Impact and Mechanism Analysis of Environmental Protection Fee and Tax Reform on the ESG Performance of Heavy Polluting Enterprises

Abstract

Recently, China has advocated for the comprehensive implementation of the new development concept and the enhancement of the national governance system and capacity, particularly in the area of ecology and environmental management. Environmental fee and tax reform has improved China’s modern environmental governance system and deepened the concept of sustainable development of enterprises. In the background of China’s strong call for green transformation and sustainable development, enterprises, as micro subjects in the operation of the market economy, are obliged to balance the concepts of business operation and sustainable development, and to practice and implement the ESG concept. Using a sample of A-share listed companies in China from 2014 to 2022, we conducted an in-depth analysis of the impact of environmental protection tax reform on corporate ESG performance through the difference-in-differences (DID) empirical approach. The results show that (1) the environmental protection fee and tax reform enhances the ESG performance of heavy polluters, and the impact coefficient is around 1.7 to 2.0. The reform exerts the strongest stimulatory effect on the environmental impact (E), with the stimulatory effect being about five times that of the aspect of social responsibility (S). (2) The environmental protection fee and tax reform enhances the ESG performance of heavily polluting firms by promoting green transformation of firms, investor attention, and the government’s focus on the environment. The impact mechanism passes a series of robustness tests, such as the parallel trend test, placebo test, and exclusion of other policy interferences. (3) The environmental protection fee and tax reform enhances the ESG performance of government-owned heavy polluting firms more than private firms. Among different regions, the ESG performance of enterprises in the central region has witnessed the largest improvement margin, while that of enterprises in the western region has the smallest improvement margin.

1. Introduction

The report of the 20th Chinese People’s Congress (CPC) has emphasized that China should speed up the green transformation of its development mode, enhance the prevention and control of environmental pollution, actively promote carbon peaking and carbon neutrality goals, and improve the modern environmental governance system. The development of ESG is one of the important solutions for pursuing sustainable development goals. It requires organizations to consider environmental impacts, social responsibilities, and corporate governance, so as to achieve green, sustainable, and high-quality development [1]. Since its introduction, ESG has received a rapid response on a global scale, which has prompted organizations to gradually shift from serving shareholders to serving stakeholders and even the whole society. In the context of China’s strong advocacy of green transformation and sustainable development, enterprises, as micro subjects in the operation of the market economy, practicing and implementing ESG concepts in their production and operation is deeply in line with China’s economic green development plan [2]. The Chinese government is also constantly introducing policies to promote green development. China’s first green tax law, the Environmental Protection Tax Law, came into effect on 1 January 2018. This marked the prelude to the reform of environmental protection fees and taxes, and China’s modern environmental governance system has been further improved. Environmental protection tax, as a market-based regulation means with a higher legislative level and much stronger constraints, solves the past problems of low legislative levels of sewage charges, susceptibility to interference by local governments, and weak rigidity of law enforcement. As such, whether it can improve the corporate ESG performance of enterprises, and which mechanisms and channels play a role in this enhancement, is a worthy area of study.

Some Western countries and the European Union were among the first to levy charges on environmental pollution, and the countries in the Organization for Economic Cooperation and Development (OECD) were the forerunners in imposing environmental taxes [3]. Among various economies in the world, it has been proven that environmental taxes can control environmental pollution and promote green development [4]. UIlah S et al. (2023) found in a study of the top seven green economies that environmental taxes can significantly improve ecological quality [5]. Al Shammre, A.S. et al. (2023), through studying the situation of environmental taxes and carbon dioxide emissions in 34 OECD countries, have confirmed that the levying of environmental taxes can restrain the carbon dioxide emissions of countries [6]. In addition, the environmental protection tax has also been proven to be able to generate double dividends, namely environmental dividends and economic dividends. This indicates that when the environmental tax is utilized rationally, it can not only protect the environment but can also take into account economic development.

The sewage charging system has been the main means of market-based environmental regulation in China since 1979, which can alleviate regional pollution and have a significant emission reduction effect [7]. But some scholars questioned the effectiveness of the sewage charging system due to the insufficient rigidity of law enforcement, more administrative influence, and a lack of standardization [8]. The Environmental Protection Tax Law adheres to the principle of ‘tax burden equalization’ and aligns with the sewage charging system in terms of the taxpayer, the scope of taxable activities, and the calculation method [9]. Thus, the reform of the environmental protection fee system to a tax framework has raised environmental protection from the realm of administrative regulations to the legal domain, making the environmental tax more enforceable and subject to greater oversight [10,11]. Moreover, the new tax law introduces a more refined collection and management system, a more equitable income distribution model, and a more adaptable institutional framework. This elevates the environmental tax above the pollutant discharge charging system, providing stronger legislative support, more robust enforcement, and greater environmental constraints. These features underscore the nation’s commitment to controlling pollution and advancing green development. With the ongoing promotion of green transformation, research has shown that strong ESG performance can yield positive financial outcomes for companies [12,13,14,15]. Therefore, exploring the effects of the environmental fee-to-tax reform not only allows for a more accurate evaluation of the environmental protection tax law’s impact but also offers both theoretical and practical insights into advancing ESG practices within companies.

To examine the impact of the environmental protection tax reform on corporate ESG performance and its underlying mechanisms, we treat the 2018 introduction of the tax in China as a quasi-natural experiment. Using the difference-in-differences (DID) approach, we conduct an empirical analysis. The possible contributions of the article lie in the following: Firstly, it reveals the internal logic of using the environmental protection fee and tax reform to promote corporate ESG enhancement, thus providing a theoretical basis for scientifically and reasonably assessing the economic benefits of environmental protection tax implementation. In the existing studies on ESG theory, most scholars focus on the economic benefits of corporate ESG performance, and there is limited attention given to the connection between environmental protection taxes and ESG outcomes. However, we explore whether the government’s environmental protection fee and tax reform can have an impact on corporate ESG performance, expand the analysis of ESG influencing factors, and reasonably assess the ecological and economic benefits of environmental protection tax implementation. Research achievement provides a theoretical basis for continuing to optimize the formulation of environmental protection tax policies. Secondly, based on the government–market–enterprise framework, it reveals the internal mechanism of environmental protection fee and tax reform to promote the ESG performance of heavily polluting enterprises, and then expands the understanding of the relationship between environmental protection tax reform and enterprise ESG performance. We enrich the research mechanism of corporate ESG influencing factors by specifically analyzing the mechanism path of the environmental fee-to-tax change on ESG performance of heavily polluting enterprises and provide theoretical references for optimizing the transmission mechanism of the role of environmental protection policies.

2. Literature Review and Research Hypotheses

A significant body of research has explored the economic impacts of China’s transition from environmental fees to taxes. The environmental protection fee and tax reform has significantly reduced haze emissions in the region [16], improved regional environmental productivity [17], promoted regional low carbon total factor productivity [18], improved the development of the green economy [19], and lowered the regional total pollutant and carbon-based emissions [20]. Furthermore, the environmental protection fee and tax reform has promoted ESG investment for enterprises in China [21], lowered the pollution emissions of enterprises [22], suppressed the environmental violations of enterprises [23], and promoted the green innovation of enterprises [24] and the environmental protection investment of enterprises [11]. When it comes to the impact of environmental protection fee and tax reform on corporate ESG performance specifically, Wang et al. (2022) found that the introduction of the Environmental Protection Tax Law significantly improved the ESG performance of heavy polluters, because of the internal factors of enterprises [25].

Both sewage charges and the environmental protection tax aim to incorporate the cost of pollutant discharge into business decisions, thereby encouraging companies to consider environmental costs in their production processes. This policy mechanism seeks to reduce emissions and promote green development. Compared to the emission charge, the environmental protection tax is characterized by a higher legislative standing, more stringent enforcement, and stronger compliance requirements. As a result, businesses will face increased emission costs and greater regulatory pressure, prompting them to adopt more sustainable and environmentally friendly development strategies [26]. Through the green transformation, the enterprise could not only reduce the long-term cost of production, but also further enhance the core competitiveness of enterprises [27], which would undoubtedly enhance the environmental performance of the enterprise. In terms of fulfilling social responsibility, the reform of environmental protection fees and taxes will also promote the performance of corporate social responsibility. On the one hand, enterprises will increase their investment in environmental protection and governance in the process of green transformation and green development [11], which in itself is an active undertaking of social responsibility. On the other hand, enterprises may also see the environmental tax reform as a signal from the government. By actively fulfilling their social responsibilities, enterprises are responding to the government’s development policies, which can promote a closer relationship between enterprises and the government and make it more convenient to obtain green and environmental protection preferential policies. In addition, the active response to government policies could bring a “reputation regulation effect” to enterprises [28]. A good market reputation can reduce the cost of capital and alleviate the constraints on financing. In the context of ESG performance, the corporate governance aspect diverges from traditional models by integrating environmental protection and social responsibility into the overall governance framework. This approach ensures that governance structures and decision-making processes prioritize sustainability, preventing an excessive focus on economic goals at the expense of environmental and social concerns. [29]. On the whole, the reform of the environmental protection fee and tax can not only improve the performance of environmental protection and social responsibility of enterprises but can also prompt enterprises to improve their self-monitoring governance systems. Combined with the above analyses, Hypothesis 1 will be tested:

Hypothesis 1: 

The collection of an environmental protection tax can promote the ESG performance of firms.

According to the discussion above, ESG influences can be categorized into three levels: corporate management, market regulation, and government control. In the study of the impact of corporate management on ESG performance, Hu (2023), Eissa, A.M et al. (2024), and Wang et al. (2023) confirmed that the digital transformation of enterprises [30], the board gender diversity [31], and the peer effects of European enterprises [32] can significantly enhance the ESG performance of enterprises, respectively. Xi et al. (2024) also found that firms’ green transformation would contribute to firms’ ESG performance [33]. According to Porter’s hypothesis, appropriate environmental regulation can promote enterprises to carry out green technological innovation. For enterprises, their active green technological innovation can not only improve their own competitiveness, but also realize the green and high-quality development and promote the green transformation of enterprises. Therefore, it will further promote the ESG performance of enterprises. On the one hand, the environmental protection fee and tax reform encourages enterprises to actively carry out green transformation [26]. On the other hand, through the market means, the polluting enterprises were under more stringent constraints on the cost of pollution emissions, which promoted the willingness of enterprises to take part in green transformation, and, thus, promoted the improvement of the ESG performance of enterprises.

Scholars have also found some market-level factors that can affect the ESG performance of enterprises. Taking African listed companies as the research objects, Dsouza, S. et al. (2023) found that higher stock liquidity can promote the ESG performance of companies [34]. Lei (2023) and He et al. (2023) agree that common institutional shareholding affects firms’ ESG performance, but the two scholars’ conclusions are diametrically opposed [35,36]. Zheng, M. et al. (2024) found in a study on 149 countries around the world that the development of information and communication technologies can significantly promote the ESG performance of countries [37]. Chen (2023) [38] argues that investor attention can enhance corporate ESG performance. From the perspective of the investment market, the environmental protection tax law mainly affects the enterprises in the heavy pollution industry as the impact on other enterprises is minimal. Enterprises in the heavy pollution industry will face greater cost pressures and business risks than in other industries, which will inevitably raise the attention of investors to heavy pollution enterprises, and the increase in investor attention will lead to greater supervision of heavy pollution enterprises, so heavy pollution enterprises will pay more attention to sustainable development, which will further enhance the ESG performance of enterprises.

In addition, there are a number of factors at the government level that influence corporate ESG performance. Niu, B. (2024) conducted a study using the data of 27 countries spanning 15 years and concluded that the government’s environmental protection expenditures in a country can significantly promote its ESG performance [39]. In comparison, the exploration of the factors influencing ESG performance at the national level by Mooneeapen, O. et al. (2022) appears to be more in-depth. They conducted a study using the data of 6035 companies from 27 countries and found that there is a negative correlation between the degree of political stability and democracy in a country and the ESG performance of enterprises [40]. In research based on China, Huang (2023), Wang et al. (2023), and Shi et al. (2023) [41,42,43] confirm that governmental environmental regulatory instruments, low-carbon city pilot policies, and the environmental pressure on local officials can all significantly enhance firms’ ESG performance, respectively. The pollutant discharge charging system faces several issues, such as local governments altering the charging standards during policy execution, often driven by concerns over economic growth and other factors. This issue has significantly undermined the effectiveness of environmental regulation in reducing emissions and managing environmental impacts. With the introduction of the environmental protection tax, the collection of sewage charges will be jointly overseen by the tax and environmental protection authorities, thereby improving compliance in tax collection. Additionally, the revised revenue distribution system will strengthen the execution of local government policies. In recent years, as environmental protection and sustainable development have progressed, local authorities, as key stakeholders in enhancing environmental quality, have placed greater emphasis on governance. This shift will drive the adoption of greener practices by polluting industries and improve their ESG performance. (See Figure 1). Hypotheses two to four will be tested, and are as follows:

Hypothesis 2: 

The collection of an environmental protection tax can promote the ESG performance of heavily polluting enterprises by promoting their green transformation.

Hypothesis 3: 

The collection of an environmental protection tax can promote the ESG performance of heavily polluting enterprises by drawing investors’ attention to heavy polluters.

Hypothesis 4: 

The collection of an environmental protection tax can promote the ESG performance of heavy polluting enterprises by enhancing the environmental importance of local governments.

3. Research Design

To investigate the impact of environmental protection tax reform on enterprise ESG performance, the implementation of the Environmental Protection Tax Law is regarded as a quasi-natural experiment in this paper. Using data from listed companies in China between 2014 and 2022 and employing the difference-in-differences (DID) econometric approach, this study examines whether the collection of environmental protection taxes has led to changes in ESG performance between heavily polluting and non-polluting enterprises. Further, we will also explore the heterogeneity among enterprises with different property rights and different regions. Referring to Hu et al. (2023) [30], the specific model is set as follows:

$$ES{G}_{it}={\beta }_0+{\beta }_{1}pollute{d}_r*Afte{r}_t+\rho X_{it}+{\eta }_j+{\lambda }_t+{\epsilon }_{it}$$

(1)

where $ES{G}_{it}$ is the dependent variable, representing the ESG performance of firm i in year t, while $pollute{d}_r$ indicates whether the firm operates in a heavily polluting industry. $Afte{r}_t$ is a dummy variable marking the policy implementation period, and $X_{it}$ includes a set of firm-level and region-level control variables. ${\eta }_j$ and ${\lambda }_t$ represent individual and time fixed effects, respectively, and ${\epsilon }_{it}$ is the random error term.

The explained variable in this paper is the ESG performance of enterprises. Although the research on ESG is quite abundant on a global scale at present, unfortunately, there is still a lack of a unified standard in terms of ESG ratings. After comprehensively considering factors, such as the rating standards, time ranges, and coverage intervals of the ESG rating data of various rating agencies, we have adopted the ESG scores provided by the Bloomberg database. The ESG scores used in the robustness test are from Shanghai Huazheng Index Information Service Company Limited. The explanatory variable is the cross-multiplier of the constructed time dummy variable ($Afte{r}_t$) with the industry pollution level dummy variable ($pollute{d}_r$). The value of the time-dependent dummy variable depends on the year when the environmental protection tax was collected, so it is taken to be 0 for years prior to 2018 and 1 for 2018 and thereafter. Similarly, the value of the industry dummy variable depends on the nature of the firm, with firms in heavily polluting industries taking a value of 1 and other firms taking a value of 0.

Heavily polluted industries were defined with reference to Li et al. (2020) [44], and a total of 18 heavily polluting industries from the A-share market are selected for the study in conjunction with the 2012 National Economy Industry Classification Standard. Specifically, we have selected the coal mining and washing industry, the oil and gas exploration industry, the ferrous metal ore mining and dressing industry, the non-ferrous metal ore mining and dressing industry, the non-metallic ore mining and dressing industry, the textile industry, the textile clothing and apparel industry, the leather, fur, feather and their products and footwear manufacturing industry, the paper making and paper products industry, the petroleum processing, coking, and nuclear fuel processing industry, the chemical raw materials and chemical products manufacturing industry, the pharmaceutical manufacturing industry, the chemical fiber manufacturing industry, the rubber and plastic products industry, the non-metallic mineral products industry, the ferrous metal smelting and rolling processing industry, the non-ferrous metal smelting and rolling processing industry, and the power, heat production, and supply industry.

The difference-in-differences method is favored by economists around the world because it can accurately evaluate the effects of policy implementation and is simple to apply. However, it cannot be denied that the conditions for the establishment of this method are relatively strict. It requires that the implemented policy should be a natural experiment and that the parallel trend should be satisfied between the control group and the experimental group. Meanwhile, when using the difference-in-differences (DID) method to evaluate the policy effects, problems, such as omitted variables and other endogeneity issues, are also prone to occur. To make up for the deficiencies of the DID method and address the possible endogeneity issues, we will conduct a parallel trend test in the robustness check. We will also reasonably select control variables by summarizing a large number of existing studies and verify the robustness of the results by using the triple difference method.

According to the previous analysis, there are three mediating variables in this paper, which are corporate green transformation, investor concern, and local government environmental emphasis. We refer to the practice of Zhou et al. (2022) [45] to measure the corporate green transformation: A total of 113 green transformation keywords of enterprises were selected, and their frequency in the annual reports of listed companies was calculated, with green transformation measured by taking the natural logarithm of the frequency plus one. At the same time, we also consider the time delay of enterprises in green transformation, so the variable is treated with a lag of one phase and recorded as GT. Investor attention, with reference to Deng et al. (2020) [46], is logarithmically processed using the web search index from the CNRDS database. The degree of the local government’s environmental importance, with reference to Chen et al. (2018) [47], is measured by analyzing the frequency share of environmental protection words in the local government’s annual work report, denoted as GEI. The ESG performance of enterprises will also be affected by the relevant factors of the enterprises themselves and the macroeconomic conditions of the cities in which they are located. Thus, the studies of Liu et al. (2022) and Wang et al. (2022) [24,25] are referred to, and the following control variables at the enterprise level and city level are selected. The specific variables are defined in

Table 1. Description of variables.

Variable Type	Variable Name	Variable Symbol	Measurement Method
Explanatory variable	ESG performance	ESG	Bloomberg ESG rating data
Explanatory variable	Reform of environmental protection fees and taxes	DID	Dummy variable for time × dummy variable for industry pollution level
Intermediary variable	Green transformation of enterprises	GT	Using the green transition word frequency text analysis measure
	Investor focus	Investor	Logarithmic web search index of listed companies
	Local government environmental priorities	GEI	Percentage of environmental protection words in the government’s annual government work report
Control variable	Asset size	Lnsize	Assets for the year in logarithmic terms
	gearing	Alr	Ratio of assets to liabilities
	Shareholding concentration	top5_hhi	Shareholdings of the top 5 shareholders during the year
	Return on assets	Roa	Revenue generated per unit of assets
	Enterprise growth capacity	Grow	Growth rate of operating income
	Return on equity	Roe	Earnings per unit of equity
	Age of business	Age	Current year minus year of establishment
	Percentage of secondary sector	Second	Percentage of GDP in the secondary sector of the city
	GDP per capita	Lnrgdp	Logarithmic GDP per capita for the year in the city
	Enterprise value	Tobin Q	Calculate the Tobin Q value of the enterprise for the year

.

The data of mediating variables and control variables are sourced from the annual reports of various companies, the Chinese Research Data Services (CNRDS) database, the China Stock Market & Accounting Research (CSMAR) Database, and the National Bureau of Statistics of China. After obtaining the initial data, we carried out significant data sorting to address the regression bias issues that might be caused by the data. First of all, we excluded the enterprises lacking basic information during the sample period. Secondly, we removed the samples of the financial industry. Then, we eliminated the enterprises that had been warned of delisting during the sample period. Finally, we matched the samples of the enterprises with the ESG information and deleted those enterprises with unsuccessful matching or with too few matching periods. After winsorizing the continuous variables at the 1% level on both tails, the descriptive statistics of the variables are shown in

Table 2. Descriptive statistics.

Variable	N	Mean	p50	SD	Min	Max
ESG	6495	32.48	30.39	8.78	14.82	70.22
lnsize	6495	23.51	23.38	1.30	20.50	28.62
alr	6495	0.478	0.494	0.191	0.03	0.955
top5 hhi	6495	0.535	0.513	0.198	0.208	0.962
roa	6495	0.046	0.037	0.052	−0.312	0.439
grow	6495	0.345	0.125	0.894	−0.787	27.09
roe	6495	0.080	0.080	0.101	−2.095	0.586
age	6495	20.23	20.00	6.296	5.00	121.0
secord	6495	37.16	38.98	11.45	3.71	71.30
lnrgdp	6495	11.54	11.65	0.464	9.636	12.46
tobinq	6495	1.946	1.466	1.397	0.737	12.98

. The mean value of ESG performance is 32.48 and the variance is 8.78, which fully indicates that there are significant differences among the sample firms. Further observation of the control variables reveals that the mean value of asset size is 23.51 with a variance of 1.30, the minimum value of equity concentration is 0.208, the maximum value is 0.962, and the mean value is 0.535. The mean values of gearing ratio and return on assets are 0.478 and 0.046, respectively, and the medians are 0.494 and 0.037, respectively, with a variance of 0.191 and 0.052, respectively. These statistical results prove that there is a certain discrepancy among the control variables, which indicates that the samples selected in this paper are representative and capable of further empirical analysis.

In addition, this paper also collected the amount of environmental protection tax in China each year and the average annual ESG performance of the sample enterprises, and then plotted them into Figure 2 below. Among them, the environmental tax data before 2018 was replaced by the sewage charges collected by the government. It can be seen from the figure that after 2018, the amount of environmental protection tax in China fluctuated greatly, which may be caused by the initial stage of the implementation of the environmental protection tax and the COVID-19 pandemic. However, the average annual ESG performance of the sample enterprises showed a trend of increasing year by year, indicating that enterprises are attaching more and more importance to the ESG framework and that the government’s ESG policy construction has achieved certain results.

4. Empirical Analysis

4.1. Benchmark Regression

This study examines whether the environmental protection fee and tax reform enhance the ESG performance of heavy polluters through regression analysis, with the detailed results presented in

Table 3. Benchmark regression results.

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7
	ESG	ESG	ESG	ESG	E	S	G
DID	1.83 *** (0.38)	1.70 *** (0.37)	1.97 *** (0.39)	1.79 *** (0.38)	4.57 *** (0.77)	0.74 * (0.43)	−0.05 (0.34)
_cons	32.12 *** (0.06)	−4.54 (11.58)	32.09 *** (0.07)	−10.27 (15.61)	−69.61 ** (29.48)	−36.33 ** (17.58)	77.35 *** (14.22)
Control variable	NO	YES	NO	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES	YES	YES
Time fixed effect	YES	YES	NO	NO	NO	NO	NO
Area × time fixed effects	NO	NO	YES	YES	YES	YES	YES

Note: ***, **, and * indicate significance at the 1%, 5%, and 10% significance levels, respectively, and standard errors for firm-level clustering are in parentheses, as in the following tables.

. In order to enhance the robustness of the regression results, this paper refers to Tian et al. (2022) [11] and controls for individual fixed effects and time fixed effects in Columns 1 and 2 and for individual fixed effects and area*time fixed effects in Columns 3 and 4, respectively. The regression further ensures the reliability of the regression results by employing firm-level clustered standard errors.

From Column 1 and Column 2 in Table 3, the DID regression coefficients are 1.83 and 1.70, respectively, controlling for individual fixed effects and time fixed effects, and both are significant at the 1% significance level. Columns 3 and 4 show that the DID regression coefficients are 1.97 and 1.79, controlling for individual fixed effects and area × time fixed effects, and both are also significant at the 1% significance level. The above regression results illustrate that the reform of environmental protection fees and taxes significantly promotes the ESG performance of heavy polluters, and the development of sewage charges into environmental protection taxes has achieved certain results; as such, Hypothesis 1 is initially verified. Further analysis of the regression results for each ESG sub-component (environment, social responsibility, and governance) in heavy polluting enterprises shows that the environmental protection fee and tax reform have regression coefficients of 4.57 and 0.74 for the environmental and social responsibility sub-components, respectively. Both are statistically significant at the 1% and 10% levels. In contrast, the regression coefficient for the corporate governance sub-component is −0.05, though it is not statistically significant. Analysis of the regression results shows that, compared to the social responsibility subcomponent, the environmental protection fee and tax reform has more significantly promoted the environmental performance of heavily polluting enterprises. It is also evident that the regression results for the corporate governance of these enterprises do not align with expectations. This may be due to the difficulty in ensuring the independence of internal governance, along with the persistent characteristics of state-controlled or family-controlled enterprises in China, which leads to the conclusion that the environmental protection fee and tax reform has not significantly enhanced corporate governance.

4.2. Robustness Tests

4.2.1. Parallel Trend Analysis

A key assumption for using DID is that there is no significant pre-policy difference between the treatment and control groups. Here, this requires that prior to the implementation of the environmental protection tax, there is no significant difference in the changing trends of ESG performance between heavily polluting enterprises and lightly polluting enterprises. Thus, we use the event study method to conduct a parallel trend test to test whether this precondition is valid. The year prior to the implementation of the environmental protection tax (2017) was set as the baseline to examine the changes in ESG performance trends of heavily polluting enterprises compared to non-heavily polluting enterprises before and after the tax was implemented. The results of the test are shown in Figure 3 below.

In Figure 3, there is no basic difference between the ESG performance of heavy polluters and light polluters before the environmental protection tax starts to be levied, and after the environmental protection tax starts to be levied, the ESG performance of heavy polluters is significantly better than that of light polluters. Based on this result, it can be reasonably concluded that the ESG performance of heavily polluting enterprises has been promoted by the implementation of the environmental protection tax. The parallel trend test is passed. It can also be seen from the figure that during the two years of policy implementation, namely from 2018 to 2019, the reform of the environmental protection fee and tax significantly promoted the ESG performance of heavily polluting enterprises. However, it is also evident that the promoting effect of the reform on the ESG performance of heavily polluting enterprises was even stronger after 2019. In other words, after fully considering the lagging effect of the policy, it can still be concluded that the ESG performance of heavily polluting enterprises was promoted by the reform, and this promoting effect is increasing year by year.

4.2.2. Placebo Test

This section employs a placebo test to confirm the causal effect of the environmental protection tax on the ESG performance of heavily polluting enterprises, controlling for potential interference from other random factors. Here, the experimental group and the timing of policy implementation were randomly selected, a policy dummy variable was constructed, and this constructed policy dummy variable was incorporated into the regression model for 1000 iterations. If the improvement in ESG performance of heavy polluters is due to the introduction of environmental protection tax, the regression coefficients of the fictitious policy dummy variables should be distributed around 0. Figure 4 illustrates the distribution of regression coefficients and significant levels of the results of 1000 regressions. As seen from the figure, the regression coefficients of the policy dummy variables are centrally distributed around 0 and have a large deviation from the estimated coefficient value of the base regression. This indicates that the policy dummy variables do not significantly enhance the ESG performance of heavy polluters, which indicates the robustness of the base regression results to some extent, and the placebo test is passed.

4.2.3. Replacement of Explanatory Variables

Furthermore, to avoid potential scoring errors caused by a single ESG rating system, the dependent variable was replaced, and the Huazheng ESG rating data was used as the dependent variable to conduct the regression again. The Huazheng ESG rating data, as a local ESG rating system in China, not only references the frameworks of mainstream international ESG evaluation systems, but also adds indicators with Chinese characteristics. In the Huazheng ESG rating system, data sources, such as national regulatory authorities, corporate social responsibility reports, and news media, are utilized, encompassing 14 themes and employing over 130 underlying data indicators, thus ensuring a degree of authority and representativeness. After replacing the explanatory variables, the base regression results are shown in

Table 4. Robustness tests for replacing explained variables.

	Column 1	Column 2	Column 3	Column 4
	ESG (2)	ESG (2)	ESG (2)	ESG (2)
DID	0.68 ** (0.32)	0.47 (0.31)	0.80 ** (0.34)	0.59 * (0.33)
_cons	75.01 *** (0.05)	48.06 *** (7.55)	74.99 *** (0.06)	55.08 *** (11.22)
Control variables	NO	YES	NO	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	YES	NO	NO
Area × time fixed effects	NO	NO	YES	YES

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

below. When individual fixed effects and time fixed effects are controlled, the DID regression coefficients are 0.68 and 0.47, respectively. When individual fixed effects and area*time fixed effects are controlled, the DID regression coefficients are 0.80 and 0.59, respectively. The regression results further indicate that the environmental protection tax reform significantly promotes the ESG performance of heavily polluting enterprises. Similarly, this also validates the robustness of the baseline regression results.

4.2.4. Exclusion of Other Policy Interference

To ensure that the improvement in ESG performance of heavily polluting enterprises is attributed to the environmental protection tax reform rather than other policies or regulations, the potential impacts of carbon trading policies and air pollutant emission restriction policies are further excluded in this paper. Here, the new regression excludes firms in carbon trading pilot regions and firms located in key air pollutant emission control zones.

Table 5. Robustness tests excluding other policy disturbances.

	Column 1	Column 2	Column 3	Column 4
	ESG	ESG	ESG	ESG
DID	1.54 *** (0.45)	1.53 *** (0.47)	2.05 *** (0.61)	2.17 ** (0.69)
_cons	−17.01 (14.39)	−12.61 (17.14)	−23.51 (21.33)	9.40 (29.65)
Control variables	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	NO	YES	NO
Area × time fixed effects	NO	YES	NO	YES

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

presents the regression results after controlling for individual fixed effects, time fixed effects, and area*time fixed effects. From the regression results, the environmental protection fee and tax reform still significantly promotes the ESG performance of heavy pollutant enterprises after excluding the carbon trading policy and the policy of key air pollutant emission control zones, which also proves the robustness of the regression results.

Although the impacts of certain policies have been excluded, our model is limited by the inability to individually exclude the interference of all other policies, which may still lead to bias in the regression results. Referring to the methodology of Zhou et al. [48] (2023), city fixed effects and time and city interaction fixed effects were sequentially added to the base regression model to control for the regression effects of the regional level factors, and the regression results are shown in

Table 6. Robustness tests to further control for fixed effects.

	Column 1	Column 2	Column 3	Column 4
	ESG	ESG	ESG	ESG
DID	1.86 *** (0.38)	1.72 *** (0.37)	2.07 *** (0.43)	1.85 *** (0.42)
_cons	32.12 *** (0.06)	−4.83 (12.30)	32.23 *** (0.07)	7.44 (25.99)
Control variables	NO	YES	NO	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	YES	NO	NO
Urban fixed effect	YES	YES	NO	NO
Urban × time fixed effects	NO	NO	YES	YES

Note: *** indicate significance at the 1% significance levels.

below. It can be seen that the regression results are still significant after controlling for city fixed effects and time*city interaction fixed effects, which confirms the robustness of the regression results.

4.2.5. Application of Triple Difference Models

The Environmental Protection Tax Law specifies that the national government sets the lower limit for the environmental protection tax, while provincial governments have the discretion to determine the upper limit based on local conditions. In other words, provincial governments have the right to determine the pollution levy on their own. As a result, 12 provinces have raised their tax rates, including Hebei, Henan, Jiangsu, Shandong, Hunan, Sichuan, Chongqing, Guizhou, Hainan, Guangxi, Shanxi, and Beijing. Referring to Liu et al. (2022) [24], the third-level difference variable, TAX, is further constructed: TAX takes the value of 1 if the firm is within the above 12 provinces; otherwise, it takes a value of 0. Next, a new triple difference variable, DDD, is constructed (DDD = After*Polluted*TAX). Based on the previous analysis, it can be reasonably inferred that the environmental protection fee and tax reform promotes the ESG performance of heavy polluters more strongly in the provinces where the tax rate has increased, which means that DDD will significantly and positively affect the ESG performance of heavy polluters. From the regression results presented in

Table 7. Robustness test using triple difference model.

	Column 1	Column 2	Column 3	Column 4
	ESG	ESG	ESG (2)	ESG (2)
DDD	0.83 * (0.48)	0.84 * (0.48)	1.05 ** (0.43)	0.95 ** (0.43)
_cons	32.37 *** (0.03)	−5.94 (11.54)	75.06 *** (0.03)	47.77 *** (7.59)
Control variables	NO	YES	NO	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	YES	YES	YES
Area fixed effect	YES	YES	YES	YES

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

, it can be observed that despite using ESG rating data from different sources, the implementation of the policy has significantly promoted the improvement of ESG performance among heavily polluting enterprises. This demonstrates the robustness of the regression results in this paper.

4.3. Analysis of Mechanisms

The benchmark regression results show that the introduction of an environmental protection tax significantly promotes the ESG performance of heavy polluters, but its enhancement mechanism still needs to be further explored. Through a literature summary and theoretical analysis, enterprise green transformation, investor concern, and government environmental importance have been selected as mediating variables to analyze the mechanism path of environmental protection fee and tax reform on enterprise ESG performance.

The environmental protection tax leverages both administrative and market mechanisms to strengthen constraints on corporate pollution behavior. The heightened rigor and standardization associated with this tax increase the cost of emitting pollutants for enterprises to a certain extent. Consequently, enterprises will seek greener and more efficient production methods during the production process, which undoubtedly promotes their green transformation behavior. An important indicator in the environmental component of ESG performance is the green development of the enterprise. The promotion of green transformation will reduce environmental pollution in the production process of the enterprise, which also represents that the enterprise is assuming a higher level of social responsibility. After the implementation of the Environmental Protection Tax Law, heavily polluting enterprises face stricter regulatory constraints and greater operational risks in their production and business activities. In order to make reasonable investment decisions, investors will increase their scrutiny of heavily polluting enterprises, which essentially strengthens the monitoring force on these enterprises. When enterprises face higher investor attention, they will further regulate their production activities and pay more attention to environmental benefits and social responsibility to enhance their brand image and attract more investment, which will further improve their ESG performance. When enterprises face higher investor attention, they will further regulate their production activities and pay more attention to environmental benefits and social responsibility to enhance their brand image and attract more investment, which will further improve their ESG performance. As China’s first green tax law, the implementation of the Environmental Protection Tax Law signifies the increasing emphasis and intensity of the government’s management on environmental protection. To a certain extent, it has broken the previous economic-centric mindset of local governments and heightened their attention to environmental protection. The heightened emphasis on the environment by local governments will prompt them to further constrain heavily polluting enterprises within their jurisdictions, thereby addressing issues, such as low legislative hierarchy, insufficient enforcement rigor during the period of pollution charges, and various fee-related irregularities. This, in turn, will drive heavily polluting enterprises to achieve green development and high-quality development. Based on the above analysis, it can be concluded that the reform of the environmental protection fee and tax can promote ESG performance of enterprises by facilitating their green transformation, increasing investor scrutiny, and enhancing the importance attached by governments to environmental protection.

The mechanism path proposed is examined by referring to Jiang et al. (2022) [49], and the regression results are shown in

Table 8. Mechanism test regression results.

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6
	GT	GT	Investor	Investor	GEI	GEI
DID	0.05 ** (0.03)	0.04 (0.03)	0.05 * (0.03)	0.05 * (0.03)	0.05 *** (0.01)	0.03 *** (0.01)
_cons	1.00 (0.88)	0.77 (1.30)	7.82 *** (0.90)	8.55 *** (1.19)	1.76 *** (0.42)	0.07 *** (0.56)
Control variables	YES	YES	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES	YES
Time fixed effect	YES	NO	YES	NO	YES	NO
Area × time fixed effects	NO	YES	NO	YES	NO	YES

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

. The regression results indicate that when individual fixed effects and time fixed effects are controlled for, the regression coefficient of the environmental protection fee and tax reform on the green transformation of heavily polluting enterprises is 0.05, which is statistically significant at the 5% level. When individual effects and area*time fixed effects are controlled for, the regression coefficient is 0.04. These regression results demonstrate that the environmental protection fee and tax reform promote the ESG performance of heavily polluting enterprises by facilitating their green transformation. Meanwhile, the regression results from Column 3 to Column 6 show that the mechanism of the environmental protection fee and tax reform promoting ESG performance through promoting investors’ concern and the government’s attention to the environment is also valid.

Referring to Tian et al. [11] (2022), a triple-difference model was constructed to examine the mechanism path. Based on the previous analysis, the variables of corporate green transformation, investor attention, and government emphasis on environmental issues were designated as mediating variables. Based on these three mediating variables, new dummy variables are constructed, respectively. Based on these three mediating variables, new dummy variables, namely GT, Investor, and GEI were, respectively, constructed. When the green transformation of enterprises is higher than the median, GT is assigned as 1, and when the green transformation of enterprises is lower than the median, GT is assigned as 0, and then we construct the triple difference variable DDD1 (DDD1= After ∗ Polluted ∗ GT). Based on the above analysis, the ESG performance enhancement of high green transformation firms will be greater, so the regression result of DDD1 on ESG performance should be significantly positive. Following the same method, DDD2 = After ∗ Polluted ∗ Investor and DDD3 = After ∗ Polluted ∗ GEI are constructed. Similarly, if the above mechanism is correct, the regression results of DDD2 and DDD3 on ESG performance should also be significantly positive, and the specific regression results are shown in

Table 9. Regression results of the mechanism test using the triple difference model.

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6
	ESG	ESG	ESG	ESG	ESG	ESG
DDDi	1.45 *** (0.34)	1.34 *** (0.33)	2.31 *** (0.51)	2.06 *** (0.48)	1.43 *** (0.33)	1.23 *** (0.33)
_cons	32.26 *** (0.04)	−6.41 (11.42)	32.30 *** (0.03)	−6.09 (11.43)	32.34 *** (0.02)	−6.89 (11.57)
Control variables	NO	YES	NO	YES	NO	YES
Individual fixed effect	YES	YES	YES	YES	YES	YES
Time fixed effect	YES	YES	YES	YES	YES	YES
Area fixed effect	YES	YES	YES	YES	YES	YES

Note: *** indicate significance at the 1% significance levels.

. The regression results of DDD1, DDD2, and DDD3 on ESG are all positive and significant at the 1% significance level by controlling different fixed effect combinations. The analysis and regression results indicate that the environmental protection fee and tax reform enhances the ESG performance of heavily polluting enterprises by promoting corporate green transformation, increasing investor attention, and elevating government emphasis on environmental issues.

4.4. Heterogeneity Test

4.4.1. Enterprises with Different Nature of Ownership

Enterprises are classified into state-owned and privately-owned companies, and separate regression analyses are conducted to examine the differences in regression outcomes based on the nature of property rights. Compared to privately-owned enterprises, state-owned enterprises are more susceptible to government influence, resulting in stronger execution of government policies. In contrast, privately-owned enterprises tend to prioritize cost and benefit considerations. When facing the same cost constraints, state-owned enterprises benefit from the inherent advantage of government endorsement, making it easier for them to access resources and investment, which aids in their investment in environmentally-friendly technologies. Furthermore, state-owned enterprises may be more driven by government environmental objectives, leading them to emphasize environmental benefits and assume greater social responsibility. These advantages collectively indicate that state-owned enterprises can enhance their ESG performance to a greater extent. The results of the regression analysis conducted on the classified sample enterprises are presented in

Table 10. Heterogeneity test regression results according to the nature of property rights.

	Column 1	Column 2	Column 3	Column 4
	State-Owned Business	State-Owned Business	Private Business	Private Business
DID	1.71 *** (0.50)	1.88 *** (0.54)	1.66 *** (0.55)	1.34 ** (0.60)
_cons	−6.70 (18.19)	7.22 (30.32)	−11.64 (14.70)	−21.05 (20.72)
Control variables	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	NO	YES	NO
Area × time fixed effects	NO	YES	NO	YES

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

below. The regression results clearly indicate that the environmental protection fee tax reform has a greater impact on enhancing the ESG performance of state-owned heavily polluting enterprises compared to privately-owned enterprises. When controlling for individual fixed effects and time fixed effects, the impact coefficient of the environmental protection fee tax reform on state-owned enterprises is 1.71, while the coefficient for privately-owned enterprises is 1.66, both of which are significant at the 1% level. When controlling for individual fixed effects and area ∗ time fixed effects, the impact coefficient of the environmental protection fee tax reform on state-owned enterprises rises to 1.88, while the coefficient for privately-owned enterprises is 1.34, significant at the 1% and 5% levels, respectively.

4.4.2. Enterprises in Different Regions

In addition to differences in property rights, regional variations can also lead to differing regression results. According to the documents issued by the Chinese government, China is divided into four major economic regions: the eastern region, the northeastern region, the western region, and the central region. The regression was conducted again based on the regional classification. During the regression process, both the individual fixed effects and the time fixed effects were controlled. The results are shown in

Table 11. Heterogeneity test regression results according to the region of the firms’ location.

	Column 1	Column 2	Column 3	Column 4
	Eastern Region	Northeast Region	West Region	Middle Region
DID	1.71 *** (0.46)	1.74 * (1.02)	1.50 (1.64)	2.17 ** (1.04)
_cons	5.77 (14.96)	−17.10 (26.60)	−41.92 (38.46)	1.64 (31.79)
Control variables	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES
Time fixed effect	YES	YES	YES	YES

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

below. The regression coefficients of the reform on heavily polluting enterprises in the eastern region, the northeastern region, the western region, and the central region are 1.71, 1.74, 1.5, and 2.17, respectively. Except for those in the western region, these regression coefficients are all statistically significant.

These results indicate that the reform has significantly promoted the ESG performance of enterprises in different regions. Moreover, the improvement in the ESG performance of enterprises in the middle region is the highest, and the results in the eastern region and the northeastern region are similar, while the improvement in the ESG performance of enterprises in the western region is the lowest. This may be related to the differing economic development levels and national strategic positioning of different regions. Among the four major regions, the eastern region has the most developed economy, followed by the middle region, the western region, and the northeastern region in turn. The strategic positioning formulated by the Chinese government for the four major regions also varies significantly, and involves revitalizing the northeastern region, the large-scale development of the western region, promoting the rise of the middle region, and modernizing the eastern region. The middle region enjoys a sound economic development, and the government has set a strategic goal of expecting this to continue, which enables enterprises in the middle region to fully integrate resources and political advantages to significantly enhance their ESG performance. For the western region, its economic development is slightly backward, and the Chinese government encourages it to conduct vigorous development, which may cause the improvement of the ESG performance of enterprises in the western region to lag behind that of other regions.

Similarly, the differences among industries were further taken into consideration. However, since there are numerous industries involved in the sample enterprises, it is not practical to conduct grouped regressions on them one by one. Therefore, we analyzed in detail the annual ESG performance and the respective performance of the three pillars in heavily polluting industries and non-heavily polluting industries through graphing. It can be seen from Figure 5 that the ESG performance of heavily polluting industries is higher than that of non-heavily polluting industries every year, and the gap between the two has gradually widened since 2018. The possible reasons for this phenomenon are that heavily polluting industries are facing greater environmental pressure, and the levy of an environmental protection tax has significantly improved the ESG performance of heavily polluting industries. From the three sub-items of ESG, the gap between heavily polluting industries and non-heavily polluting industries is the largest in terms of environmental impact (E), while the gap is the smallest in terms of corporate governance (G). By observing the sample data after 2018, it can also be found that the implementation of the environmental protection tax has significantly promoted the performance of heavily polluting industries in terms of environmental impact and social responsibility but has little stimulating effect on corporate governance. This is basically consistent with the key results of the benchmark regression.

The results of the heterogeneity discussion not only provide data support for analyzing enterprises with different property rights natures and in different regions, but also offer some insights for us to further analyze the situations in different countries. In some developed Western countries, the levying of environmental taxes and the construction and implementation of the ESG concept have already had a relatively complete framework. Some of these economies, such as the European Union, are leading the construction of the global ESG framework. However, it cannot be denied that there are still many developing countries in the world that are facing challenges in environmental protection and the construction of the ESG concept due to the pressures of economic development, technological limitations, lack of funds, and imperfect systems. The implementation of China’s environmental tax reform and its improvement on the ESG performance of heavily polluting enterprises serve as a good precedent, enabling more countries to realize that economic development and environmental protection may not necessarily be in conflict. For countries that are seeking the path of sustainable development, they can learn some experiences from China’s reform of environmental protection fees and taxes. For example, the government can adopt a gradual strategy for pollution charges and, by formulating clear laws, clarify the objects of collection, tax rates, collection scopes, etc. of the environmental protection tax. Meanwhile, the government should also focus on the collaboration between tax authorities and local governments and encourage enterprises to undergo green transformation through the setting of differentiated tax rates.

5. Conclusions and Recommendations

The implementation of the environmental protection tax was used as a quasi-natural experiment to investigate the impact of the reform from pollution discharge fees to environmental protection taxes on the ESG performance of heavily polluting enterprises. The study found that the environmental protection fee and tax reform promoted the ESG performance of heavy polluters by promoting the green transformation of enterprises, investor attention, and the government’s emphasis on the environment. This finding passes a series of robustness tests, such as the parallel trend test, placebo test, and exclusion of other policy interferences. In-depth analysis revealed that the environmental protection fee and tax reform enhances the ESG performance of state-owned heavy polluters more than that of privately-owned firms. Among different regions, the ESG performance of enterprises in the central region has witnessed the largest improvement margin, while that of enterprises in the western region has the smallest improvement margin.

Based on the previous research findings, combining the research conclusions of this paper with some discussions on less developed countries, several recommendations were proposed to better enhance the pollution control and green development functions of the environmental protection tax:

Firstly, supporting policies for the environmental protection tax have been rationally introduced, and the reward and punishment mechanism has been improved. Establishing a sound and effective tax system is crucial for enhancing corporate environmental responsibility and supporting sustainable development. A well-designed market-based environmental regulation framework can drive the green transformation and upgrading of economic and industrial structures, improving resource efficiency and fostering the sustainable and high-quality growth of the economy. In addition, a perfect institutional framework can provide a clearer and stricter normative system for environmental management, thus realizing the win–win situation of environmental quality improvement and green and efficient economic development. This has very good enlightening significance for underdeveloped countries seeking green transformation and sustainable development. Those governments can establish a complete set of environmental tax legal systems in the form of laws, promote local pollution charging policies in a gradual manner, combine charging with incentives, and then achieve the dual dividends of pollution charging.

Secondly, support for environmental technology and the green transformation of enterprises should also be increased. As key players in the market economy, enterprises’ responsiveness to policy can influence the success of policy implementation. The adoption of environmental protection technologies and green transformation strategies can significantly reduce pollutant emissions and resource consumption, enhancing resource efficiency and economic returns. Moreover, this shift supports the emergence of new green industries, creating fresh drivers for economic growth. However, enterprises which undertake the research and development of environmental protection technology and green transformation alone will face higher cost pressures and operational risks. Thus, the government should also introduce relevant policies to support incentives. This is even more important for the governments in underdeveloped regions, as in these countries, enterprises tend to be more vulnerable and lack resilience. These measures not only promote the transformation of the industrial structure to the direction of environmentally friendly, to fundamentally improve the quality of the environment, but also accelerate the realization of the sustainable development of the whole society.

Finally, the ESG concept should be vigorously promoted, and the awareness of ESG development among enterprises should be strengthened. Whether it is enterprises in developed countries or developing countries, the promotion and development of the ESG (environmental, social, and governance) concept is a core element of the current corporate sustainable development strategy. This is even more important for some enterprises in underdeveloped regions and those without development advantages. When various governments promote the ESG concept, they can encourage enterprises to actively participate in improving the ecological environment, undertake more social responsibilities, and enhance the public’s trust in and social acceptance of enterprises. Accepting and implementing the ESG concept also helps enterprises to establish a fairer, more transparent, and more efficient governance structure, thereby reducing enterprise risks, improving operational efficiency, and enhancing market competitiveness. Strengthening enterprises’ awareness of ESG development and integrating ESG concepts into their strategies and operations not only helps them mitigate and prevent environmental and social risks, but also helps them find a new foothold for development. Therefore, promoting the development of ESG concepts is not only a necessity for enterprises to adapt to the future development trend, but also the key to achieving long-term success and sustainable development in China and many other developing countries economy in the future.