The Impact of Environmental Protection Tax on Green Behaviors and ESG Performance of Industrial Enterprises

Abstract

Environmental protection tax is levied based on various types of emitted pollutants and has a significant impact on the green behaviors and ESG (environmental, social, and corporate governance) performance of enterprises. This article explores the green effect and the impact of environmental protection tax on the green behavior of listed companies with in-depth empirical analysis based on the data of industrial enterprises listed on the A-shares from 2018 to 2022 in China. Research has found that the implementation of environmental protection tax has played a significant driving role in improving the overall performance level of corporate ESG, and this tax system has formed a driving force mechanism for enterprises to increase investment in green innovation and effectively improve their comprehensive ESG performance. Green innovation plays a significant intermediary role between environmental protection tax and corporate ESG performance. It is suggested that regions should adjust the applicable amount of environmental protection tax, increase green innovation, and standardize pollution control and emission reduction regulations.

1. Introduction

Environmental regulation serves as a critical strategy to guide the transition toward green development and achieve high-quality economic growth, especially the environmental tax system. As an effective market-oriented tool for environmental governance, environmental protection tax plays a pivotal role in China’s green transformation. It not only contributes to the harmonious integration of the economy, society, and the environment but also provides strong and consistent support for sustainable development. Since its official implementation on 1 January 2018, the Environmental Protection Tax Law of the People’s Republic of China has focused on taxing four categories of pollutants based on their emission volume or equivalent amount. This policy reflects a transition from pollution discharge fees to environmental taxation. The core objective of introducing the environmental protection tax is to deepen environmental tax system reforms and maximize the realization of the “dual dividend” effect at the macro level. This effect entails achieving a balance between economic and environmental costs for polluters. The first component, the “environmental dividend”, refers to the improvement in regional environmental quality through the binding effect of taxation. The second, the “social dividend”, pertains to the mitigation of the economic burden caused by other taxes, thereby promoting regional economic development. In this context, the concept of Environmental, Social, and Corporate Governance (ESG), which integrates sustainable development with responsible investment, aligns closely with China’s pursuit of high-quality and green economic growth [1]. The ESG framework not only reflects a company’s commitment to environmental stewardship, social responsibility, and corporate governance but also highlights its strategic orientation toward sustainable and high-quality development. Against the backdrop of rising operational costs driven by the dual carbon goals and the imposition of environmental protection taxes, enterprises can leverage green behaviors to fulfill environmental obligations and enhance their ESG performance.

The implementation of the environmental tax system has exerted a dual impact on enterprise green behaviors. On the one hand, it has positively stimulated green technological innovation. Empirical studies, such as those based on industrial enterprises listed on the Chinese A-share markets, demonstrate a significant positive correlation between the introduction of the tax and enterprise-level green innovation [2]. They confirm that the environmental protection tax reform has significantly promoted enterprise green development [3]. On the other hand, certain studies suggest that in specific industries, such as the chemical sector, the environmental protection tax may have a negative impact on green innovation [4]. They find that the short-term effects of the tax on enterprise performance are not significant [5]. There is a dispute regarding that the tax burden remains within a reasonable range and has not substantially affected corporate financial performance [6]. Conversely, they observe that the shift from fees to taxes has enhanced environmental and innovation performance, albeit with limited influence on economic performance [7]. Moreover, the composite green performance index—comprising environmental, innovation, and economic dimensions—has shown improvement. They suggest that the environmental protection tax positively influences ESG performance, although this effect may be delayed [8]. The prevailing view holds that ESG performance can significantly enhance green innovation. It reveals that ESG performance improves green innovation efficiency, primarily through increased risk tolerance and enhanced supply chain leadership [9]. It affirms that ESG initiatives positively affect corporate value and performance [10]. It further demonstrates that ESG performance promotes enterprise innovation through internal and external mechanisms [11].

While existing research primarily explores the relationship between environmental protection tax and green behaviors, as well as its impact on enterprise performance, few studies systematically examine the interplay among environmental tax system, green technology development, and enterprise performance. This paper addresses this gap by examining the micro-level green behaviors and effects of the environmental protection tax. Using green innovation as a mediating variable, it investigates how the tax influences enterprise ESG performance, offering new insights for the development of low-carbon economy.

2. Theoretical Framework and Hypotheses

When assessing the green behaviors and overall performance of enterprises, ESG (Environmental, Social, and Corporate Governance) indicators play a crucial role [12]. This evaluation framework not only captures the environmental achievements of enterprises but also thoroughly examines their fulfillment of social responsibilities and the development of corporate governance structures. Specifically, environmental performance focuses on the environmental challenges enterprises encounter and the effectiveness of their mitigation strategies [13]. It involves the identification and assessment of potential environmental risks, as well as the formulation and implementation of targeted management measures aimed at minimizing the environmental impact of operational activities and promoting sustainable development practices. Social responsibility performance centers on how enterprises fulfill their obligations to various stakeholders, including shareholders, employees, customers, and consumers [14]. It encompasses both the fundamental societal commitments of enterprises and their active participation in national economic development and transformation strategies, thereby demonstrating a strong sense of social responsibility and mission. Corporate governance performance serves as a key indicator of an enterprise’s effectiveness in strategic planning, organizational structure, operational mechanisms, and management practices [15]. It reflects the standardization and efficiency of internal decision-making and execution processes. This indicator not only evaluates the robustness of the corporate governance framework but also highlights the coordination and efficiency of internal operations.

2.1. Environmental Protection Tax and Corporate ESG Performance

The introduction of the environmental tax system, which is levied based on the volume of pollutants discharged by enterprises, has effectively prompted companies to internalize environmental costs into their accounting systems. This has increased cost sensitivity and strengthened corporate accountability for environmental stewardship. As a result, enterprises have placed greater emphasis on pollution control, integrating it into their risk management strategies. This policy has also driven companies to develop a business strategy centered on green and sustainable development, leading to increased investment in environmental protection [16]. Through green technological innovation, firms have transitioned toward more environmentally friendly production models, achieving dual benefits in environmental and economic performance [17]. Additionally, enterprises are encouraged to leverage the oversight functions of the board of supervisors and independent directors to optimize resource allocation in environmental protection, including investments in environmental funds, advanced equipment, and specialized personnel. These efforts contribute to energy conservation, emission reduction, and the enhancement of environmental performance [18]. Improved corporate governance and environmental performance collectively foster a healthier ecological environment, thereby boosting social responsibility performance and ultimately elevating overall ESG performance.

The environmental protection tax serves not only as a regulatory tool to control pollution but also as an incentive for enterprises to enhance their environmental performance [19]. Meanwhile, environmental, social, and governance performance have increasingly become key indicators of corporate competitiveness. The reform of the environmental protection tax represents a significant government initiative to regulate corporate environmental behavior through market-based mechanisms [20]. By increasing tax rates, this reform enhances corporate environmental awareness and encourages proactive environmental measures, resulting in a win-win outcome for environmental protection and economic growth (Figure 1). Based on this analysis, the following Hypothesis 1 is proposed.

Hypothesis 1. 

The imposition of environmental protection taxes can effectively enhance enterprises’ ESG performance.

2.2. Environmental Protection Tax and Green Innovation

The implementation of the environmental tax system has intensified environmental regulation and increased policy rigidity. While this may raise pollution control costs for enterprises, the primary objective of the tax is not merely to impose financial burdens but to reduce environmental pollution [21]. To this end, the policy provides tax incentives for enterprises that meet emission standards, encouraging proactive innovation and the adoption of effective emission reduction measures, thereby fostering a virtuous cycle of environmental protection and economic development [22]. The Porter Hypothesis posits that moderate environmental regulation can stimulate technological innovation. The environmental protection tax adheres to the “polluter pays” principle, with tax liability determined by the volume of pollutants discharged [23]. Developing green products-such as recyclable and reusable ones—and producing goods aligned with sustainable development principles can alleviate environmental pressures, generate innovation-driven benefits, and enhance enterprises’ motivation for green technological innovation [24]. Therefore, the environmental protection tax functions as an economically incentivized regulatory mechanism that can offset the compliance costs of enterprises through innovation compensation [25]. In other words, increased profitability can mitigate the negative impacts of the tax. Furthermore, the tax can serve as a substitute for other forms of taxation. In terms of product structure, transitioning from less environmentally friendly products to “green” or “innovative” alternatives can provide enterprises with competitive advantages in market share and pricing [26]. From a production factor perspective, investing in new pollution control technologies and scientific emission reduction processes can effectively counteract the adverse effects of tax increases.

Under the environmental tax system, the principle of “polluter pays” is strictly enforced, meaning that the tax burden corresponds directly to the volume of pollutants discharged, ensuring fairness through the principle of “more pollution, more tax” [27]. Consequently, increased emissions lead to higher tax liabilities, intensifying the financial pressure on enterprises. To mitigate this burden, companies are not only optimizing internal management and improving operational efficiency but are also increasing investments in green innovation to pursue more sustainable development paths [28]. Through green innovation, enterprises can adopt cleaner production methods, reduce emissions, and achieve both tax savings and long-term sustainability [29]. By adopting advanced process equipment and innovative waste treatment technologies, enterprises can significantly lower pollutant emissions, ensuring compliance with—and even exceeding—national and regional environmental standards, thereby optimizing tax expenditures [30]. Based on this reasoning, the following Hypothesis 2 is formulated.

Hypothesis 2. 

Environmental protection taxes compel enterprises to enhance green innovation.

2.3. The Mediating Effect of Green Innovation on Environmental Protection Tax

The implementation of the environmental tax system has heightened corporate awareness of environmental investment and the urgency of emission reduction [31]. In this context, green innovation plays a pivotal role in achieving emission reduction targets. Enterprises have begun to adjust their product structures and optimize production processes. A key strategy involves focusing on the research, development, and production of environmentally friendly and recyclable products [32]. Traditional non-environmentally friendly products are often associated with high pollution and energy consumption, which are major contributors to environmental degradation [33]. By developing and producing more sustainable products, enterprises can effectively reduce emissions and advance sustainable development goals [34].

Additionally, to minimize pollution during production, enterprises are increasingly engaging in green innovation to improve pollutant treatment processes and enhance their efficiency [35]. This not only delivers environmental benefits but also offsets the costs associated with environmental protection, generating a “process compensation effect.” Overall, the environmental protection tax has incentivized enterprises to prioritize and increase investment in environmental protection. Through green innovation, these investments have effectively contributed to energy conservation and emission reduction, significantly improving corporate environmental performance [36]. Moreover, green innovation helps reduce energy consumption, lower waste disposal costs, and create safer working environments [37]. At the same time, enterprises aim to provide eco-friendly products and proactively contribute to society, thereby enhancing their social responsibility performance [38]. Strong ESG performance can serve as a mechanism for market oversight and motivation, encouraging companies to shift toward green development and improve green innovation efficiency (Figure 1). Based on this, the following Hypothesis 3 is proposed.

Hypothesis 3. 

Green innovation mediates the positive impact of environmental protection taxes on corporate ESG performance.

3. Methods and Data

3.1. Sample Selection

This paper selects the data of industrial enterprises listed on the Shanghai and Shenzhen A-share markets in China from 2018 to 2022 as the research sample. If the sample is extended to more recent years (e.g., 2023–2024), it will help to test the long-term effect of environmental tax. In principle, this expansion is feasible when the same data source can provide timely and comparable observation data. But in practice, it often faces two constraints: one is the lag of public reports and database updates (which will delay the ESG score or the availability of final data for patent authorization); Second, changes in disclosure practices or local law enforcement may lead to cross year comparability problems. Based on considerations of transparency and measurement consistency, the current analysis is limited to 2018–2022. Regarding data collection, the ESG performance data of enterprises in this article mainly comes from the Bloomberg database, the relevant data on green innovation is from the China Research Data Service Platform (CNRDS), and the data on control variables adopts the information from the CSMAR database. When processing the raw data, the following steps were taken in this paper to ensure the quality and integrity of the data: First, all samples that had been specially treated (ST, *ST) were removed; The second is to eliminate the samples with missing observations among all variables. Third, the financial industry has been excluded. The fourth step is to truncate the 1% and 99% quantiles of continuous variables. After the above screening and organization, this paper finally collected a total of 2395 observations from 479 enterprises. The reason why the empirical sample is limited to industrial listed companies of Shanghai and Shenzhen A-shares from 2018 to 2022 is that (1) it directly corresponds to the implementation time of the environmental protection tax law in 2018; (2) the industrial sector is the main source of pollutant emissions and is also the focus of policy; and (3) the information disclosed by A-share listed companies is relatively complete, and the company level tax, Bloomberg ESG score and CNRDS green patent data required in this article can be obtained, so as to support the subsequent panel data analysis. In order to make the location of enterprises more intuitive and clear, a map is inserted to show the average number of enterprises in each province from 2016 to 2020. Since some enterprises are involved in relocation, only the integer part is reserved here. It can be seen that, on the whole, the number of enterprises in the central and eastern regions is more than that in other regions (Figure 2).

3.2. Variable Definition

3.2.1. Explanatory Explained Variable: Environmental Protection Tax

The explanatory variable set in this article is Tax, that is, the tax amount of environmental protection tax. Since 2018, China has introduced environmental protection tax as a policy continuation of the previous pollution discharge fee. Although there are differences in legal forms and collection procedures, they are continuous in policy objectives and tax base (both based on pollutant emissions or equivalent units). In view of this, the research scope of this article will cover both the historical data of the pollution discharge fee and the latest practice of the environmental protection tax. In order to ensure the transparency of measurement, this study adopts the following principles for the construction of tax sequence: when the data source clearly reports the amount of environmental protection tax in 2018 and later for a company in a certain year, the tax is preferred; The only record of the historical discharge fee shall be kept as a historical observation; When the data can be aligned, we align the caliber according to the same pollution base.

Table 1. Variable definition table.

Variable Type	Variable Name	Variable Symbol	Variable Description
Explained variable	the ESG performance of the enterprise	ESG	The ESG Bloomberg database ESG rating score
Explanatory variable	Environmental protection Tax	Tax	The natural logarithm of the amount of environmental protection tax
Mediating variable	Green innovation behavior	GTI	Take the natural logarithm of the total number of green patent applications by enterprises after adding 1
Control variable	Growth capacity	Growth	Growth rate of current operating income
	Capital expenditure	Capital	The natural logarithm of capital expenditure
	Returns on assets	ROA	Net profit of the enterprise divided by total assets
	Asset-liability ratio	Lev	The total liabilities divided by the total assets
	Cash flow	Cash	The ratio of a company’s net cash flow to its total assets
	Share concentration	Share	The shareholding ratio of the top ten shareholders
	Board size	Board	The natural logarithm of the number of people on the board

has indicated that tax includes historical sewage charge and environmental protection tax. In order to clearly show the cost of environmental protection tax in different regions, a map is inserted below. Figure 3 shows the average annual cost of environmental protection tax in each province from 2016 to 2020, with three decimal places reserved. In order to make the figure more concise and beautiful, specific values are not marked one by one.

3.2.2. Explained Variable: ESG Performance of Enterprises

The explained variable in this article is the ESG performance of enterprises, that is, the ESG score of enterprises. The ESG score of enterprises is derived from the Bloomberg database. It is based on public information available to investors, such as corporate social responsibility reports, annual reports, and websites, aiming to comprehensively assess the performance of enterprises in the three key dimensions of the environment, society, and corporate governance. The score range is set between 0.1 and 100 points. To objectively reflect the ESG performance of enterprises.

3.2.3. Moderating Variables: Green Innovation Behavior

The mediating variable in this article is GTI, which stands for the quantitative indicator of enterprise green innovation behavior—the number of green patent applications. Given that many scholars in existing research generally tend to use the number of green patent applications as an effective indicator to measure the green technological innovation achievements of enterprises, this paper also adopts a similar approach, choosing to take the total number of green patent applications of enterprises (including the number of green inventions and green utility models) as the benchmark, and adding 1 to the total number of patents to take the natural logarithm as the mediating variable to comprehensively measure the performance of enterprises in green innovation.

3.2.4. Controlled Variables

To provide a more comprehensive analysis of the influencing factors, the following control variables are included in this paper: The specific definitions and measurement standards of variables such as Growth capacity (Growth), Capital expenditure (Capital), Returns on assets (ROA), Asset-liability ratio (Lev), Cash flow (Cash), Share concentration (Share), and Board size (Board) are detailed in Table 1.

3.3. Model Specification

The impact effect of the implementation of environmental protection tax policies on the ESG performance of enterprises, Model 1 is established [39].

$${ESG}_{i,t}={\phi }_0+{\phi }_1{Tax}_{i,t}+{\sigma X}_{i,t}+{\gamma }_t+{\vartheta }_{i,t}+{\epsilon }_{i,t}$$

(1)

To explore the impact of the implementation of environmental protection tax policies on the green innovation capabilities of enterprises, Model 2 is established.

$${GTI}_{i,t}={\beta }_0+{\beta }_1{Tax}_{i,t}+{\sigma X}_{i,t}+{\gamma }_t+{\vartheta }_{i,t}+{\epsilon }_{i,t}$$

(2)

To study the impact of an enterprise’s green innovation level on its ESG performance, Model 3 is established.

$${ESG}_{i,t}={\mu }_0+{\mu }_{1}GTIi,t+\sigma Xi,t+{\gamma }_t+{\vartheta }_{i,t}{+\epsilon }_{i,t}$$

(3)

To examine whether green innovation plays a mediating role between environmental protection tax and the ESG performance of enterprises, Model 4 is established.

$${ESG}_{i,t}={ \alpha }_0+{\alpha }_1{Tax}_{i,t}+{\alpha }_2{GTI}_{i,t}+{\sigma X}_{i,t}+{\gamma }_t+{\vartheta }_{i,t}+{\epsilon }_{i,t}$$

(4)

where Subscripts i and t indicate company and year, respectively. ${Tax}_{i,t}$ is the explanatory variable, and the natural logarithm of the company’s annual environmental protection tax amount is taken as the measurement (see Table 1). ${ESG}_{i,t}$ is the explained variable, that is, the Bloomberg score of company i in year t. ${GTI}_{i,t}$ is the mediating variable, indicating green innovation, and operationalization, =ln(1 + number of green patent applications of company i in year t). $X_{i,t}$ are all the control variables, including Growth, Capital, ROA, Lev, Cash, Share and Board. ${\gamma }_t$ represents the annual fixed effect,${\vartheta }_{i,t}$ represents the individual fixed effect, and ${\epsilon }_{i,t}$ represents the error term. ${\phi }_0$, ${\beta }_0$, ${\mu }_0$, ${\alpha }_0$ are intercept items.

3.4. Descriptive Statistics

Table 2. Descriptive statistics of main variables.

Variable	Sample Observations	Mean	Standard Deviation	Minimum Value	Maximum Value
ROA	2395	0.053	0.0568	−0.253	0.372
Lev	2395	0.449	0.180	0.008	0.941
Cash	2395	0.080	0.077	−0.319	0.706
Share	2395	0.594	0.152	0.188	1.010
Growth	2395	0.215	0.575	−0.863	9.235
Board	2395	2.178	0.207	1.099	2.833
Capital	2395	19.970	1.613	13.700	26.510
ESG	2395	33.270	7.773	19.010	67.940
GTI	2395	1.756	1.543	0.000	7.307
Tax	2395	15.070	1.397	10.160	21.070

provides a detailed descriptive statistical overview of the main variables discussed in this paper. The average ESG score of enterprises is 33.27, the standard deviation is 7.773, the minimum value is 19.01, and the maximum value is 67.94, indicating that there are significant differences in ESG levels among different enterprises, and the ESG performance of enterprises needs to be improved. The average total number of green patents (GTI) is 1.756, the standard deviation is 1.543, and the minimum and maximum values are 0 and 7.307, respectively. It can be seen that there is a significant gap between various enterprises in terms of green technology innovation patents, which highlights the importance of strengthening the environmental protection awareness of industrial enterprises. The road to achieving green development is both long and full of challenges. The statistical analysis results of the control variables are basically consistent with the existing research achievements.

4. Empirical Analysis and Tests

4.1. Correlation Analysis

After a detailed analysis of the correlation coefficients in the table, it can be clearly observed that there is a significant correlation between environmental protection tax, green innovation and the ESG performance of enterprises. This discovery not only verifies the rationality of the measurement indicators selected in this paper, but also further highlights the value of in-depth research among these variables. As shown in

Table 3. Correlation analysis results.

Variable	ESG	GTI	Tax	ROA	Lev	Cash	Share	Growth	Board	Capital
ESG	1
GTI	0.304 ***	1
Tax	0.480 ***	0.585 ***	1
ROA	0.022	−0.151 ***	0.006	1
Lev	0.128 ***	0.390 ***	0.475 ***	−0.436 ***	1
Cash	0.097 ***	−0.108 ***	0.094 ***	0.616 ***	−0.187 ***	1
Share	0.249 ***	0.094 ***	0.326 ***	0.176 ***	0.011	0.167 ***	1
Growth	−0.117 ***	−0.002	0.014	−0.047 **	0.053 ***	−0.092 ***	0.016	1
Board	0.105 ***	0.082 ***	0.160 ***	−0.063 ***	0.161 ***	−0.013	0.064 ***	0.074 ***	1
Capital	0.442 ***	0.517 ***	0.741 ***	0.001	0.389 ***	0.150 ***	0.295 ***	−0.117 ***	0.164 ***	1

Note: The Person correlation coefficient, ** indicates p < 0.05, and *** indicates p < 0.01.

, the correlation coefficient between environmental protection tax and the ESG performance of enterprises is 0.48, and it is significant at the set 1% significance level, fully demonstrating that there is a positive and constructive connection between environmental protection tax and the ESG performance of enterprises. There is a significant positive correlation between the implementation of environmental protection tax and the green innovation level of enterprises. Specifically, the correlation coefficient between the two reaches 0.304, and a clear positive correlation is shown at the significance level of 1%. When evaluating the ESG performance of enterprises, except for Returns on assets (ROA), all other control variables show significant correlations with the ESG performance of enterprises. This discovery not only confirms the appropriateness of the selection of control variables, but also validates the validity of the assumptions of the linear model. The lack of a significant association between ROA and ESG performance may be attributed to the heterogeneous characteristics of the samples, which downplays or offsets the potential impact of this indicator on the ESG performance of enterprises in the overall dataset.

4.2. Basic Regression Analysis

Given the potential issues associated with multicollinearity, this study employs the variance inflation factor (VIF) to rigorously assess each variable, ensuring the robustness of the data model. The VIF values for all variables were found to be substantially below 5, indicating the absence of significant multicollinearity among the independent variables. Based on this finding, the research proceeds with multiple regression analysis to explore the potential relationships and interactions among environmental protection taxes, corporate green innovation behavior, and ESG performance.

First, this study utilizes multiple regression analysis to examine the correlations among environmental protection taxes, green innovation behavior, and ESG performance, yielding corresponding empirical results. Detailed data are presented in

Table 4. Basic empirical results.

Variables	Model 1	Model 2	Model 3	Model 4
	ESG	GTI	ESG	ESG
Tax	0.717 ***	0.115 ***		0.665 ***
	(−0.232)	(−0.041)		(−0.232)
ROA	−2.504	0.008	−0.472	−2.508
	(−2.437)	(−0.428)	(−2.328)	(−2.43)
Lev	−2.293	−0.153	−1.222	−2.224
	(−1.411)	(−0.248)	(−1.366)	(−1.407)
Cash	−1.127	−0.251	−0.742	−1.015
	(−1.444)	(−0.254)	(−1.44)	(−1.44)
Share	5.119 **	−0.048	5.293 ***	5.140 ***
	(−1.988)	(−0.349)	(−1.985)	(−1.982)
Growth	−0.422 **	−0.073 **	−0.262	−0.389 **
	(−0.192)	(−0.034)	(−0.187)	(−0.192)
Board	−0.09	0.144	−0.081	−0.155
	(−0.883)	(−0.155)	(−0.882)	(−0.88)
Capital	0.246 *	0.100 ***	0.228 *	0.201
	(−0.134)	(−0.024)	(−0.135)	(−0.135)
GTI			0.471 ***	0.447 ***
			(−0.13)	(−0.13)
Cons	13.234 ***	−2.403 ***	22.754 ***	14.309 ***
	(−4.454)	(−0.783)	(−3.345)	(−4.453)
N	2395	2395	2395	2395
R2	0.372	0.149	0.374	0.376
Individual fixed	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t, the same below.

. In Model 1, the analysis focuses on the impact of environmental protection taxes on ESG performance. The results indicate that the coefficient for the environmental protection tax variable is positive and statistically significant at the 1% level. This suggests that the imposition of environmental protection taxes positively influences corporate ESG performance, thereby supporting Hypothesis 1. In Model 2, a regression analysis is conducted to investigate the relationship between environmental protection taxes and green innovation behavior. The regression results reveal a positive and statistically significant coefficient at the 1% significance level, indicating that environmental protection taxes encourage enterprises to enhance their green innovation efforts. This supports Hypothesis 2. Model 3 examines the effect of green innovation on ESG performance. The results show a statistically significant positive impact at the 1% level, confirming that green innovation contributes to improved ESG performance.

Secondly, this paper analyzes the mediating effect of environmental protection tax on the green technological innovation of enterprises [40]. To assess the mediating role of green innovation behavior, this study adopts the three-step regression method. Model 1 confirms that environmental protection taxes positively influence corporate ESG performance. Model 2 further demonstrates a significant positive correlation between environmental protection taxes and green innovation, suggesting that taxation policies stimulate green innovation behavior. Model 4 incorporates both explanatory and mediating variables in the regression analysis. The results show that both the impact of green innovation and environmental protection taxes on ESG performance remain statistically significant at the 1% level. These findings indicate that green innovation partially mediates the relationship between environmental protection taxes and ESG performance, thereby validating Hypothesis 3. Further analysis using an extended mediation model confirms that environmental protection taxes significantly influence ESG performance through green innovation pathways.

4.3. Robustness Test

4.3.1. Robustness Tests: Alternative Measurement of Core Variables

To enhance the reliability and validity of the empirical findings, a series of robustness tests were conducted. Green innovation was alternatively measured using the number of authorized green utility model patents (GTI1), as patent applications may not fully reflect actual innovation outcomes. Regression analysis using this refined indicator yielded results consistent with the original findings (

Table 5. Robustness test results after changing core variables.

Variables	Model 1	Model 2	Model 3	Model 4
	ESG	GTI1	ESG	ESG
Tax	0.717 ***	0.068 *		0.690 ***
	(−0.232)	(−0.04)		(−0.232)
ROA	−2.504	−0.313	−0.257	−2.38
	(−2.437)	(−0.415)	(−2.33)	(−2.432)
Lev	−2.293	−0.221	−1.163	−2.205
	(−1.411)	(−0.24)	(−1.367)	(−1.408)
Cash	−1.127	−0.276	−0.736	−1.017
	(−1.444)	(−0.246)	(−1.441)	(−1.442)
Share	5.119 **	0.005	5.274 ***	5.117 ***
	(−1.988)	(−0.338)	(−1.987)	(−1.984)
Growth	−0.422 **	−0.071 **	−0.262	−0.394 **
	(−0.192)	(−0.033)	(−0.187)	(−0.192)
Board	−0.09	0.048	−0.03	−0.11
	(−0.883)	(−0.15)	(−0.882)	(−0.881)
Capital	0.246 *	0.077 ***	0.244 *	0.215
	(−0.134)	(−0.023)	(−0.135)	(−0.135)
GTI1			0.413 ***	0.398 ***
			(−0.135)	(−0.134)
Cons	13.234 ***	−1.457 *	22.560 ***	13.814 ***
	(−4.454)	(−0.758)	(−3.347)	(−4.45)
N	2395	2395	2395	2395
R2	0.372	0.102	0.372	0.375
Individual fixed	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t.

).

4.3.2. Robustness Tests: Impact of Capital Governance

Capital intensity (Density) was introduced as an additional control variable. The regression results remained consistent with previous conclusions, reinforcing the robustness of the findings, as

Table 6. Empirical results after adding capital intensity.

Variables	Model 1	Model 2	Model 3	Model 4
	ESG	GTI1	ESG	ESG
Tax	0.865 ***	0.125 ***		0.810 ***
	(−0.247)	(−0.043)		(−0.247)
ROA	−1.673	0.064	−0.057	−1.701
	(−2.482)	(−0.437)	(−2.43)	(−2.475)
Lev	−2.426 *	−0.162	−1.193	−2.354 *
	(−1.412)	(−0.248)	(−1.367)	(−1.409)
Cash	−0.98	−0.241	−0.675	−0.873
	(−1.446)	(−0.254)	(−1.444)	(−1.442)
Share	4.940 **	−0.06	5.246 ***	4.967 **
	(−1.989)	(−0.35)	(−1.987)	(−1.984)
Growth	−0.474 **	−0.077 **	−0.269	−0.440 **
	(−0.195)	(−0.034)	(−0.188)	(−0.194)
Board	−0.123	0.141	−0.086	−0.186
	(−0.882)	(−0.155)	(−0.882)	(−0.88)
Capital	0.209	0.098 ***	0.218	0.165
	(−0.136)	(−0.024)	−0.136)	(−0.136)
Density	0.201 *	0.013	0.065	0.195 *
	(−0.115)	(−0.02)	(−0.108)	(−0.115)
GTI1			0.472 ***	0.444 ***
			(−0.13)	(−0.13)
Cons	11.501 **	−2.519 ***	22.804 ***	12.620 ***
	(−4.562)	(−0.803)	(−3.347)	(−4.561)
N	2395	2395	2395	2395
R2	0.373	0.15	0.374	0.377
Individual fixed	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t.

.

The shareholding ratio of the top ten shareholders was replaced with the shareholding ratio of the largest shareholder (Top1), as

Table 7. Empirical results after adding shareholding ratio.

Variables	Model 1	Model 2	Model 3	Model 4
	ESG	GTI1	ESG	ESG
Tax	0.740 ***	0.112 ***		0.689 ***
	(−0.233)	(−0.041)		(−0.232)
ROA	−2.425	0.052	−0.315	−2.449
	(−2.444)	(−0.429)	(−2.334)	(−2.437)
Lev	−2.499 *	−0.14	−1.398	−2.436 *
	(−1.412)	(−0.248)	(−1.367)	(−1.408)
Cash	−1.29	−0.262	−0.901	−1.172
	(−1.445)	(−0.253)	(−1.441)	(−1.441)
Top1	1.828	−0.682 *	1.86	2.136
	(−2.34)	(−0.411)	(−2.338)	(−2.335)
Growth	−0.391 **	−0.072 **	−0.224	−0.359 *
	(−0.192)	(−0.034)	(−0.187)	(−0.192)
Board	−0.136	0.137	−0.126	−0.198
	(−0.884)	(−0.155)	(−0.883)	(−0.882)
Capital	0.270 **	0.099 ***	0.253 *	0.225 *
	(−0.134)	(−0.024)	(−0.135)	(−0.135)
GTI1			0.475 ***	0.451 ***
			(−0.13)	(−0.13)
Cons	14.971 ***	−2.098 ***	24.880 ***	15.917 ***
	(−4.488)	(−0.787)	(−3.32)	(−4.484)
N	2395	2395	2395	2395
R2	0.37	0.151	0.371	0.374
Individual fixed	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t.

. The regression results remained largely unchanged, further confirming the stability of the conclusions.

4.4. Further Analysis

4.4.1. Heterogeneity Analysis by Pollution Level

Enterprise pollution levels may influence the effectiveness of environmental protection taxes on ESG performance and green innovation behavior. Group analysis based on pollution levels (

Table 8. Results of heterogeneity test by pollution level.

Variables	Model (1)	Model (2)	Model (3)	Model (4)
	Non-Heavily Polluting Enterprises	Heavily Polluting Enterprises	Non-Heavily Polluting Enterprises	Heavily Polluting Enterprises
	ESG	ESG	GTI	GTI
Tax	0.828 **	0.625 **	0.142 **	0.100 *
	(−0.345)	(−0.312)	(−0.063)	(−0.054)
ROA	−4.956	−1.345	0.049	−0.062
	(−3.365)	(−3.459)	(−0.617)	(−0.598)
Lev	−2.47	−1.315	−0.496	0.017
	(−2.124)	(−1.944)	(−0.39)	(−0.336)
Cash	−1.721	−0.669	−0.654 *	0.05
	(−2.079)	(−2.026)	(−0.381)	(−0.35)
Share	−0.57	7.675 ***	−0.436	0.056
	(−3.103)	(−2.606)	(−0.569)	(−0.451)
Growth	−0.562 *	−0.437 *	−0.03	−0.099 **
	(−0.309)	(−0.248)	(−0.057)	(−0.043)
Board	−0.696	0.159	0.363	−0.005
	(−1.228)	(−1.233)	(−0.225)	(−0.213)
Capital	0.458 **	0.011	0.109 ***	0.087 ***
	(−0.191)	(−0.187)	(−0.035)	(−0.032)
Cons	11.813 *	16.922 ***	−2.505 **	−2.129 *
	(−6.091)	(−6.364)	(−1.117)	(−1.101)
N	993	1402	993	1402
R2	0.36	0.391	0.164	0.148
Individual fixed	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t.

) reveals that ESG performance is positively correlated with environmental protection taxes in both heavily and non-heavily polluting enterprises. However, the effect is more pronounced in non-heavily polluting firms, likely due to lower emission baselines and greater social responsibility awareness. Similarly, green innovation is positively associated with environmental taxes across pollution levels, with stronger effects in non-heavily polluting enterprises, possibly due to cost constraints in heavily polluting firms.

4.4.2. Regional Heterogeneity Analysis (Central and Western Regions)

Regional differences may affect how environmental protection taxes influence ESG performance and green innovation behavior. Dividing the sample into eastern, central, and western regions (

Table 9. Heterogeneity test results by region.

Variables	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
	Eastern Region	Western Region	Central Region	Eastern Region	Western Region	Central Region
	ESG	ESG	ESG	GTI	GTI	GTI
Tax	0.787 ***	1.238 **	0.063	0.150 ***	0.011	0.003
	(−0.294)	(−0.567)	(−0.511)	(−0.051)	(−0.101)	(−0.096)
ROA	0.014	−19.539 ***	1.207	−0.291	−0.091	0.948
	(−3.023)	(−6.714)	(−5.224)	(−0.523)	(−1.199)	(−0.979)
Lev	−2.529	−4.693	1.422	−0.51	0.941	0.104
	(−1.806)	(−3.474)	(−2.926)	(−0.312)	(−0.621)	(−0.548)
Cash	−3.190 *	1.272	3.051	−0.605 *	−0.621	0.991 *
	(−1.841)	(−3.613)	(−3.085)	(−0.319)	(−0.645)	(−0.578)
Share	0.828	15.653 ***	13.228 ***	−0.082	−0.224	−0.15
	(−2.503)	(4.916)	(−4.5)	(−0.433)	(−0.878)	(−0.843)
Growth	−0.487 **	−0.671	0.144	−0.057	−0.056	−0.101
	(−0.243)	(−0.453)	(−0.457)	(−0.042)	(−0.081)	(−0.086)
Board	−1.592	2.108	2.153	0.317	−0.55	0.11
	(−1.199)	(−2.152)	(−1.624)	(−0.207)	(−0.384)	(−0.304)
Capital	0.216	0.384	0.078	0.115 ***	0.081	0.091 *
	(−0.175)	(−0.308)	(−0.284)	(−0.03)	(−0.055)	(−0.053)
Cons	18.993 ***	−5.983	12.835	−3.273 ***	0.576	−1.001
	−5.816	−11.066	−8.984	−1.006	−1.977	−1.684
N	1490	478	427	1490	478	427
R2	0.407	0.305	0.397	0.156	0.144	0.196
Individual fixed	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed	Yes	Yes	Yes	Yes	Yes	Yes

Note: * indicates p < 0.1, ** indicates p < 0.05, and *** indicates p < 0.01; The values in parentheses are t.

), the results show that ESG performance is significantly positive in both eastern and western regions, with a stronger effect in the eastern region. This may be attributed to the more developed economic environment and better resource availability in eastern China. Regarding green innovation, only the eastern region shows a significant positive relationship, likely due to its more mature green innovation ecosystem.

4.5. Discussion—Comparison with Prior Studies

Our empirical results show that the implementation of environmental protection taxes significantly improves corporate ESG performance (Model 1: Tax = 0.717, p < 0.01) and stimulates firm green innovation (Model 2: Tax = 0.115, p < 0.01). These findings are broadly consistent with a growing stream of literature that finds that market-based environmental regulation—including environmental taxation—can induce firms to internalize environmental costs and increase green investment and innovation [1,2,8]. Mechanically, the results of this paper support an ‘innovation compensation’ view similar to the Porter hypothesis: by raising the marginal cost of pollution, the tax motivates firms to adopt cleaner technologies and product/process innovations, which in turn improve their ESG outcomes.

At the same time, some early research reports pointed out that environmental tax would have different or insignificant effects under certain circumstances. For example, it would have negative or insignificant short-term effects in some heavily polluted sectors. Differences have been found at the industry level [4], and the short-term effects are sometimes limited [5]. This difference can be reconciled by three factors. First of all, industry and pollution intensity are important: enterprises in heavy pollution industries often face greater short-term compliance costs and financing constraints, which may inhibit the immediate innovation response. The heterogeneity test results show that for non heavy polluting enterprises, the tax effect is stronger, which is consistent with channel differences. Second, regional institutions and innovation capabilities (such as the eastern and central/western regions) affect the ability of tax incentives to transform into innovative output; The regional test of this paper shows that the innovation response in the eastern region is more obvious. Third, the issue of measurement and time is important: studies using short windows (2016–2020) or only patent authorizations (relative to applications) or different ESG scoring sources may capture different stages of the innovation cycle. The follow-up study can further test the mechanism proposed in this paper by using longer panel, micro data of enterprise investment and trans regional migration indicators. Law enforcement and supplementary policies (subsidies, tax credits, local law enforcement) also determine the final result.

4.6. Limitations

4.6.1. Limitations of Excessive Environmental Taxes and Potential Enterprise Relocation

While empirical results show that environmental protection taxes generally stimulate green innovation and improve corporate ESG performance through innovation pathways, it is important to acknowledge boundary conditions. If tax rates increase too rapidly or to levels beyond firms’ short-term capacity to adapt—due to technological constraints, financing limitations, or long implementation lags—firms may experience rising operating costs, compressed margins, and reduced investment, with potential consequences including output contraction, employment reduction, or even market exit for the most vulnerable firms. Some firms may consider relocation to jurisdictions with lower tax burdens or laxer enforcement as a cost-minimizing response; however, relocation is constrained by sunk capital, supply chain dependencies, permitting and regulatory procedures, and reputational and contractual risks, and therefore is more feasible for mobile, light-asset activities than for heavy-industry plants. Notably, our heterogeneity results indicate that heavily polluting firms and less developed regions tend to have weaker innovation responses to tax incentives, making them particularly vulnerable if taxes rise sharply (see Table 8 and Table 9). Based on this, to avoid unintended adverse outcomes, policymakers should consider phased tax adjustments, complementary innovation subsidies or tax credits targeted at vulnerable firms, strengthened regional coordination to reduce incentives for relocation (and associated carbon leakage), and dedicated technical and financing support for high-polluting enterprises to facilitate low-cost abatement options.

4.6.2. Institutional Differences & Limitation

There are differences in legal status and administrative procedures between the pollution charge (before 2018) and the environmental protection tax (after 2018), which may affect the reporting behavior, law enforcement intensity and short-term response of enterprises. Although this paper constructs a continuous tax series to measure the monetary burden related to emissions through the unified caliber, institutional differences may bring heterogeneity in measurement (for example, in terms of recording time or local law enforcement differences). When interpreting the research results, especially the short-term effects, we should pay attention to this assumption and limitation. Therefore, this paper limits the conclusion to the unified framework adopted, and suggests that future research should further separate the impact of institutional channels by using administrative taxation records or law enforcement intensity indicators when available.

5. Conclusions

Using panel data from industrial enterprises listed on the Shanghai and Shenzhen A-share markets in China from 2018 to 2022, this study constructs a research sample and applies multiple regression analysis to explore the relationships among environmental protection taxes, green innovation behavior, and ESG performance. The key findings are as follows: (1) Environmental protection taxes positively influence corporate ESG performance; (2) environmental taxes drive enterprises to engage in green innovation; and (3) green innovation partially mediates the relationship between environmental protection taxes and ESG performance, indicating that taxation enhances ESG outcomes through green innovation. In general, the results of the study are consistent with the conclusion generally supported by existing studies, that is, environmental tax can promote green innovation and improve enterprise ESG performance. By combining benchmark regression, mediation effect and heterogeneity test, this study helps to reconcile the differences in the literature, and highlights the differential effect of environmental tax in different industries and regions.

Based on this, the following policy suggestions are put forward: (1) Expand the Scope and Adjust the Rate of Environmental Protection Taxes. Currently, the environmental tax rate in China is relatively low, limiting its effectiveness in promoting pollution reduction. Regions should align tax burdens with pollution control costs and adjust tax rates according to regional development needs to encourage environmental innovation and tax reduction incentives. (2) Promote Green Innovation in Enterprises. Green innovation not only reduces tax burdens but also enhances ESG performance. Enterprises should optimize production processes, adopt sustainable practices, meet or exceed environmental standards, and improve waste treatment efficiency. Additionally, they should explore solid waste resource utilization. (3) Strengthen Corporate Environmental Awareness. Enhancing environmental consciousness is essential for sustainable development. Enterprises should establish environmental value systems, disclose environmental information, and accept public and governmental oversight. This improves transparency, builds a positive image, and facilitates continuous improvement in environmental management. (4) Establish Supportive Policies for Green Innovation. Environmental protection taxes alone may not sufficiently incentivize green innovation. A coordinated approach combining environmental regulations and innovation incentives is necessary. This includes reducing reliance on administrative enforcement, strengthening legal compliance, and introducing economic incentives to promote green development.