Greening Corporate Environmental, Social, and Governance Performance: The Impact of China’s Carbon Emissions Trading Pilot Policy on Listed Companies

Abstract

With carbon emissions continuing to rise, global warming has become a popular research topic. To address climate change, China is taking proactive measures to reduce its carbon emissions. Covering the period between 2009 and 2021, this study utilizes data from 3668 publicly listed companies in China, along with data from their respective cities, to investigate the impact of the carbon emissions trading pilot policy on their environmental, social, and governance (ESG) performance. The conclusions show that the policy has greatly improved corporate ESG performance. However, its impact on the corporate ESG performance has varied over time, across different entities, and among different cities. Furthermore, the companies’ level of green innovation plays a crucial intermediary role. Additionally, a company’s risk-bearing capacity and level of urban green credit support positively moderate the effectiveness of the policy. These findings enrich our understanding of the relationship between the pilot policy and corporate ESG performance.

1. Introduction

Global carbon emissions and temperatures have been rising continuously over the past decade [1]. Global warming is leading to a redistribution of precipitation, permafrost, and rising sea levels, reducing biodiversity, decreasing crop yields, and damaging various infrastructures. These impacts seriously threaten human sustainable development. To cope with these global changes, curbing greenhouse gas emissions, mainly CO₂, has become a necessary goal of climate policies in many countries. As the largest greenhouse gas producer, China has taken active measures to promote sustainable development [2,3]. Referring to the results of the emissions trading system (ETS) in Japan and the European Union (EU), China launched a carbon emissions trading pilot policy (CETPP) in 2013 to give full play to the positive role of market mechanisms in promoting carbon emissions reduction. According to China’s Carbon Market: Reviews and Prospects (2022) published by the Beijing Institute of Technology, more than 2900 emission enterprises and units were included in the CETPP by the end of 2021. The cumulative volume of carbon emission allowances is about 8 billion tons, with a cumulative trading volume of 36,262,420 tons. Additionally, China encourages and guides enterprises to disclose their ESG information to promote voluntary emission reductions in businesses. Based on data from the Wind database, the number of ESG reports disclosed by listed Chinese companies increased from 597 in 2012 to 1513 in 2022. Specifically, the proportions of ESG reports disclosed by listed companies were 40.54% on the Shanghai Stock Exchange and 21.89% on the Shenzhen Stock Exchange. Meanwhile, more and more listed companies are planning to disclose their ESG information.

The carbon ETS is designed to reduce greenhouse gas emissions through a cap-and-trade mechanism. Regarding its specific arrangements, the government or another designated authority first determines the amount of greenhouse gas emissions allowed within a specified period and the allowance allocation method. Secondly, enterprises obtain allowances for free or through an auction, thus formulating the greenhouse gas emission cap for each enterprise within the specified period. Finally, if the greenhouse gas emissions of an enterprise exceed its corresponding allowance, it is required to purchase emission rights from enterprises with excess allowances. Conversely, enterprises with greenhouse gas emissions lower than their allowances are permitted to sell their allowances on the carbon market. However, in practice, the carbon ETS introduces new costs to enterprises despite the possibility that some enterprises gain revenue by selling their allowances. This increases the operational pressure on enterprises. For example, enterprises participating in the EU Carbon Emissions Trading Scheme generally claim that the scheme has increased the risks of reduced employment, competitiveness, and market shares [4].

Coping with the carbon ETS-induced cost pressures and maintaining competitiveness have become key challenges for enterprises under the carbon ETS. During exploration and practice, they found that improving their ESG performance was an effective approach. On the one hand, ESG performance emphasizes that enterprises should prioritize environmental issues during operation and incorporate carbon emission reductions into their goals, which is in line with the purpose of the carbon ETS. On the other hand, by virtue of the systematic interactions among ESG systems, ESG performance can not only help enhance the green total factor productivity (GTFP) to alleviate cost pressures, but it can also enable enterprises to build a reputation of pursuing sustainable development, appealing to their investors, consumers, and other stakeholders, and strengthening their competitiveness.

Can China’s CETPP improve the ESG performance of enterprises? If so, what is the mechanism of action? To address these two questions, the impact of the pilot policy on corporate ESG performance was assessed using a difference-in-difference (DID) approach, with the listed Chinese companies used as samples assigned to a treatment group (affected) or a reference group (not affected). Subsequently, since the CETPP may increase enterprises’ costs, the cost pressure may compel enterprises to adopt green technologies to reduce carbon emissions and thus, lower their production costs [5,6]. Hence, green innovation was established as a mediating variable between the policy and the corporate ESG performance, and the mediating effect was analyzed. On this basis, considering that enterprises with stronger risk tolerance or a higher level of support for the green credit policy in their cities are more likely to possess a greater ability to improve their ESG performance, the risk tolerance of enterprises and support for the green credit policy in their city were used as moderating variables to probe into the moderating effect between the policy and corporate ESG performance.

As one of the largest carbon emitting countries, China has demonstrated a sense of responsibility as a major developing nation [7]. In 2020, China introduced the goal of “striving to peak carbon emissions before 2030 and aiming for carbon neutrality before 2060” [8,9]. The achievement of this target inevitably depends on utilizing a market-oriented low-carbon policy. In fact, to facilitate the transition toward a low-carbon development pattern, the Chinese government issued a policy document in October 2011, officially approving Beijing, Shanghai, and other cities as the first batch. In addition, by the end of 2016, Fujian Province joined the program, marking a significant milestone in China’s adoption of a market-oriented low-carbon policy. These eight provinces and cities initiated carbon market operations in the order listed in Figure 1. In 2021, the Chinese carbon emissions market officially opened. Compared with the world’s largest EU-ETS, China’s carbon emissions trading pilot policy has significant differences in its trading mode, quota allocation, coverage, and price restrictions. For example, China’s CETPP uses protocol transfers and one-way bidding, while the EU-ETS conducts quota auctions through designated auctioneers. According to the administrative division, Shenzhen was considered part of Guangdong for research purposes.

This paper is composed of six parts: The first part is the introduction. The second part introduces studies related to the CETPP and corporate ESG performance. The third part explores the details of the related theoretical mechanisms and hypotheses. The fourth part describes the models, variables, and data sources. The fifth part presents the results. The sixth part presents a conclusion and policy recommendations.

2. Literature Review

2.1. Impacts of China’s Carbon Emissions Trading Pilot Policy

With the comprehensive initiation of China’s carbon emissions trading market, ex-post evaluations of the effects of the CETPP have become an important topic in the research on its role in addressing climate change. According to the literature, the impacts of the CETPP have been primarily assessed from the following aspects.

The purpose of implementing the policy is to investigate whether it reduces carbon emissions and emission intensity, so assessing the impacts of China’s CETPP on carbon emissions and emission intensity has become the primary focus of research. For example, Zhang et al. [10] and Shi et al. [11] concluded that such a policy has reduced carbon emissions. Wu et al. [12] argued that the policy has reduced carbon emissions and emission intensity. Zhang et al. [5] indicated that this policy has decreased the carbon emissions of industries under its coverage but has failed to mitigate the carbon intensity. Li and Wang [13] denoted that the policy has not only reduced the carbon emissions and its intensity in the pilot region, but it has also lowered carbon emissions in surrounding regions.

In China’s energy structure, the traditional energy source of coal comprises a large proportion, and such a situation is challenging to change in the short term. Therefore, the reduction in CO₂ emissions is necessary to improve energy efficiency, making the study of the impact of the CETPP on energy efficiency a crucial area of research. For instance, Hong et al. [14] discovered that China’s CETPP has improved the single- and total-factor energy efficiencies in urban areas. Tan et al. [15] also revealed that this policy can increase energy efficiency. In addition, Liu et al. [16] found that the policy has an impact on the energy structure and influences energy efficiency through technological innovation.

The updating or improving of the total factor productivity (TFP) has resulted in some scholars discussing the influences of China’s CETPP on innovation and changes in the TFP. For instance, Dong et al. [17] demonstrated that the policy can more likely achieve a Porter effect in the long run rather than in the short run. Chen et al. [18] discovered that this policy does not produce a weak Porter effect. Furthermore, Hu et al. [19] illustrated that this policy has had positive effects on innovation in enterprises, and Pan et al. [20] found that the policy has had a significant effect on the TFP of enterprises. Additionally, Wu and Wang [21] stated that the TFP of Chinese enterprises would increase by about 22.73% if they met the carbon emissions prices set in the EU.

Furthermore, the cost constraints imposed on enterprises by the CETPP have also altered the behaviors of enterprises and affected their financial performance, investment choices, market performance, and so on. Researchers are gradually conducting investigations in this area. For example, Yan et al. [22] demonstrated that the policy can improve financial performance while suppressing the market performance of enterprises. Yu et al. [23] concluded that this policy can enhance the financial performance of enterprises under its coverage. Qi et al. [24] found that the enterprises under the policy have expanded their local production layout and reduced cross-regional investment in production.

2.2. Influencing Factors of ESG

In recent years, research on corporate ESG performance has developed rapidly. According to statistics, the number of articles published on Web of Science with ESG as the topic increased from 94 in 2015 to 1410 in 2022 [25]. Among them, the number of papers related to ESG in China surged from 3 in 2015 to 117 in 2022. The exploration of factors influencing corporate ESG performance is one of the main research areas in the literature [26]. In summary, the research in this area mainly involves the following aspects:

Digital transformation reshapes the management modes and processes of enterprises, which is conducive to improving their ESG performance. For example, Wang et al. [27] stated that a digital strategy impacts the ameliorating ESG performance. Huang et al. [28] concluded that digital innovation exerts a positive effect on ESG performance. Moreover, Lu et al. [29], Wu and Li [30], Yang and Han [31], and Zhong et al. [32] indicated that corporate ESG performance can be enhanced through digital transformation, especially by strengthening internal control and environmental protection innovation.

Green financial policies help enterprises that are focused on green and low-carbon development to obtain credit support, thus efficiently alleviating their financing constraints. As revealed by Lei and Yu [33] and Xue et al. [34], green financial policies can increase corporate ESG performance. Additionally, Chen et al. [35] and Zheng et al. [36] revealed that the issuance of green bonds has also promoted increases in corporate ESG performance. In addition, Cao et al. [37] argued that improvements in corporate ESG performance through green investments will attract investors and clients with an awareness of environmental protection.

The governance mechanisms and structures of enterprises determine their ESG performance to a certain extent. For instance, Adeneye et al. [38] found that effective corporate governance mechanisms can improve ESG performance. The study conducted by Jiang et al. [39] showed that institutional investors play a crucial role in shaping corporate ESG performance. Alkurdi et al. [40] concluded that the attendance at board meetings has a positive impact on ESG performance. Gurol and Lagasio [41] revealed a significantly positive correlation between board size and corporate ESG performance. Velte [42] found that establishing a corporate social responsibility (CSR) committee can significantly contribute to ESG performance improvement.

Environmental regulations pose coercive pressure on enterprises, and such regulatory pressure is considered an important driving factor for sustainable environmental, societal, and economic development [43]. For instance, Wang et al. [44] found that China’s central environmental protection inspection has evidently improved corporate ESG performance. According to Baraibar-Diez et al. [45], sustainable compensation policies affect corporate ESG scores. In particular, the establishment of CSR committees may result in better nonfinancial performance. Additionally, Lu and Cheng [46] concluded that China’s Environmental Protection Law led to better ESG performance among state-owned enterprises. Moreover, He et al. [47] discovered that China’s Environmental Protection Tax Law has increased the ESG performance of heavily polluting enterprises.

2.3. Carbon Emissions Trading Pilot Policy and Corporate ESG Performance

Some studies have partially or indirectly reflected the impact of China’s CETPP on corporate ESG performance. Firstly, such a policy affects one or more aspects of corporate ESG performance, such as carbon emissions, energy efficiency, and corporate finance. Secondly, given the function of China’s CETPP as a market-based incentive for environmental regulation, it is important to investigate its influence on corporate ESG performance. This also involves researching the impacts of the implementation of the central environmental protection inspection system, Environmental Protection Law, and Environmental Protection Tax Law on corporate ESG performance.

Inspired by the existing studies, some scholars have started to explore the impact of China’s CETPP on corporate ESG performance more comprehensively. For example, Zhang et al. [48] found that China’s CETPP encouraged enterprises to increase their R&D investment and improve their internal controls, ultimately enhancing their ESG performance. Additionally, Tian et al. [49] discovered that the pilot policy can improve ESG performance and has had a more remarkable promotion effect on the ESG performance of non-state-owned enterprises.

To further enrich the research on the impact of China’s CETPP on corporate ESG performance, the marginal contributions of this study are as follows: Firstly, the influencing mechanism of the impact of China’s CETPP on corporate ESG performance is systematically proposed and elucidated. Secondly, the interference of urban heterogeneity in the research on the influence of China’s CETPP on corporate ESG performance is fully considered, using urban heterogeneity as the control in the model. Thirdly, the mediating effect of the policy on corporate ESG performance is re-tested, and the moderating effect of the policy on corporate ESG performance is investigated.

3. Theorized Mechanisms and Hypotheses

3.1. Direct Impact of Carbon Emissions Trading Pilot Policy on ESG Performance

The promotion of a company’s ESG performance through a carbon trading pilot policy primarily involves external pressures and internal drivers. On the one hand, external pressure driven by environmental concerns compels companies to elevate their focus on their ESG performance. On the other hand, the increasing cost of carbon emissions encourages companies to undergo green transformations and enhance their ESG performance. In terms of external pressures, companies that do not comply with their environmental responsibilities are more likely to face constraints from environmental protection policies. Moreover, they may face penalties in the financial markets. For instance, financial market investors tend to avoid companies with such issues, even if these companies have a strong financial performance [50]. Consequently, companies often seek to improve their public reputation by addressing their environmental protection and social governance issues.

When companies start focusing on such matters, they align more closely with national policies and are more likely to receive favor from both the government and the public. Even though this increased recognition can enhance a company’s visibility, it also magnifies any misconduct in their operations, significantly increasing both the internal and external pressures they face [51]. From an internal perspective, the carbon market has transformed carbon quotas into commodities. Purchasing carbon quotas to meet operational needs only provides a temporary solution, and the prices of carbon quotas fluctuate with demand. Therefore, from a long-term perspective, companies engage in green technological innovation to reduce their dependence on carbon quotas, thereby minimizing the uncertainty of their operating costs. This, in turn, facilitates the transformation of the company toward green practices and increases their ESG performance.

Hypothesis 1.

The carbon emissions trading pilot policy can enhance a company’s ESG performance.

3.2. Influence Mechanisms of the Carbon Emissions Trading Pilot Policy on ESG Performance

The CETPP encourages companies to engage in green innovation, resulting in an improvement in their ESG performance. The Porter hypothesis posits that appropriate environmental regulations can encourage companies to internalize external environmental costs, thereby motivating them to engage in technological innovation [52]. Companies often weigh the costs of procuring carbon allowances against the costs of developing low-carbon technologies. When the cost of purchasing carbon allowances exceeds the cost of low-carbon innovation, companies tend to choose the latter [53]. Conversely, they will purchase carbon allowances. Considering the fluctuations and instability of carbon prices, the risk associated with carbon emissions intensifies the compensation effect [54,55]. Consequently, to reduce the adverse impact of carbon risk on their ongoing operations, companies focus on and increase their investments in low-carbon technology innovation. The emphasis that companies place on green development helps promote the implementation of a series of measures related to carbon emissions reduction. Furthermore, it compels companies to pay more attention to environmental protection and improve their ESG performance. The potential influence mechanism of the policy on the corporate ESG performance is shown in Figure 2.

Hypothesis 2.

The carbon emissions trading pilot policy can incentivize companies to engage in green innovation, leading to an improvement in their ESG performance.

The impact of the CETPP on a corporation’s ESG performance is often subject to the moderating influence of the company’s risk-bearing capacity and green credit support level. On the one hand, to maximize their profits, companies reallocate more funds to non-profit ESG-related projects to adapt to this policy environment. Resource allocation is crucial for a company’s operations. Achieving high ESG performance requires flexible adjustments to their business strategies, but the risks associated with such changes can be detrimental to the company. This is, to some extent, a matter of the company’s long-term strategic decisions rather than short-term adjustments [56]. Therefore, a company’s risk-bearing capacity is crucial, as it helps accelerate the accumulation of capital and the level of investment in innovation [57]. This, in turn, helps companies overcome the challenges of green innovation and lower their carbon emissions, in line with this policy. With a higher risk-bearing capacity, companies can effectively adjust to enhance their ESG performance.

On the other hand, the increasing level of support for green credit provides more robust backing for companies’ green transformation efforts [58]. Strengthening the level of urban green credit support can optimize the credit structure between polluting industries and environmental protection industries, reducing financial difficulties for companies seeking to enhance their green innovation capabilities. This makes it easier for the goals of the carbon emissions trading pilot policy to be translated into practical actions at the company level, facilitating further optimization of production and laying a solid foundation for improving ESG performance.

Hypothesis 3.

A company’s risk-bearing capacity has a positive moderating effect on the ability of the carbon emissions trading pilot policy to improve the company’s ESG performance.

Hypothesis 4.

Urban green credit support has a positive moderating effect on the ability of the carbon emissions trading pilot policy to improve a company’s ESG performance.

4. Research Design

4.1. Variables

4.1.1. Explained Variable

The explained variable in this study was ESG performance. The ESG data on companies were sourced from the Huazheng ESG Rating Index within the Wind database. To mitigate the influence of different rating systems, this study also used alternative rating indices for robustness analysis.

4.1.2. Core Explanatory Variable

The core explanatory variable was a dummy variable (Carbonpilot) based on the carbon trading pilot policy. When a city was chosen as a pilot city in year t, the value was 1 for years t and beyond; otherwise, it was 0. A total of 3668 listed firms in 249 cities were included. Given the policy regulations by the Chinese government stipulating that a company’s registered location must align with its operational location, this study matched the details of the registered locations obtained from the company registration information system with the cities where the policy has been implemented. This process allowed us to derive a company-level variable indicating the implementation of the pilot policy. The matching principles for the other variables followed the same criteria. Among the firms, 2903 made up the treatment group, and the other 765 firms comprised the control group. Additionally, considering that the pilot policy is typically implemented at the end of the year, this study set its implementation time at the beginning of the following year when constructing the dummy variables.

4.1.3. Mediating and Moderating Variables

In this study, we chose the number of green patent applications (Greenpatent), measured as the number of annual green patents granted to public companies, as a mediating variable. The company’s risk-bearing capacity (ROI) and support for green credit (Greenloan) were selected as the moderating variables. The ROI was calculated as the adjusted asset return rate based on the industry and year medians, while the Greenloan was calculated as the ratio of the total credit for environmental protection projects to the total credit.

4.1.4. Control Variables

This study incorporated control variables from both the city [59] and company levels [60,61]. At the company level, the control variables included the following: company age (Lnage), measured as the total number of years since the company’s establishment (in natural logarithm); ownership concentration (Concentration), measured as the ownership percentage of the largest shareholder; asset-to-liability ratio (Alratio), measured as the ratio of company assets to liabilities; return on investment (ROI), measured as the ratio of company investment returns to its investment amount; current ratio (Crlratio), measured as the ratio of operating funds to borrowed funds; and the Herfindahl–Hirschman Index (HHI) for industry concentration, calculated as the square of the ratio of the company’s revenue in its primary business area to the total revenue of its industry. At the city level, the control variables included the following: industrial advancement (Advancement), measured as the ratio of annual output in the third industry to the combined output of the first and second industries in the city; GDP (LnGDP), measured as the GDP level in the base year of 2009, adjusted for inflation (as a natural logarithm).

4.2. Data Resource and Descriptive Statistics

The ESG performance data were from the Wind database. Company-level data were derived from the annual report of A-share listed companies, excluding those categorized as ST (Special Treatment), PT (Special Treatment for Delisting), insurance companies, and those with missing key variables. This study also excluded samples for which the company’s location was not clear. The time period of this sample was 2009–2021. The city-level data were from statistical yearbooks of various cities. All GDP-related data were adjusted to 2009 as the base year. In addition, the interpolation method was applied when missing values occurred.

Table 1. Descriptive statistics.

Abbreviation	N	Mean	SD	Min	Max
E	24,935	60.374	7.771	29.460	95.160
S	24,935	73.867	10.855	0	100
G	24,935	78.514	7.809	24.330	97.330
ESG	24,935	72.854	5.685	36.620	91.460
Carbonpilot	24,935	0.309	0.462	0	1
Lnage	24,935	2.241	0.805	0	3.466
Concentration	24,935	34.081	15.275	0.286	89.991
Alratio	24,935	0.497	0.579	0.014	55.409
ROI	24,935	0.511	2.349	−1.284	19.792
Crlratio	24,935	15.256	64.916	−3.251	529.156
HHI	24,935	0.147	0.143	0.032	1
Advancement	24,935	1.497	1.139	0.001	5.199
LnGDP	24,935	8.780	1.063	4.796	10.466

presents the descriptive statistics.

4.3. Econometric Model

4.3.1. Difference-in-Difference (DID) Approach

This study used the CETPP as the basis and employed a multiple period DID model to examine the impact of the carbon pilot policy on the listed companies’ ESG performance. The baseline regression model is as follows:

$${ESG}_{itc}={\beta }_0+{\beta }_1{Carbonpilot}_{itc}+\lambda {Control}_{itc}+v_i+{\mu }_t+{\omega }_c+{\epsilon }_{itc}$$

(1)

In Equation (1), ${ESG}_{itc}$ represents a company’s ESG performance, and ${Carbonpilot}_{itc}$ is a dummy variable. If company i joins the policy in year t, the value is 1; otherwise, it is 0. ${Control}_{itc}$ represents the control variables. Here, i, t, and c represent the company, time, and city where the company is located, respectively. ${\beta }_0$ is the constant term, and ${\beta }_1$ is the coefficient of the policy effect in this study. If the pilot policy promotes an improvement in company ESG performance, ${\beta }_1$ should be significantly positive. ${\epsilon }_{itc}$ is the error term. Furthermore, ${\nu }_i$, ${\mu }_t, and$ ${\omega }_c$ represent the individual, time, and city fixed effects, respectively.

4.3.2. Parallel Trend Test

The premise for using the DID approach is that, prior to the policy, there were no significant differences between the groups. In other words, without the policy intervention, the trends in the ESG performance of the companies were parallel. A significant difference between the groups after implementing the policy suggests that the difference might be attributed to the policy. To verify whether the treatment group and the control group had similar trends before the pilot policy, this study adopted the event study method as follows:

$${ESG}_{itc}={\sum }_{k=-6}^{k=8}{\beta }_k({Carbonpilot}_i\times {Post}_{t-k})+\lambda {Control}_{itc}+{\upsilon }_i+{\mu }_t+{\omega }_c+{\epsilon }_{itc}$$

(2)

In Equation (2), ${Post}_{t-k}$ is a series of dummy variables representing the pre-pilot periods, pilot year, and post-pilot period. k represents the cumulative number of years that company i has been involved in the policy. ${\beta }_k$ represents the estimated coefficient.

4.3.3. Mediating Effect Model

This study used the level of green innovation of a company as a mediating variable and conducted analysis through a stepwise regression approach. The following models were used for the mediating effects: Equation (3) is the same as Equation (1) and was used to examine the direct impact of the policy. Equation (4) tests how the pilot policy affects companies’ levels of green innovation. Equation (5) incorporates company ESG performance, the Carbonpilot dummy variable, and the companies’ levels of green innovation to determine whether the green innovation of a company acts as a mediator of the impact.

$${ESG}_{itc}={\beta }_0+{\beta }_1{Carbonpilot}_{itc}+\lambda {Control}_{itc}+v_i+{\mu }_t+{\omega }_c+{\epsilon }_{itc}$$

(3)

$${Greenpatent}_{itc}={\gamma }_0+{\gamma }_1{Carbonpilot}_{itc}+\lambda {Control}_{itc}+v_i+{\mu }_t+{\omega }_c+{\epsilon }_{itc}$$

(4)

$${ESG}_{itc}={\phi }_0+{\phi }_1{Carbonpilot}_{itc}+{\phi }_2{Greenpatent}_{itc}+\lambda {Control}_{itc}+v_i+{\mu }_i+{\omega }_c+{\epsilon }_{itc}$$

(5)

In Equations (4) and (5), ${Greenpatent}_{itc}$ represents a company’s green innovation level, which served as the mediating variable. ${\gamma }_0$ and ${\phi }_0$ are constant terms, and ${\gamma }_1$, ${\phi }_1$, and ${\phi }_2$ are coefficients of the related variables.

4.3.4. Moderating Effect Models

According to the theoretical analysis, the impact of the policy on a company’s ESG performance is subject to moderation by its risk-bearing capacity and level of urban green credit support. Therefore, this study constructed Equation (6) to test whether moderation effects exist.

$${ESG}_{itc}={\rho }_0+{\rho }_1{Carbonpilot}_{itc}+{\rho }_2{Mod}_{itc}+{\rho }_3({Carbonpilot}_{itc}\times {Mod}_{itc})+\lambda {Control}_{itc}+v_i+{\mu }_i+{\omega }_c+{\epsilon }_{itc}$$

(6)

In Equation (6), ${Mod}_{itc}$ represents the moderator variable. In this study, a company’s risk-bearing capacity and level of urban green credit support were considered moderator variables. ${Carbonpilot}_{itc}\times {Mod}_{itc}$ is the interaction term, which was obtained by centralizing the product of ${Carbonpilot}_{itc}$ and ${Mod}_{itc}$. ${\rho }_0$ is the constant term, while ${\rho }_1 ,{\rho }_2$, and ${\rho }_3$ are the related regression coefficients.

5. Empirical Analyses

5.1. Results of Baseline Regression

Table 2. Benchmark regression results.

Variables	(1) E	(2) S	(3) G	(4) ESG
Carbonpilot	0.093	−0.186	0.975 ***	0.443 ***
	(0.66)	(−0.84)	(5.42)	(3.67)
Lnage	−0.209	1.503 ***	−2.952 ***	−0.982 ***
	(−1.52)	(6.94)	(−16.88)	(−8.35)
Concentration	−0.015 ***	0.004	0.047 ***	0.021 ***
	(−2.78)	(0.49)	(7.14)	(4.72)
Alratio	−0.199 **	−0.434 ***	−0.948 ***	−0.599 ***
	(−2.22)	(−3.08)	(−8.34)	(−7.85)
ROI	−0.000	−0.014	−0.126 ***	−0.062 ***
	(−0.02)	(−0.61)	(−6.97)	(−5.09)
Ctlratio	0.000	−0.001	0.003 ***	0.001 **
	(0.52)	(−1.32)	(4.07)	(2.25)
HHI	−2.254 ***	−1.700 ***	−0.812 *	−1.415 ***
	(−5.85)	(−2.81)	(−1.66)	(−4.31)
Advancement	0.269 ***	−0.110	−0.317 ***	−0.102
	(2.78)	(−0.72)	(−2.58)	(−1.24)
LnGDP	−0.603	−2.749 ***	−1.010	−1.340 **
	(−0.93)	(−2.69)	(−1.23)	(−2.42)
Year FE	Yes	Yes	Yes	Yes
Id FE	Yes	Yes	Yes	Yes
City FE	Yes	Yes	Yes	Yes
Constant	66.619 ***	95.096 ***	93.181 ***	86.615 ***
	(11.65)	(10.59)	(12.85)	(17.78)
Obs.	24,420	24,420	24,420	24,420
R-squared	0.702	0.626	0.528	0.597

Note: T-statistics are in parentheses. ***, **, and * indicate the 1%, 5%, and 10% significance levels, respectively (the same below). Due to limited page space, the control variables are not displayed in the subsequent tables.

presents the results of baseline regression. Columns (1) to (3) present the effects of the policy on three aspects of corporate ESG performance: environmental protection, social responsibility, and corporate governance, respectively. Column (4) displays the impact of the policy on companies’ ESG performance. According to columns (1) to (3), the policy had positive and negative impacts on environmental protection and social responsibility, but these effects were not significant. However, corporate governance was significantly influenced by the implementation of the policy. This may be attributed to the significant influence of the policy on a company’s profitability and profit distribution, compelling management to optimize corporate governance to adapt to the market-oriented carbon trading model. According to column (4), the coefficient of the policy on a company’s ESG performance was positive and significant, indicating that the policy enhanced the overall ESG level of companies. This suggests that the pilot policy draws attention to green and sustainable development for local governments and market entities. Local governments can tailor their environmental protection policies and green production standards to align with national policy direction. Additionally, the establishment of a carbon trading system forces companies to include carbon emissions in their considerations, motivating them to improve production technologies and processes to reduce their carbon emissions [62,63] and boost their environmental awareness and social responsibility [64,65].

5.2. Parallel Trend Test

As per the previous analysis, the DID method could be employed only when the treatment and control groups exhibited a parallel trend before implementing the policy [66,67]. Therefore, this study conducted a parallel trend test to demonstrate the feasibility of the DID model, and the results are presented in Figure 3. It can be observed that the variable was insignificant before the policy. This shows that there was no significant difference in the ESG performance between the companies before the policy was implemented, and the sample satisfied the parallel trend.

5.3. Robustness Test

5.3.1. Other ESG Rating Systems

To reduce the possibility of coincidental results due to differences between the Huazheng ESG rating and other ESG rating systems, this study selected 1209 listed companies with ESG ratings from both Huazheng and Bloomberg for the years 2011–2020. By comparing the results in

Table 3. Test results of other ESG rating systems.

Variables	Huazheng ESG Rating System		Bloomberg ESG Rating System
	(1)	(2)	(3)	(4)
Carbonpilot	0.728 ***	0.773 ***	0.628 ***	0.677 ***
	(3.78)	(4.01)	(3.55)	(3.82)
Control		Yes		Yes
Year FE	Yes	Yes	Yes	Yes
Id FE	Yes	Yes	Yes	Yes
City FE	Yes	Yes	Yes	Yes
Constant	74.300 ***	90.756 ***	20.956 ***	20.738 **
	(962.02)	(9.57)	(295.61)	(2.38)
Obs.	8166	8166	8166	8166
R-squared	0.593	0.597	0.796	0.798

Note: T-statistics are in parentheses. *** and **, indicate the 1% and 5%, significance levels, respectively.

, it is evident that even when replacing the dependent variable, the findings remain consistent with those in the earlier sections. In other words, the conclusion that the CETPP influences a company’s ESG performance remains robust.

5.3.2. PSM-DID Estimation

The CETPP encompasses eight provinces and cities. However, the selection of these regions may be influenced by many factors [68]. To reduce potential selection bias, the PSM-DID approach was employed to retest the baseline regression, as shown in

Table 4. PSM-DID estimation results.

Variables	ESG
	(1)	(2)
Carbonpilot	0.462 ***	0.502 ***
	(3.73)	(4.08)
Control		Yes
Year FE	Yes	Yes
Id FE	Yes	Yes
City FE	Yes	Yes
Constant	72.648 ***	87.515 ***
	(1581.58)	(17.94)
Obs.	24,127	24,127
R-squared	0.594	0.603

Note: T-statistics are in parentheses. *** indicate the 1% significance levels, respectively.

. First, through 1:1 nearest neighbor matching and replacement sampling, the treatment group was matched with the control group. P-values for differences in the means of variables between the groups after re-matching were not significant. Then, based on the PSM method, regression was performed on the matched samples. The coefficient of the Carbonpilot after matching was significant, consistent with the results in the previous sections and demonstrating that the results of the impact of the CETPP on a company’s ESG performance are robust.

5.3.3. Placebo Test

A placebo test was conducted to check whether the baseline regression results could be attributed to omitted variables [69]. Firstly, this study randomly selected 1623 cities from the entire sample of listed companies as the pseudo-treatment group, assuming that there were companies affected by the CETPP. The distributions of the estimated coefficients and related p-values were plotted after this procedure was carried out 500 times. As shown in Figure 4, a dashed vertical line on the x-axis represents the true coefficient. It is evident that the coefficients obtained from random matching differed significantly from the true value, with the observed density plot centered around 0. Therefore, the placebo test was passed.

5.3.4. Other Robustness Tests

Table 5. Results of other robustness tests.

Variables	Outliers	Bidirectional Causality		Sample Period (Excluding 2018)
	(1)	(2)	(3)	(4)
Carbonpilot	0.533 ***			0.235 *
	(4.60)			(1.83)
L. Carbonpilot		0.619 ***
		(4.80)
L2. Carbonpilot			0.743 ***
			(5.49)
Control	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Id FE	Yes	Yes	Yes	Yes
City FE	Yes	Yes	Yes	Yes
Constant	87.022 ***	83.590 ***	76.443 ***	82.318 ***
	(20.00)	(14.45)	(11.08)	(13.82)
Obs.	24,420	19,938	17,010	14,078
R-squared	0.608	0.627	0.651	0.629

Note: T-statistics are in parentheses. *** and * indicate the 1%, and 10% significance levels, respectively (the same below).

presents the results of other robustness tests. First, to mitigate the disturbance caused by outliers, this study applied a trimming method to the sample data by excluding the dummy variables and restricting the data to the lower and upper 1% values [70]. This was followed by parameter estimation, as shown in column (1) of Table 5. Second, this study employed first-order and second-order lagged variables of the independent variables to conduct tests, replacing the original variables, to address potential issues of bidirectional causality [71], as indicated in columns (2) and (3). Third, to mitigate the potential influence of other policies [67], this study adopted a method of shortening the time window. Considering that the Chinese government began implementing the environmental protection tax in early 2018, and to avoid the impact of this policy, this study excluded the years 2018 and beyond and reexamined the results. The results are shown in column (4) and remain consistent with the previous analysis. In summary, the conclusions are robust.

5.4. Heterogeneity Test

5.4.1. Temporal and Spatial Heterogeneity

Figure 3 illustrates the evolving impact of the CETPP on ESG performance over time. As depicted in Figure 3, the pilot policy demonstrated a time-lagged effect. As time progressed, its influence on corporate ESG performance increased. The impact of environmental regulatory policies on a company’s ESG ratings was gradual. During the initial implementation of the pilot policy, it may increase a company’s production and operational costs, requiring the allocation of funds from their production and operations toward ESG investments [72,73]. After a period of green technological innovation, companies improve their level of green production, optimize management processes, and increase their focus on social governance [74,75].

According to the geographical classification by the National Bureau of Statistics of China, this study divided the sample into three regions and conducted separate regressions to examine the spatial heterogeneity. Columns (1), (2), and (3) of

Table 6. Spatial heterogeneity test results.

Variables	Eastern	Central	Western
	(1)	(2)	(3)
Carbonpilot	0.510 ***	−0.052	−0.871 *
	(3.61)	(−0.14)	(−1.71)
Control	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Id FE	Yes	Yes	Yes
City FE	Yes	Yes	Yes
Constant	89.733 ***	65.290 ***	40.422 ***
	(14.30)	(4.51)	(3.14)
Obs.	17,732	3479	3188
R-squared	0.610	0.555	0.579

Note: T-statistics are in parentheses. *** and * indicate the 1% and 10% significance levels, respectively.

represent the results for the Eastern, Central, and Western regions, respectively. The results indicate that the promotion effect of the policy on ESG performance was most significant in Eastern China, while it was not significant in Central China. In Western China, the pilot policy had a suppressive effect on ESG performance. This could be attributed to the higher economic development level in Eastern China, where companies have more advanced low-carbon and clean production technology. In contrast, Central and Western China are less economically developed, with less green R&D investment, making it more difficult for the inland regions to undergo a green transformation. Therefore, following the implementation of the pilot policy, these companies must allocate some of their funds originally designated for ESG investments to purchase carbon emissions allowances to maintain their normal production. Consequently, the impact of the pilot policy is less favorable.

5.4.2. Ownership and Industry Heterogeneity

To explore the impact of ownership heterogeneity, this study conducted a consistency check on the sample based on whether the companies were state-owned. Columns (1) and (2) in

Table 7. Ownership and industry heterogeneity test results.

Variables	Ownership		Industry Types
	State-Owned	Non-State-Owned	Real Estate	Utility	Industrial	Financial	Commercial	Conglomerates
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Carbonpilot	0.924 ***	−0.268	0.543	0.893 ***	0.387 **	0.671	−0.168	0.394
	(5.96)	(−1.41)	(1.48)	(2.84)	(2.54)	(0.40)	(−0.40)	(0.39)
Control	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Id FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
City FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	82.216 ***	70.683 ***	143.057 ***	69.541 ***	82.639 ***	174.335 **	102.194 ***	15.284
	(12.66)	(9.67)	(7.61)	(5.21)	(14.66)	(2.19)	(5.14)	(0.28)
Obs.	10,262	14,048	1784	4152	16,193	266	1456	564
R-squared	0.620	0.607	0.633	0.624	0.594	0.562	0.530	0.495

Note: T-statistics are in parentheses. *** and ** indicate the 1% and 5%, significance levels, respectively.

display the results for the state-owned and non-state-owned companies, respectively. The results indicate that the ESG performance of state-owned enterprises was significantly affected by the pilot policy. In contrast, non-state-owned enterprises did not show significant effects. This is because, in accordance with national policies (for example, environmental and sustainability policies), state-owned enterprises are required to participate in ESG practices [58], and they often have higher ESG demands compared to other types of enterprises [76]. Therefore, when influenced by the pilot policy, the ESG performance of state-owned enterprises is more likely to improve.

When considering the influence of different industries, this study conducted another heterogeneity test based on industry types, as shown in columns (3) to (8) of Table 7. Specifically, columns (3) to (8) represent the regression results of real estate enterprises, utility enterprises, industrial enterprises, financial enterprises, commercial enterprises, and conglomerates, respectively. These results indicate that utility enterprises and industrial enterprises are more likely to be affected by the CETPP in terms of ESG performance. For utility enterprises, the proportion of state-owned enterprises is relatively high, and they are more responsive to government environmental policies than non-state-owned enterprises. Additionally, some public utility enterprises are involved in the power industry, which is covered by China’s CETPP, making them more likely to improve their ESG performance. For industrial enterprises, China’s CETPP covers multiple heavy chemical industries, imposing direct constraints on the enterprises in these industries. In addition, Industry 5.0 establishes a link between business performance and carbon emissions [77]. Industry 5.0, which is characterized by green technology and intelligence, not only reduces operating costs, and improves product competitiveness but also encourages enterprises to participate more in green innovation. As a result, they are more likely to optimize their ESG performance [78,79].

5.5. Results of the Mechanism Influence

5.5.1. Mediating Effect of Green Innovation Level

Based on Equations (3) to (5), this study utilized a stepwise regression method to investigate the mediating effect of a company’s green innovation level. In

Table 8. Mechanism influence test results.

Variables	Greenpatent	ESG
	(1)	(2)	(3)	(4)
Carbonpilot	1.215 ***	0.432 ***	0.451 ***	0.357 ***
	(3.45)	(3.58)	(3.74)	(2.80)
Greenpatent		0.009 ***
		(3.66)
ROA			3.352 ***
			(12.56)
Carbonpilot × ROA			1.373 **
			(2.52)
Greenloan				−0.105
				(−0.05)
Carbonpilot × Greenloan				7.875 **
				(2.06)
Control	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Id FE	Yes	Yes	Yes	Yes
City FE	Yes	Yes	Yes	Yes
Constant	−41.413 ***	86.972 ***	86.873 ***	84.540 ***
	(−2.92)	(17.85)	(17.90)	(16.99)
Obs.	24,420	24,420	24,420	24,420
R-squared	0.623	0.598	0.601	0.598

Note: T-statistics are in parentheses. *** and ** indicate the 1% and 5%, significance levels, respectively.

, column (1) presents the effect of the CETPP on the green innovation level, while column (2) presents the impact of the policy and a company’s green innovation level on its ESG performance. The coefficients of the Carbonpilot and green innovation levels in both columns passed significance tests at the 1% level, indicating that the green innovation level acts as a mediator in the pathway through which the policy affects a company’s ESG performance. To further confirm the validity of this mediating effect, this study used Sobel and Bootstrap tests, the results of which suggest that the green innovation level is an effective mediating variable. Therefore, Hypothesis 2 is supported.

The policy provides companies with clear market price information and innovation directions for technology, creating incentives for low-carbon technological innovation within enterprises [80]. This can reduce the risks that companies face and enhance the certainty of the value obtained when investing in carbon emissions reduction. Additionally, the technological innovation of a company is affected by various cost factors. The pilot policy raises the relative price of carbon emissions allowances, and technological innovation can significantly reduce the long-term costs of carbon reduction [81]. As a result, companies prefer to invest in R&D for low-carbon technologies to lower their carbon reduction costs. Additionally, technological innovation contributes to reducing the environmental burden on stakeholders and decreasing the emissions of other pollutants, ultimately leading to higher ESG performance of companies [82,83].

5.5.2. Moderating Effects of Risk-Bearing Capacity and Urban Green Credit Support Level

This study examined the moderating effects of companies’ risk-bearing capacity and urban green credit support level, with the results shown in columns (3) and (4) of Table 8. The coefficients of the interaction terms are significant at the 5% level, indicating that a company’s risk-bearing capacity and urban green credit support level have positive moderating effects on the impact of the CETPP on the company’s ESG performance. The moderating effect values were 1.373% and 7.875%, supporting Hypotheses 3 and 4.

For companies with a lower risk-bearing capacity, the implementation of the policy forces them to engage in green innovation to reduce carbon emissions, leading to additional risks and crowding-out effects. This reduces the companies’ operational efficiency and hinders their further investment in ESG. However, when a company’s risk-bearing capacity reaches a certain level, it absorbs the uncertainty risks associated with its investment in ESG. In this scenario, the risks of green innovation are reduced, and companies are more proactive in allocating resources to reduce their carbon emissions. This fosters a sustainable development mindset and a social governance perspective, making the impact of the policy on companies’ ESG performance more pronounced. Furthermore, the development of green credit reduces the costs of green innovation, effectively mitigating the financial burden of ESG investment. This enhances companies’ commitment to addressing social and environmental concerns for green innovation, leading to their improved ESG performance.

6. Conclusions and Limitations

6.1. Conclusions

This study investigated the impact of China’s carbon emissions trading pilot policy on corporate ESG performance using data from 3668 listed companies and their respective cities from 2009 to 2021. Furthermore, this study employed intermediate and moderation effect models to provide a detailed analysis of the influencing mechanisms and pathways. In this study, the pilot policy was found to have a favorable influence on corporate ESG performance. This effect remained consistent across several rigorous tests. Notably, the policy primarily enhanced corporate governance standards. When a heterogeneity analysis was conducted, it became evident that the policy’s effectiveness intensified two years after its implementation, with a more significant impact observed in the eastern regions of China. Furthermore, state-owned enterprises, utilities, and industrial companies were more prone to experiencing positive effects on their ESG performance due to the policy. The pilot policy also encourages green innovation among companies, which, in turn, boosts their ESG performance. It is worth noting that the listed companies with stronger risk-bearing capacities and those located in areas with higher levels of urban green credit support experienced greater improvements in their ESG ratings due to the policy. Based on these findings, government departments should consider expanding the carbon market, fostering green innovation, and crafting targeted policies that account for the distinct characteristics of enterprises across different ownership structures and regions.

6.2. Limitations

Although this study enriches the research on the effects of China’s CETPP on corporate ESG performance and draws some valuable conclusions, there are still three limitations that warrant further exploitation. First, the impact of China’s CETPP on the ESG performance of the enterprises under its coverage was not investigated. Second, there are other external factors that influence ESG performance beyond China’s carbon emissions trading pilot policy, such as China’s “dual control of energy consumption” policy. However, due to the availability of data and the limitations of research topics, this study did not expand on these other factors, and further exploration is needed in the future. Third, there may be other mediating or moderating variables that have yet to be discovered.