations mainly include parametric and nonparametric methods. Among parametric methods, stochastic frontier analysis (SFA) is the most popular, considering the influence of stochastic factors on the results. However, it is limited to “multiple inputs, single output” scenarios and is susceptible to model-setting errors [25]. On the nonparametric side, Data Envelopment Analysis (DEA) and its extended models dominate the field [26]. Traditional DEA, while measuring GTFP, necessitates proportional changes in inputs and outputs and focuses only on economic gains, sidelining energy inputs and environmental costs. To address these limitations, Tone (2001) [27] introduced the non-directional, relaxed, and non-radial slack-based model (SBM) that solves deviations caused by radial and angular directions during efficiency measurement. In reality, when multiple decision-making units (DMUs) are effective simultaneously, sorting and comparing their effectiveness is challenging [28]. In response, Tone (2002) [29] proposed the super-SBM that could further distinguish efficient DMUs. Since then, many scholars have adopted it to assess GTFP at different levels, such as national [30], industrial [31], and corporate [28]. Meanwhile, there is extensive research examining the factors that influence GTFP, such as investment [32], trade [33], the digital economy [34], environmental regulation [26], and innovation [28]. Yet, there is no consensus on the impact of most factors on GTFP. For example, some argue that the essence of implementing environmental regulation is to internalize firms’ negative externalities, which increases operating costs and hinders GTFP growth [26]. However, proponents of Porter’s hypothesis contend that well-designed environmental regulations can enhance GTFP by promoting firm innovation and structural optimization [35]. Nonetheless, the existing literature has paid limited attention to the impact of ESG ratings on GTFP.

China’s significant economic growth has led to serious ecological deterioration. As the “world’s factory”, its enterprises bear a substantial burden of international transfer emissions, with over 80% of environmental pollutants originating from enterprise production [36]. China has embraced ESG, which reflects business ethics, as a strategy for fostering corporate sustainability. Its ESG ratings started relatively late but developed rapidly. An increasing number of companies are being assessed by both domestic and international rating agencies. These ESG ratings for companies offer a value proposition that can be shared by multiple market players [37], but they also face numerous challenges, such as relatively low utilization by investors [38]. Given that China is both a major polluter and an active advocate of ESG, the impact of its ESG ratings on GTFP is an important reference for the global community, especially developing countries, in terms of sustainable transformation.

To empirically investigate the role of ESG ratings in promoting corporate sustainability, we construct a panel dataset of 17,559 firm-year observations from Chinese A-share listed companies during 2011–2019. By applying a two-way fixed effects model and instrumental variable regression, we examine the relationship between ESG ratings and corporate GTFP. Our analysis reveals a significant positive association between ESG ratings and GTFP. Moreover, we find that this relationship is mediated through reduced financing constraints and enhanced green innovation. These mediating effects are particularly pronounced in firms with higher dependence on external capital. Heterogeneity analysis further shows that the ESG–GTFP link is stronger in eastern regions, pollution-intensive sectors, and state-owned enterprises, underscoring the conditional effectiveness of ESG practices in fostering sustainable development.

This study makes several contributions. First, it takes the lead in exploring the causal relationship between business ethics practices, quantified by ESG, and sustainable development, quantified by GTFP, demonstrating that ESG advantages have a tangible positive impact on GTFP. In doing so, this study extends the discourse on the intrinsic value of business ethics, framing ESG ratings as a carrot that steers business entities towards a sustainable and ethically conscious future.

Second, the research identifies financial constraints and green innovation as key mediators in the ESG–GTFP relationship, revealing how ethically driven financial strategies and innovation practices enable firms to improve sustainable performance. Specifically, the alleviation of financial constraints and advancements in green innovation, both driven by business ethics advantages, are critical to a firm’s long-term sustainability. We further reveal that for companies heavily reliant on external financing, the mediating role of financial constraints is more significant than that of companies with minimal reliance on external financing.

Finally, our analysis of how regional, industrial, and ownership characteristics weaken or strengthen the ESG–GTFP relationship provides deeper insights into the variability of ESG impacts. We find that ESG ratings are more effective in promoting GTFP only in developed regions, heavily polluting industries and state-owned enterprises. These findings suggest that policymakers and businesses should design targeted interventions based on the specific dynamics of regions, industries, and firms to foster corporate sustainability.

The remainder of this paper is organized as follows. Section 2 develops the hypotheses. Section 3 explains our sample, data, and methodology. Section 4 reports the baseline results, including robustness checks, heterogeneity, and mediation effects. Section 5 summarizes the main conclusions and discusses the policy implications and research limitations.

2. Hypothesis

2.1. ESG Ratings and GTFP

According to stakeholder theory, firms that actively engage in corporate social responsibility (CSR), as systematically measured through environmental, social, and governance (ESG) ratings, establish stronger stakeholder relationships, thereby supporting sustained growth and corporate resilience [39,40]. ESG ratings capture not merely isolated ethical behaviors but reflect a firm’s structured commitment to sustainability, transparency, and stakeholder responsiveness [41,42]. Empirical evidence highlights that stakeholders, particularly younger, environmentally conscious consumers, increasingly pressure corporations to enhance their ESG practices, shaping firm performance through heightened stakeholder expectations [43]. Firms with high ESG ratings thus attract stakeholder support, facilitating stable cooperative relationships essential for resource efficiency, operational consistency, and ultimately enhanced GTFP.

From a signaling theory perspective, robust ESG ratings communicate credible and transparent signals regarding a firm’s long-term strategic orientation and sustainability alignment, thereby enhancing trust among business partners and investors [44]. Belas et al. [45] empirically support this by demonstrating that companies adopting comprehensive ESG standards significantly improve their market position, attract reliable business partners, and reduce transaction costs through enhanced reputational capital and trust. These empirical findings affirm that signaling sustainability commitment reduces informational asymmetry, enhances resource allocation efficiency, and facilitates efficient market interactions, thereby positively influencing overall GTFP.

Agency theory further elucidates how ESG performance directly relates to GTFP. Superior ESG ratings indicate strong corporate governance mechanisms and aligned managerial incentives, mitigating agency conflicts between shareholders and management. Such governance structures ensure managerial decisions consistently prioritize sustainable value creation rather than short-term benefits, directly contributing to higher productivity and environmental efficiency [46]. Rubáček et al. [47] provide empirical validation by showing that mandatory ESG disclosure, as institutionalized in the EU regulatory frameworks, significantly reduces agency costs and improves managerial accountability and corporate governance effectiveness. This improvement in governance transparency thus contributes directly to enhanced sustainability performance and higher GTFP.

Moreover, according to social identity theory, ESG-driven corporate actions positively shape employee and consumer identification with the firm’s values, fostering internal cohesion, productivity, and market success [48]. Pelikanova et al. [49] empirically demonstrate that firms achieving authentic ESG performance effectively build stronger internal stakeholder identification and external brand loyalty, while firms perceived as engaging in superficial ESG reporting risk losing stakeholder trust, thereby undermining productivity and growth potential. Enhanced identification and cohesion motivate employees to engage actively in efficiency-enhancing practices, thereby positively influencing GTFP. Based on the above theoretical and empirical insights, we hypothesize the following:

H1: 

ESG ratings are positively associated with GTFP.

2.2. The Mediating Role of Financing Constraints

According to signaling theory, ESG ratings function as credible signals that effectively communicate a firm’s strategic alignment with sustainability and robust governance practices, thereby substantially reducing the information asymmetry faced by external investors [44,50]. Transparent ESG disclosures signal reduced operational risks, higher management quality, and stronger long-term growth potential, leading to more favorable market perceptions and lower risk premiums [51,52]. Empirical studies further support this perspective. For instance, Chava [53] demonstrates that firms with poor environmental performance face significantly higher costs of debt capital, underscoring the financial market’s sensitivity to environmental risks. Similarly, Yan et al. [54] and Guo et al. [55] find robust evidence in Chinese capital markets indicating that firms with superior ESG performance achieve significantly lower borrowing costs, better credit ratings, and enhanced capital access, largely due to reduced informational opacity and improved investor trust. Moreover, European regulatory experience, such as the EU’s Corporate Sustainability Reporting Directive (CSRD), provides further validation, demonstrating that mandatory ESG reporting significantly alleviates financing constraints by enhancing corporate transparency and accountability, thus attracting more long-term institutional investment [47,56].

Financing is essential for firms to achieve sustainable development, yet financing constraints can hinder the execution of diverse GTFP-oriented strategic investment initiatives by firms. By alleviating financing constraints, ethically minded enterprises can efficiently allocate capital toward GTFP. Specifically, first, firms with eased financing constraints can optimize capital flows by reducing their reliance on high-pollution, high-energy projects [57]. Second, firms with sufficient funding can carry out previously unfeasible but GTFP-enhancing investments, such as increased research and development (R&D) activities, technological upgrades, and operational efficiency improvements, thereby better practicing sustainable development [58,59]. Finally, with secure funding, companies are more likely to fulfill their environmental compliance obligations, reducing shutdown risks and ensuring stable green production [57]. Therefore, the ESG rating advantages of firms help ease financing constraints, enhance capacity for sustainable productivity improvements, and thus positively impact GTFP. Based on the above arguments, we propose the following hypothesis:

H2: 

Financing constraints mediate the relationship between ESG ratings and GTFP.

2.3. The Mediating Role of Green Innovation

From the perspective of upper echelons theory, managerial strategic decisions are profoundly influenced by executive values, priorities, and incentive structures [60]. When ESG metrics are explicitly integrated into executive performance evaluations and corporate governance frameworks, managerial incentives shift toward prioritizing green and sustainable innovation strategies [61]. Empirical evidence consistently supports this argument: Wu et al. [28] report that firms that incorporate ESG criteria into executive evaluations generate significantly more green patents, environmental innovations, and sustainable products, indicating enhanced innovation output driven by ESG-oriented managerial incentives. Moreover, institutional theory further reinforces this mechanism: stringent ESG regulatory frameworks, such as mandatory sustainability disclosures under the EU’s CSRD, generate institutional pressures compelling firms to innovate sustainably and maintain competitive legitimacy [56,62]. Chen et al. [63] confirm that regulatory environments characterized by high ESG demands actively shape corporate strategies toward innovation, aligning compliance obligations with market advantages. Thus, strong ESG performance significantly incentivizes firms to pursue green innovation.

Green innovation is a critical driver of corporate sustainability. On one hand, companies can reduce pollution emissions, improve resource utilization efficiency, and reduce operational costs through green process innovations (e.g., dismantling outdated equipment, optimizing production techniques, and developing new processes), all of which are conducive to enhancing GTFP. Extensive empirical research supports this linkage. For example, Zhang et al. [64] demonstrate empirically that green innovation significantly improves firms’ environmental efficiency and resource productivity, leading directly to increased total factor productivity. In a study across multiple industries, Hamid et al. [65] identify green technological advancements as the primary driver behind sustained productivity improvements in emerging economies. Similarly, Zhang and Li [66] report that firms adopting comprehensive digital and environmental innovation frameworks demonstrate consistently higher GTFP levels, due primarily to enhanced operational efficiencies and lower environmental liabilities. On the other hand, firms can develop green products by incorporating sustainable designs and using eco-friendly materials to meet the dual demands of regulators and consumers. This not only helps firms achieve synergy between economic performance and environmental protection but also improves resource efficiency, thereby enhancing GTFP. Mady et al. [67] emphasize the importance of incorporating sustainable designs and eco-friendly materials in products to meet regulatory requirements and consumer expectations, thereby enhancing competitiveness and sustainability. In summary, green innovation aims to develop processes and products with minimal environmental impact throughout their entire lifecycle, with the key focus on guiding companies to adopt more sustainable production methods. Therefore, we propose the following hypothesis:

H3: 

Green innovation mediates the relationship between ESG ratings and GTFP.

3. Research Design

3.1. Sample and Data Sources

The sample for this research comprises A-share firms from 2011 to 2019, considering data availability and the economic disruptions caused by the 2020 pandemic. A-shares are the stock shares of mainland China-based companies that are traded in RMB on the Shanghai Stock Exchange (SSE) and the Shenzhen Stock Exchange (SZSE), primarily by domestic investors. China’s A-share market is the second largest in the world, second only to the United States stock market. Selecting A-share listed companies as the research sample is representative and realistic. Several criteria are used to clean the sample [68]: (1) exclude ST (Special Treatment), *ST (Particular Special Treatment), and PT (Particular Transfer) enterprises; in China’s stock markets, companies are labeled as ST if they incur losses for two consecutive years, *ST if losses continue for three years or if they exhibit severe financial issues such as insolvency, and PT refers to firms already delisted and transferred to special trading boards; (2) exclude firms with missing dependent, independent variables, or those with more than two missing control variables; (3) exclude financial firms; and (4) winsorize all continuous variables at the 1% and 99% levels. Ultimately, the cleaned sample consisted of 17,559 observations from 1951 firms, forming a balanced panel. All data are from the China Stock Market and Accounting Research (CSMAR) database, the China Research Data Service Platform (CNRDS) database, the Wind Financial Terminal, and the China City Statistical Yearbook.

3.2. Variable Measures

3.2.1. Dependent Variable: GTFP

We select the GTFP index as a proxy for corporate sustainability, which is calculated via the super-SBM and ML index methods. Considering that the essence of GTFP lies in balancing environmental conservation with economic growth, the inputs for GTFP assessment include capital inputs (measured by the firm’s net fixed assets), labor inputs (measured by the firm’s number of employees), and energy consumption (estimated by multiplying the industrial electricity consumption of the firm’s prefecture-level city by the ratio of its operating income to the local GDP). The outputs are categorized into two categories: desirable outputs, measured by the firm’s total operating revenue, and undesirable outputs, measured by multiplying the total emissions of wastewater, dust, and sulfur dioxide in the company’s prefecture-level city by the ratio of the firm’s operating income to the local GDP.

We regard each listed company as a decision-making unit (DMU), and there are n DMU records denoted as DMU_k (k = 1, 2, …, n), with a total of t periods (t = 2011, …, 2019), to construct the optimal production technology boundary. Each DMU has three types of input–output indices [26], including m types of input x_i (i = 1, 2, …, m), q₁ types of expected output y_r (r = 1, 2, …, q₁), and q₂ types of undesirable output b_t (t = 1, 2, …, q₂). The production possibility can be expressed as follows:

$$P(x_i)=\left\{\left.(x_i,y_r,b_i)\right|x_i\ge {\displaystyle \sum _{k=1}^n{\lambda }_k{x}_k,y_r\le {\displaystyle \sum _{k=1}^n{\lambda }_k{y}_r},b_t\le {\displaystyle \sum _{k=1}^n{\lambda }_k{b}_t},\lambda \ge 0}\right\}$$

(1)

where subscript k denotes the DMU under measurement and where λ represents the nonnegative weight vector assigned to the output and input factors. According to Equation (1), the super-SBM for undesired outputs can be expressed as

$$\mathit{min}{\rho }_{\mathit{k}}={\displaystyle \frac{1-\frac{1}{\mathit{m}}{\displaystyle {\sum }_{\mathit{i}=1}^{\mathit{m}}\frac{S_{\mathit{i}}^{-}}{{\mathit{x}}_{\mathit{i}0}}}}{1+\frac{1}{{\mathit{q}}_1+{\mathit{q}}_2}\left({\displaystyle {\sum }_{\mathit{r}=1}^{{\mathit{q}}_1}\frac{S_{\mathit{r}}^{+}}{{\mathit{y}}_{\mathit{r}0}}}+{\displaystyle {\sum }_{\mathit{k}=1}^{{\mathit{q}}_2}\frac{S_{\mathit{t}}^{\mathit{b}-}}{{\mathit{b}}_{\mathit{r}0}}}\right)}}$$

(2)

$$\mathit{s}.\mathit{t}.\left\{\begin{array}{l}{\displaystyle {\sum }_{\mathit{j}=1}^{\mathit{n}}{\mathit{x}}_{\mathrm{ij}}{\lambda }_{\mathit{j}}+\stackrel{\_}{S_i}=}{\mathit{x}}_{\mathit{i}0}\hspace{1em}\hspace{1em}\hspace{1em}i=1,2,\dots m;\\ {\displaystyle {\sum }_{\mathit{j}=1}^{\mathit{n}}{\mathit{y}}_{\mathrm{ij}}{\lambda }_{\mathit{j}}-S_{\mathit{r}}^{+}=}{\mathit{y}}_{\mathit{r}0}\hspace{1em}\hspace{1em} r=1,2,\dots q_1;\\ {\displaystyle {\sum }_{\mathit{j}=1}^{\mathit{n}}{\mathit{b}}_{\mathrm{tj}}{\lambda }_{\mathit{j}}+S_{\mathit{t}}^{b-}=}{\mathit{b}}_{\mathit{t}0}\hspace{1em}\hspace{1em} t=1,2,\dots q_2;\\ {\displaystyle {\sum }_{\mathit{j}=1}^{\mathit{n}}{\lambda }_{\mathit{j}}=}1\\ \lambda \ge 0({\forall }_j),S_{\mathit{i}}^{-}\ge 0({\forall }_i),S_{\mathit{r}}^{+}\ge 0({\forall }_r),S_{\mathit{t}}^{b-}\ge 0({\forall }_t)\end{array}\right.$$

where $\overline{S_i}$, $S_r^{+}$, and $S_t^{b-}$ represent the slack values of the inputs, desired outputs, and undesired outputs, respectively. ρ_k denotes the superefficiency value of the evaluated DMU. If the production efficiency of the j-th DMU improves, this value is greater than or equal to 1; otherwise, it is less than 1.

Given that the ML index represents the change in GTFP from period t to t + 1, which lacks comparability [69], we transform it into a cumulative index. The firm’s GTFP in the base period is assumed to be 1. This value is then multiplied by the ML index for each period to produce the firm’s GTFP from 2011 to 2019.

3.2.2. Independent Variable: ESG Score

We use Huazheng’s ESG ratings as a proxy for corporate ESG performance, aiming to accurately reflect firms’ ethical practices. Huazheng ESG rating developed by Shanghai Huazheng Index Information Service Co., Ltd., is widely used to assess the ESG performance of A-share listed companies in China. Apart from the Huazheng ESG rating, there are several ESG rating systems in China, such as CSI ESG Rating, Shangdao Ronglv ESG Rating, Wind ESG Rating, FTSE Russell ESG Rating, Jia Shi ESG Rating, and Social Value Investment Alliance ESG Rating. Compared to other ratings, the Huazheng ESG rating has the following advantages. First, drawing on the framework of internationally recognized ESG evaluation systems, the Huazheng ESG rating has been customized for the Chinese market, policies, and listed companies by excluding or refining specific indicators. The system consists of three pillars: environmental, social, and governance. Each pillar encompasses multiple topics, covering a total of 44 key concerns, thereby providing a more comprehensive reflection of a company’s ethical performance. Second, about 78% of the data in the Huazheng ESG rating comes from publicly available and high compliant company disclosures, such as regular reports, social responsibility reports, and interim announcements. Third, the Huazheng ESG rating covers all A-share listed companies and offers relatively long time-series data, dating back to 2009. This rating system divides the ESG levels of listed enterprises into nine grades: C, CC, CCC, B, BB, BBB, A, AA, and AAA. Following Lin et al. (2021) [70], the ESG ratings in this study are mapped to numerical values ranging from 10 to 90, with higher values indicating better ESG performance.

3.2.3. Control Variables

In accordance with the literature [71], we enter a set of firm-level control variables as follows: (1) firm size (Lnl), measured as the logarithm of the number of employees; (2) establishment age (Lnage), calculated as the logarithm of “current year—establishment year + 1”; (3) management expenses (Lnm), calculated as the natural logarithm of management expenses; (4) financial leverage (Lev), measured as total liabilities divided by total assets; (5) asset liquidity (Liq), measured by current assets divided by total assets; (6) return on assets (Roa), calculated as net profit divided by total assets; (7) return on equity (Roe), defined as net income divided by shareholders’ equity; (8) return on human capital (Rop), defined as net profit divided by total employee compensation and benefits; and (9) firm value (Tobin’s Q), measured by the sum of equity market value and net debt market value divided by total assets.

3.2.4. Baseline Model

Referring to Attig (2023) [72], this study employs a two-way fixed effects model—an econometric approach that controls for unobserved heterogeneity across firms and time—to estimate the effect of ESG ratings on GTFP. By accounting for both firm-specific and time-specific effects, this method effectively mitigates bias caused by time-invariant omitted variables and unobserved macroeconomic shocks, thereby improving the credibility of the estimation results.

$$GTF{P}_{\mathit{it}}={\beta }_0+{\beta }_{1}ES{G}_{\mathit{it}}+{\displaystyle {\sum }_{i=2}^9{\beta }_{i}Contro{l}_{it}+{\phi }_{\mathrm{i}}+{\mu }_t+{\epsilon }_{it}}$$

(3)

where the subscripts i and t refer to the firm and year, respectively. GTFP_it denotes firm sustainability, and ESG_it represents the ESG score of a firm. Control_it is a set of firm-level variables as defined above, and ε_it is the random error term. φ_i and μ_t refer to the fixed effects controlling for firms and years, respectively. β₀ is the intercept, and β₁ is the coefficient of ESG.

4. Empirical Results

4.1. Descriptive Statistics

Table 1. Descriptive statistics.

Variable	Variable Description	N	Mean	Standard Deviation	Minimum	Maximum
GTFP	Green total factor productivity	17,599	1.219	0.729	0.038	25.346
pESG	Bloomberg ESG score	7928	20.465	6.601	1.240	59.917
ESG	Huazheng ESG Rating Index	17,599	40.553	11.205	10.000	80.000
Lnage	Establishment age	17,599	2.922	0.302	0.000	4.174
Lnl	Enterprise scale	17,599	7.780	1.333	1.609	12.722
Lnm	Management expense	17,599	5.158	1.198	0.315	10.351
Lev	Asset liability ratio	17,599	46.037	41.660	−19.470	3146.670
Flow	Ratio of current assets to total assets	17,599	55.538	21.043	0.868	100.000
Roa	Return on assets	17,599	5.633	21.571	−215.894	2078.546
Roe	Return on equity	17,433	64.232	127.130	−15,824.416	1104.102
Rop	Return on corporate manpower investment	17,336	142.233	1506.452	−414,58.569	180,408.502
Tobin’s Q	Firm value	17,069	2.272	8.877	0.684	729.629

summarizes the descriptive statistics for each variable. The mean ESG score is 40.533, which is consistent with the findings of Deng et al. (2023) [18], indicating that the average ESG ratings of companies fall between B and BB. The maximum ESG value is 80, suggesting that no company achieved an AAA rating during the sample period. For the Bloomberg ESG score (pESG), the mean is 20.465, with a standard deviation of 6.601, a minimum of 1.240, and a maximum of 59.917, indicating a lower average ESG disclosure level under the Bloomberg system. The corporate GTFP values fluctuate between 0.038 and 25.346. The generally low levels of GTFP, with a mean of 1.219, align with the results of Sun et al. (2023) [73], highlighting ample room for improvement in corporate GTFP. The values of the control variables are in line with the actual situation of listed companies and the statistical findings of the literature. For example, the mean of lev is close to the result of Li et al. (2023) [74]. The descriptive statistics indicate that there are no outliers in the sample that would affect the subsequent analysis.

4.2. Baseline Regression

Column (1) of

Table 2. Results of the basic regression.

Variable	(1) GTFP	(2) GTFP
ESG	0.0011 *** (0.0004)	0.0013 *** (0.0004)
Lnage	0.5698 *** (0.0230)	−0.5235 *** (0.0786)
Lnl	−0.0991 *** (0.0092)	−0.1048 *** (0.0091)
Lnm	0.1410 *** (0.0097)	0.1358 *** (0.0098)
Lev	0.0024 *** (0.0003)	0.0030 *** (0.0003)
Liq	0.0041 *** (0.0003)	0.0041 *** (0.0003)
Roa	−0.0047 *** (0.0013)	−0.0038 ** (0.0013)
Roe	0.0014 *** (0.0005)	0.0013 *** (0.0005)
Rop	0.0002 *** (0.0000)	0.0002 *** (0.0000)
Tobin’Q	0.0238 *** (0.0033)	0.0126 *** (0.0037)
Constant	−0.8573 *** (0.0811)	2.3906 *** (0.2360)
N	16,767	16,767
r2	0.5336	0.5533
F	161.5371	58.6720
Firm FE	Yes	Yes
Year FE	No	Yes

Note: *** and ** indicate statistical significance at the 1% and 5% levels, respectively.

presents the regression results, where the firms’ fixed effects are included to control for microlevel, unobservable, and time-invariant company characteristics. Column (2) augments the regression model by incorporating year fixed effects, which mitigate bias by controlling for common time-varying factors across firms. The coefficients of ESG performance in Columns (1) and (2) are statistically significant at the 1% levels, with corresponding values of 0.0011 and 0.0013, respectively. These results indicate a significant positive correlation between ESG ratings and GTFP, which supports H1. Specifically, for every one-level increase in ESG ratings, the corporate GTFP level increases by 13%, as shown in Column (2).

4.3. Robustness Test

Three methods are employed to verify the robustness of the results. First, following Arouri and Pijourlet (2017) [75], the core explanatory variable, the Huazheng ESG rating (ESG), is replaced with the Bloomberg ESG score (pESG). The Bloomberg ESG data framework covers over 15,000 companies worldwide. It sources information from 25 types of public documents and ensures that the data represents at least 80% of a company’s operations and workforce [76]. Column (1) of

Table 3. Results of robustness testing.

Variable	(1)	(2)	(3)	(4)
	GTFP	GTFP	GTFP	GTFP
ESG		0.0013 * (0.0006)	0.0013 ** (0.0006)
pESG	0.0042 *** (0.0013)
L.ESG				0.0011 ** (0.0004)
Control	Yes	Yes	Yes	No
L.Control	No	No	No	Yes
Constant	1.2091 *** (0.3825)	2.3906 *** (0.5572)	2.3906 ** (1.1124)	2.2942 *** (0.2726)
N	7705	16,767	16,767	14,833
r2	0.5851	0.5533	0.5533	0.5875
F	24.3362	143.2212	16.1188	29.5891
Firm FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Industry clustering	No	Yes	No	No
City clustering	No	No	Yes	No

Note: ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

shows that the coefficient of ESG performance is 0.0042, which is significant at the 1% level. This coefficient is larger than the base regression’s coefficient of 0.0013, likely due to differences in ESG scores arising from the varying rating criteria used by the two agencies. Second, consistent with Aouadi and Marsat (2018) [77], we examine the robustness of the clustering hierarchy for standard errors. The results in columns (2)–(3) show that the ESG coefficients for both industry and regional clusters are equal and significantly positive at the 10% and 5% level, respectively, suggesting that the conclusion is robust and unaffected by changes in the clustering hierarchy. Third, we account for the potential time lag effect of ESG ratings on enterprise GTFP performance. Specifically, drawing on Ge et al. (2022) [78], we reexamine the baseline model with all explanatory and control variables lagged by one period. As shown in column (4), the main explanatory variable retains its sign and statistical significance, further demonstrating the robustness of the results.

4.4. Instrumental Variable Estimation

Although the fixed effects model can mitigate endogeneity problems caused by omitted variables by controlling for numerous unobservable variables, endogeneity arising from the bidirectional causality between business ethics and sustainable development still exists [68]. To address this issue, we refer to Chu et al. (2023) [79] and employ a two-stage least squares (2SLS) regression with instrumental variables. This method helps to isolate the exogenous variation in ESG ratings that is not driven by unobserved confounders or reverse causality. Compared with ordinary least squares (OLS), 2SLS improves the consistency and reliability of coefficient estimates when explanatory variables are endogenous [80]. The first instrumental variable is the ESG industry mean (IV1), which is calculated as the average ESG score of all firms in a specific industry and year, excluding the ESG score of firm i in the same year. The rationale for IV1 is that a company’s ESG score is likely correlated with that of its industry peers due to similar external developmental conditions; however, the industry average is unlikely to directly influence the company’s GTFP. Thus, the ESG industry mean serves as a valid instrumental variable. The second instrumental variable, a one-period lagged ESG score (IV2), is strongly associated with endogenous elements but uncorrelated with current-period disruptions that have already occurred; this ensures that the exogeneity condition is satisfied. We then conduct the 2SLS regression as follows:

$${\mathit{ESG}}_{\mathit{it}}={\mathit{\beta }}_0+{\mathit{\beta }}_{1}I{V}_{\mathit{it}}+{\displaystyle {\sum }_{i=2}^9{\mathit{\beta }}_{i}Control}+{\mathit{\phi }}_i+{\mathit{\mu }}_t+{\mathit{\epsilon }}_{it}$$

(4)

$$GTF{P}_{\mathit{it}}={\mathit{\beta }}_0+{\mathit{\beta }}_1\stackrel{\wedge }{ES{G}_{\mathit{it}}}+{\displaystyle {\sum }_{i=2}^9{\mathit{\beta }}_{i}Control}+{\mathit{\phi }}_i+{\mathit{\mu }}_t+{\mathit{\epsilon }}_{it}$$

(5)

Here, Equations (4) and (5) represent the first- and second-stage regressions, respectively, where $\widehat{{ESG}_{it}}$ is the fitted value estimated in the first stage, IV_it refers to the instrumental variable, and the other variables are defined consistently with Equation (3).

Columns (1)–(4) of

Table 4. Results of the 2SLS regressions using instrumental variables.

	ESG	GTFP	ESG	GTFP
IV1	0.3051 *** (0.0685)
IV2			0.2577 *** (0.0088)
ESG		0.0105 *** (0.0027)		0.0035 ** (0.0018)
Control	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Kleibergen–Paap rk LM statistic	729.281 (0.0000)		588.792 (0.0000)
Kleibergen–Paap rk Wald F statistic	966.888		856.786
N	16,767	16,769	14,915	14,915
r2	0.5824	0.0112	0.6299	0.0301
F	44.9059	30.5132	120.1906	31.3569

Note: *** and ** indicate statistical significance at the 1% and 5% levels, respectively.

present the results of the 2SLS regressions using instrumental variables. The Kleibergen–Paap rk LM statistics are significant at the 1% level, strongly rejecting the null hypothesis that the instrumental variables are unidentifiable. Moreover, the Kleibergen–Paap rk Wald F statistic for the two stages are 966.888 and 856.786, respectively, both significantly exceeding the Stock–Yogo critical value of 16.38 at the 10% significance level, indicating that there is no weak instrumental variable problem. After potential endogeneity concerns with two appropriate instrumental variables and 2SLS are addressed, the regression coefficients of ESG in columns (2) and (4) are significantly positive at the 1% and 5% levels, respectively, further validating the robustness of the estimation results.

4.5. Mechanistic Analysis

We adopt the three-step approach proposed by Muller and Judd (2005) [81], a widely recognized method for mediation analysis, to test the mechanisms in H2 and H3. This approach enables us to investigate how ESG performance impacts GTFP by systematically examining the pathways through which ESG affects productivity. The following system of equations is applied:

$$GTF{P}_{\mathit{it}}={\mathit{\beta }}_0+{\mathit{\beta }}_{1}ES{G}_{\mathit{it}}+{\displaystyle {\sum }_{i=2}^9{\mathit{\beta }}_{i}Contro{l}_{it}+{\mathit{\phi }}_i+{\mathit{\mu }}_t+{\epsilon }_{it}}$$

(6)

$${\mathit{Mediator}}_{\mathit{it}}={\mathit{\beta }}_0+{\mathit{\beta }}_{1}ES{G}_{\mathit{it}}+{\displaystyle {\sum }_{i=2}^9{\mathit{\beta }}_{i}Contro{l}_{it}+{\mathit{\phi }}_i+{\mathit{\mu }}_t+{\mathit{\epsilon }}_{it}}$$

(7)

$$GTF{P}_{\mathit{it}}={\mathit{\beta }}_0+{\mathit{\beta }}_{1}ES{G}_{\mathit{it}}+{\mathit{\beta }}_2{\mathit{Mediator}}_{\mathit{it}}+{\displaystyle {\sum }_{i=3}^{10}{\mathit{\beta }}_{i}Contro{l}_{it}+{\mathit{\phi }}_i+{\mathit{\mu }}_t+{\mathit{\epsilon }}_{it}}$$

(8)

Here, Mediator_it denotes the mediating variables, specifically financial constraints and green innovation, whereas the other variables follow the definitions used in the baseline model. Financial constraints are commonly measured by investment-cash flow sensitivity [82], the Kaplan–Zingales (KZ) index [83], the Whited–Wu (WW) index [70], and the Sah–Stiglitz (SA) index [84]. Since the SA index does not include variables with endogenous characteristics, we follow Glavas (2023) [85] and use the absolute value of the SA index as a proxy for firms’ financial constraints, where higher absolute values indicate greater constraints. In addition, following Wang et al. (2023) [86], corporate green innovation is measured by two widely adopted indicators in previous studies [87,88,89,90]: the number of applied green patents and the number of granted green patents. These indicators are from the CNRDS database, which classifies all patents into green or nongreen patents according to the Guide to the International Patent Classification (IPC) of the World Intellectual Property Organization.

Column (1) of

Table 5. Results of mechanism analysis.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	GTFP	SA	GTFP	GTFP	Patent1	GTFP	Patent2	GTFP
ESG	0.0013 *** (0.0004)	−0.0004 *** (0.0001)	0.0010 *** (0.0004)	0.0017 *** (0.0004)	0.1134 *** (0.0171)	0.0012 *** (0.0004)	0.0419 *** (0.0092)	0.0013 *** (0.0004)
SA			−0.4950 *** (0.0395)	−0.7441 *** (0.0543)
Interact				0.0001 *** (0.0000)
Patent1						0.0013 *** (0.0002)
Patent2								0.0014 *** (0.0003)
Control	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.3906 *** (0.2360)	3.5993 *** (0.0473)	4.1700 *** (0.2771)	4.7321 *** (0.3045)	−24.1113 ** (10.3677)	2.4213 *** (0.2357)	−12.2540 ** (5.5954)	2.4079 *** (0.2359)
N	16,767	16,767	16,767	14,384	16,767	16,767	16,767	16,767
r2	0.5533	0.9356	0.5577	0.5729	0.7504	0.5547	0.7560	0.5538
F	58.6720	64.5327	67.1807	73.9908	9.4455	57.7164	5.6167	54.9093
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Note: *** and ** indicate statistical significance at the 1% and 5% levels, respectively.

shows that ESG rating advantages have a notably favorable effect on GTFP (p < 0.01). Columns (2), (5), and (7) examine whether corporate ESG performance serves as a significant determinant of the two intermediary variables. All of the results are significant at the 1% confidence interval, indicating that high ethical performance mitigates corporate financial constraints and enhances green innovation. Overall, the treatment effects of ESG performance on the outcome (i.e., GTFP) and the mediators (i.e., financial constraints and green innovation) are significant, which fulfills the prerequisites for financial constraints and green innovation as mediators. Controlling for the independent variables in columns (3), (6), and (8), the mediating variable significantly affects enterprise sustainable development, with a slight decrease in the main effects of ESG performance. The coefficients of ESG in columns (3), (6), and (8) are lower than those in column (1). The above findings affirm H2 and H3.

The H2 results align with those of Bai et al. (2022) [91] and Zhang and Vigne (2021) [92]. Bai et al. (2022) [91] revealed that a firm’s ESG strengths mitigate financial constraints by sending positive signals to the market, whereas Zhang and Vigne (2021) [92] further highlighted the positive impact of reduced financial constraints on firm productivity. These findings can be explained by the fact that high ESG scores, which reflect a firm’s commitment to ethical standards, enhance market legitimacy, help firms gain stakeholder support and, in turn, better alleviate financial constraints. In fact, relaxing financial constraints incentivizes firms to pursue more green investments that might otherwise be foregone [93], ultimately contributing to higher GTFP. Furthermore, the results of H3 are supported by Wang et al. (2023) [86], who reported that ESG ratings significantly enhance the quantity and quality of corporate green innovation, and by Wu et al. (2022) [28], who reported that green innovation contributes to improved corporate sustainable growth. These findings can be attributed to the role of business ethics as a catalyst for green innovation within firms, which boosts green business performance and strengthens competitive advantage.

To further explore whether the mediating role of financial constraints varies with a firm’s reliance on external financing, we draw on the approach of Hsu et al. (2014) [94] and construct an indicator of external financing dependence for listed companies, denoted as Dependence. In Table 5, the estimated coefficient of the cross-multiplier term Interact in column (4) is significantly positive at the 1% level, indicating a stronger positive impact of alleviating financial constraints on firm sustainability as the firm’s external financing dependence increases. The possible reason is that firms with high levels of dependence have a greater need for external financing, and the alleviation of financial constraints is more sensitive to the enhancement effect of their sustainability. This finding aligns with that of Bağır and Seven (2022) [95], who found that, in Turkish, firms facing greater financial constraints exhibit a greater sensitivity of total factor productivity growth to debt growth.

4.6. Heterogeneous Effects

We perform a series of heterogeneity tests across groups to determine whether our results differ by region, industry, and ownership type. First, the sample is divided into three subgroups for testing: eastern, central, and western enterprises [96]. Columns (1)–(3) of

Table 6. Results of heterogeneity analysis.

Variable	(1) Eastern	(2) Central	(3) Western	(4) Heavy Polluting	(5) Non-Heavy Polluting	(6) SOEs	(7) Non-SOEs
	GTFP	GTFP	GTFP	GTFP	GTFP	GTFP	GTFP
ESG	0.0016 *** (0.0007)	0.0004 (0.0016)	0.0009 (0.0014)	0.0014 *** (0.0006)	0.0015 (0.00014)	0.0029 *** (0.0007)	−0.0001 (0.0009)
Control	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.5153 *** (0.3560)	1.5440 * (0.8217)	0.5145 (0.9183)	1.4685 *** (0.3378)	2.6746 *** (0.6765)	1.1513 *** (0.3634)	3.1484 *** (0.4997)
N	11,339	3067	2361	11,937	4830	7895	8872
r2	0.5053	0.4318	0.5101	0.4650	0.4984	0.4876	0.4672
F	51.9495	5.5383	13.5623	26.9605	25.3701	27.4523	32.2119
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Note: *** and * indicate statistical significance at the 1% and 10% levels, respectively.

reveal that the effect of ESGs on GTFP is significant only in the eastern region and not in the central or western region. This phenomenon could be attributed to the pivotal role of regional financial systems in facilitating or impeding corporate financing. The eastern region’s superior financial infrastructure and reduced information asymmetry create fertile ground for ESG initiatives to mitigate financial constraints effectively [97]. Additionally, firms in Eastern China, operating under stricter regulatory frameworks than their counterparts in Central China and Western China, are likely more motivated to upgrade their GTFP levels. This result differs from Zhang et al. (2024) [98], who found that the positive relationship between ESG and productivity is more pronounced among firms in the central region. A possible reason is that the study focuses on listed firms in the textile industry, which began shifting to the central region in 2006. The industry has higher investment and better facilities in the central region compared to the east and west, which amplifies the positive impact of ESG on firm productivity.

Second, we draw on Zhang et al. (2019) [99] to classify companies into heavily polluting and non-heavily polluting categories. As shown in columns (4)–(5) of Table 6, the effect of ESG ratings on GTFP is significant at the 1% level for heavily polluting companies but not significant for non-polluting counterparts. This disparity can be attributed to the heightened scrutiny and regulatory pressure faced by heavily polluting companies from the public, investors, and regulators due to their substantial environmental impact [100]. For example, according to the Administrative Measures on Statutory Disclosure of Enterprise Environmental Information [101], heavily polluting firms in China are mandated to report their pollutant and carbon emissions. Faced with prevailing ethical norms and severe environmental regulations, these firms are motivated to pursue green innovations and advance production technologies to achieve sustainable development goals. This rationale closely aligns with legitimacy theory, which posits that firms with high-environmental ethics are more likely to gain facilitated access to diverse resource pools held by key stakeholders [102]. This finding is consistent with Gu et al. (2025) [103], which argues that highly polluting firms are assigned greater ESG responsibilities, and the external networks formed during ESG practices facilitate access to innovation-related experience, thereby enhancing TFP.

Third, we categorize the samples into state-owned enterprises (SOEs) and nonstate-owned enterprises (non-SOEs) to examine the impact of ownership type. Columns (6)–(7) of Table 6 show a more pronounced influence of ESG ratings on GTFP in listed SOEs. There are two possible reasons. On the one hand, the government has been progressively tightening ESG regulatory standards for SOEs. For instance, the State-owned Assets Supervision and Administration Commission issued the Guiding Opinions on Promoting Information Disclosure of Centralized Enterprises, mandating stringent ESG disclosure requirements for SOEs [104]. This initiative underscores the ethical obligations of SOEs to set an example in the transition towards a greener economy. On the other hand, the career trajectories of SOE officials, including promotions and dismissals, are often contingent on their compliance with and implementation of central government policies. This places SOEs under substantial pressure to commit to high-quality ESG disclosures. In return, the government may reward SOEs through preferential resource allocation, subsidies, and tax incentives [105]. This support not only effectively alleviates the financial constraints of SOEs but also significantly reduces their innovation costs, thereby further enhancing their GTFP. This result is consistent with the findings of Deng et al. [18], who concluded that ESG has a more significant impact on the productivity of state-owned enterprises, while its effect is insignificant for non-state-owned enterprises.

5. Discussion and Conclusions

5.1. Conclusions

Amid increasing global emphasis on sustainable development, this study investigates whether and how firms’ ESG performance affects green total factor productivity (GTFP). Using panel data from Chinese A-share listed companies, we find that higher ESG ratings are significantly associated with increased GTFP. Mechanism analysis shows that this effect operates through two key pathways: alleviating financing constraints and promoting green innovation. Notably, the mediating role of financing constraints is more pronounced in firms with greater reliance on external capital, underscoring the importance of ESG performance in enhancing financial access for sustainability-related investments. Heterogeneity analysis reveals that the positive impact of ESG ratings on GTFP is more evident in firms located in eastern China, operating in pollution-intensive sectors, and with state ownership. These results suggest that institutional development, environmental exposure, and political alignment shape the effectiveness of ESG initiatives in driving green productivity. These findings contribute to the literature by introducing GTFP as a composite sustainability outcome that bridges environmental responsibility and production efficiency, constructing a dual-pathway framework to clarify the micro-level mechanisms behind ESG impacts, and highlighting the contextual variability of ESG effectiveness across industries, ownership types, and regions. Overall, the study provides empirical evidence that ESG engagement can serve as a strategic lever to promote green productivity, deepen sustainable development, and inform more targeted ESG evaluation and governance practices.

5.2. Policy Implications

This research reveals that ESG ratings have a clear and positive effect on corporate sustainability, as assessed through GTFP. According to the resource-based view, an outstanding corporate culture, characterized by its value, rarity, inimitability, and non-substitutability, can be considered a key resource for achieving sustainable competitive advantage [106]. Therefore, managers should cultivate ESG-oriented corporate cultures, aligning socially responsible behaviors with key ESG rating criteria to ensure differentiated advantages at the organizational level and to become leaders in sustainable development. For ESG rating agencies, we propose integrating GTFP into existing ESG evaluation frameworks to enhance sustainability assessments. For companies, we suggest detailed reporting on GTFP outcomes as part of annual sustainability reports. Companies should provide the calculation method, raw data, and assessment period for GTFP to allow investors and stakeholders to monitor the company’s sustainable progress.

By elucidating the pathways through which ESG ratings enhance GTFP—specifically through financial constraints and green innovation—the study provides a robust foundation for understanding how companies can achieve sustainable development goals. On one hand, to ease the financial constraints of enterprises, financial institutions should adopt a science-based approach to evaluate borrowing enterprises and pursue responsible investments guided by ESG principles. They should adopt an external oversight role to monitor fund usage by borrowing entities and prevent the misallocation of green financing for non-sustainable projects. Banks often value a firm’s immediate financial health over its long-term profitability when assessing loan repayment capabilities [107]. Without policy intervention, financial support tends to favor entities with stronger collateral and asset bases. Echoing Bağır and Seven (2022) [95], our findings suggest that if limited resources can be allocated to financially constrained firms, they are more likely to increase their green productivity, resulting in greater overall productivity gains from financial borrowing in an economy. Thus, implementing indirect government policies that incentivize private banks to extend credit to financially strapped businesses through guaranteed schemes, alongside direct support from public banks and financial institutions, is crucial for accelerating capital flow to companies with significant room for GTFP enhancement. On the other hand, to promote the green innovation of enterprises, the Chinese government could offer tax incentives, subsidies, and innovation fund support to enterprises committed to green innovation; this may include direct grants for the development of green technologies or tax benefits for companies that meet specific environmental standards. Moreover, the government could adopt green procurement policies, prioritizing goods and services produced via innovative green technologies and methods; this would not only directly support green businesses but also encourage more enterprises to invest in green innovation.

In addition, our study concludes that ESG ratings fail to increase the GTFP of firms in undeveloped regions, non-state enterprises and non-heavily polluting industries; this is mainly attributed to these firms having insufficient awareness, cognition, and motivation to carry out ESG practices, leading to inactive ESG behaviors and poor ESG performance. Specifically, first, in China’s less developed central and western regions, with their lower marketization and lax oversight [97], the urgency of economic development often overshadows the importance of ESG ratings for sustainable growth; this reflects a critical ethical dilemma of prioritizing short-term economic gains over long-term sustainability, highlighting the ethical duty of corporate decision-makers to integrate sustainability with economic growth. Second, the Chinese government’s encouraging yet nonmandatory approach to ESG disclosure creates a notable ethical divide between SOEs and non-SOEs, as former managers are more active in responding to the state’s call for political purposes. As stated in the Report on ESG Actions of Chinese Listed Companies (2022–2023), the ESG disclosure rate among centralized SOEs was 57.88% as of March 2023, far higher than the 24.47% rate observed among non-SOEs [108]. This disparity not only highlights the varying incentives for ESG disclosure between different types of enterprises but also underscores the importance of policies that ensure that all firms, regardless of ownership structure, are equally committed to ethical practices. Third, non-heavily polluting industries, which face minimal environmental shock, lack the drive to increase their environmental stewardship. Additionally, when investors incorporate ESG criteria into their investment strategies, there is a disproportionate focus on environmental aspects (E) over social (S) and governance (G) considerations [11]. Consequently, firms not categorized as heavy polluters also tend to neglect the S and G dimensions, leading to a general disinterest in comprehensive ESG engagement. This misalignment calls for a recalibration in the approach of investors and businesses, advocating for a holistic embrace of ethical practices that equally prioritize all dimensions of corporate responsibility.

To give full play to the role of ESG ratings in facilitating the GTFP development of Chinese companies across regions, industries, and attributes, it is imperative for the government to establish a unified set of ESG disclosure standards and enforce the stringent regulation of disclosed information. First, the absence of harmonized ESG performance metrics significantly hinders the effective utilization of ESG data. A promising development in this arena is the release of the first global ESG disclosure standards—IFRS S1 and IFRS S2—by the International Sustainability Standards Board (ISSB) in June 2023 [109]. These standards establish a unified framework for sustainability-related financial reporting, aiming to improve the consistency, comparability, and decision-usefulness of ESG information across jurisdictions. China has begun aligning with this global trend. In 2023, the ISSB established an office in Beijing [110], and Chinese regulatory authorities have initiated preliminary efforts to localize ESG disclosure practices. However, significant work remains to develop a coherent, nationally unified evaluation system. Future efforts should focus on standardizing ESG terminology, refining indicator definitions, and clarifying data boundaries to reduce ambiguity and discretionary interpretation. In particular, constructing a modular indicator framework that distinguishes core universal metrics from sector-specific extensions would help balance international alignment with local relevance. Such standardization is essential to ensure ESG data consistency across firms, enhance comparability over time, and lay the groundwork for credible ESG assessment, assurance, and market application.

Second, to address the issue of insufficient information required for ESG ratings, mainland China should consider implementing mandatory ESG disclosure. Globally, some countries have already enforced compulsory ESG reporting, obligating firms to provide high-quality ESG information either in conjunction with traditional financial disclosures or in specialized standalone reports [10]. A notable example is the European Union’s Corporate Sustainability Reporting Directive (CSRD), which came into force on January 5, 2023, replacing the earlier Non-Financial Reporting Directive (NFRD) [111]. The CSRD significantly expands the scope of sustainability reporting, mandating that approximately 50,000 companies disclose detailed information on environmental, social, and governance (ESG) factors [112]. These disclosures must adhere to the European Sustainability Reporting Standards (ESRS), developed by the European Financial Reporting Advisory Group (EFRAG), which emphasize a “double materiality” approach—requiring companies to report both on how sustainability issues affect their business and how their operations impact society and the environment [113]. The first set of ESRS was adopted by the European Commission on July 31, 2023, with reporting obligations starting from the 2024 financial year and disclosures to be published in 2025 [114]. Although the Hong Kong Stock Exchange mandates ESG disclosures, mainland China’s A-share market still implements voluntary reporting, resulting in more than 60% of A-share companies not publishing ESG reports by 2022 [108]. Therefore, it is imperative for the Chinese government to expand mandatory ESG disclosure requirements from Hong Kong to the entire nation.

Third, regulatory bodies should empower third-party auditors and financial market supervisors to verify ESG claims, investigate discrepancies, and impose penalties for misrepresentation. In the United States, the Securities and Exchange Commission (SEC) has intensified enforcement against ESG misrepresentation by establishing a Climate and ESG Task Force in 2021. Although the task force was later absorbed into broader operations, ESG-related enforcement actions have continued. Notably, in 2022, BNY Mellon was fined $1.5 million for overstating ESG review procedures, and Goldman Sachs Asset Management was fined $4 million for failing to follow internal ESG investment protocols [115,116]. The United Kingdom has adopted a parallel approach: the Financial Conduct Authority (FCA) implemented a new anti-greenwashing rule (ESG 4.3.1R) effective from May 2024, requiring that any sustainability claims made by regulated firms be fair, clear, and not misleading. The FCA’s accompanying guidance (FG24/3) outlines standards for substantiating ESG statements, including requirements for data transparency, completeness, and appropriate use of third-party information [117,118]. Similarly, in Australia, the Australian Securities and Investments Commission (ASIC) imposed a record AUD 12.9 million penalty on Vanguard Investments in 2024 for misleading claims about ESG exclusions, demonstrating a growing trend toward legal accountability in green finance [119]. Drawing on these international precedents, China could establish a comparable compliance framework that integrates ESG disclosures into its legal infrastructure, mandates third-party verification, enables public reporting channels, and imposes meaningful sanctions for misleading disclosures. Such a shift from voluntary reporting to enforceable compliance would enhance ESG data credibility and incentivize firms to engage more substantively with sustainable development goals.

5.3. Limitations and Further Research

This paper has several limitations. It is focused primarily on the ESG profiles of Chinese A-share companies listed on the Shanghai and Shenzhen stock exchanges, excluding firms listed in markets outside A-shares, such as Jingdong (JD. US; 09618. HK), Alibaba (BABA. US; 09988. HK), and Tencent (00700. HK), as well as unlisted entities, including major corporations such as Huawei, which boasted assets worth CNY 10,637.55 billion by the end of 2022 [120], and a significant number of micro-, small-, and medium-sized enterprises (MSMEs), which represent China’s largest and most vibrant business sector with over 52 million entities [121]. These diverse enterprise categories may face unique ESG challenges and have varying perceptions of ESG, consequently yielding diverse impacts on their GTFP. Thus, future studies should aim to encompass these companies, contingent on ESG assessment bodies that broaden their scope beyond China’s listed companies, to include a wider array of business types.

Moreover, the applicability of our findings to firms outside China is subject to further investigation. The maturity of ESG frameworks and disclosure mandates differs significantly across countries. Based on ESG disclosure practices, Singhania and Saini (2022) [122] categorized nations into four groups: those with well-established ESG frameworks (e.g., Norway, Sweden, and France), those with rapidly evolving ESG frameworks (e.g., Germany, Italy, and Australia), those with an ESG framework in the development phase (e.g., Singapore, India, and China), and those with nascent ESG frameworks (e.g., Russia, Indonesia, and Thailand). With respect to ESG disclosure norms, although regions such as Hong Kong (China), Singapore, and Vietnam have implemented mandatory disclosures, others such as mainland China have opted for semi-mandatory disclosures [123], and some, including Japan, Australia, and Thailand, follow voluntary disclosure practices. Thus, future research should broaden the sample scope and undertake multi-country analyses to explore how factors such as disclosure requirements and developmental stages influence the relationship between ESG ratings and GTFP.