Environmental Protection Tax Policy and Green Innovation of Heavy-Polluting Enterprises: Evidence from China

: Since 2018, China has implemented the Environmental Protection Tax as a market-based environmental regulation tool to encourage enterprises, especially heavy-polluting ones, to reduce emissions and engage in green innovation. The effectiveness of this policy in promoting green innovation has become a focal point of attention for Chinese society, as it directly relates to the coordinated high-quality development of the economy and environmental protection, as well as the achievement of China's “carbon peaking and carbon neutrality goals”. This study evaluates the green innovation effect of the Environmental Protection Tax policy on heavy-polluting enterprises in China using micro-data of listed enterprises from 2003 to 2019, based on the new market-oriented environmental regulation tool. We employ the propensity score matching and difference-in-differences method (PSM-DID) to follow up on the green innovation effect of China's Environmental Protection Tax policy and reveal its implementation effect. Empirical results indicate that, first, the Environmental Protection Tax policy contributes to heavy-polluting enterprises' R&D innovation of green patents, and this conclusion holds through various robustness tests. Second, the policy's impact on green innovation types of heavy-polluting enterprises exhibits heterogeneity, with a more significant positive effect on the utility model patent than on the invention patent. Third, the green innovation effect of the policy differs according to regional economic development and enterprise ownership attributes. Finally, financing constraints do not significantly affect the green innovation of heavy-polluting enterprises. These enterprises remain pro-active in green innovation, even under high financing constraints. To further promote green innovation of heavy-polluting enterprises and achieve China's dual carbon goals, the government should continue to implement and improve the Environmental Protection Tax policy, leveraging its positive effect on green innovation.


Introduction
With the increasingly severe global environmental issues such as climate change and the continuous advancement of ecological civilization construction in our country, as a responsible developing country, China proposed in 2021 to strive to achieve the goals of "carbon peak" by 2030 and "carbon neutrality" by 2060, respectively. However, due to the public nature of the environment, while industrial production and emissions create economic benefits for enterprises, the resulting environmental damage is borne by the entire society. Therefore, the fundamental approach to prevent enterprises from externalizing environmental costs to society and to achieve emission reduction targets is to encourage companies, especially heavily polluting ones, to internalize the costs of environmental pollution, thereby incorporating the economic losses caused by environmental pollution into their production and emission reduction decision-making. At the same time, as the mainstay of the market economy, enterprises play an important role in the green transformation of the economy and society. Market-based environmental regulations can more effectively incentivize emissions reduction. To this end, China implemented the Environmental Protection Tax in 2018 as a market-based environmental regulatory policy to replace the previous environmental fee system, which had been in operation for thirty years. The aim is to promote enterprises to more effectively internalize the marginal social costs of environmental pollution. Meanwhile, as the world's largest developing country, China faces the arduous task of economic development and lags behind developed countries in terms of clean production technology. For example, according to the Environmental Performance Index jointly released by Yale and Columbia University in 2022, China scored 28.4, ranking 160th globally and relatively low. Innovation is the driving force behind economic and social development, and green innovation, achieved through the green transformation of enterprise technologies, can reduce environmental pollution. It is an environmentally friendly approach to innovation. Using green innovation to improve a company's pollutant emissions is an important means to narrow the gap between China and developed countries in terms of clean production technology and to achieve the carbon neutrality target effectively.
However, China's current level of environmental technological innovation still lags behind that of developed countries. For instance, according to the "Global Analysis Report on Environmental Technology Patent Activity 2008-2017," China only accounts for 5.6% of patents in the field of air pollution prevention and control, which is equivalent to 13% of the United States' patent volume. Moreover, due to the market scale and advantages of nongreen technologies in terms of initial productivity, the adoption of green technologies by enterprises has a minimal impact on their final product returns. Combined with organizational inertia and other factors, it is challenging for enterprises to spontaneously and independently undertake green technological innovation to reduce environmental pollution, leading to an increase in environmental pollution. Among them, incentivizing green innovation in heavily polluting enterprises, which have greater emissions and cause more severe environmental damage compared to general polluting enterprises, without compromising their competitiveness, is a crucial issue that needs to be urgently addressed.Therefore, whether China's environmental protection tax can promote green innovation in heavily polluting enterprises is an important question that determines the success of achieving the dual carbon targets. It is also a key factor in achieving high-quality development by balancing China's economic growth and environmental protection.
This study has the following characteristics: Firstly, starting from the fundamental purpose of environmental protection tax policies, it uses the PSM-DID method to investigate the green technological innovation of heavily polluting enterprises and test the effectiveness of the policies, enriching the research on policy evaluation regarding green technological innovation in heavily polluting enterprises. Secondly, based on the PSM-DID method, the experiment and control group cities are made as similar as possible in various characteristics, eliminating selection biases, in order to more accurately evaluate the effects of environmental protection tax policies on green innovation in heavily polluting enterprises and enrich the research on the effects of environmental protection taxes on green technological innovation in companies. Thirdly, by using micro-level data from listed companies for the period 2003-2019, it verifies the heterogeneous impact of environmental protection taxes on green technological innovation in enterprises, providing reference for government policy adjustments.

Environmental Protection Tax and Green Innovation in Heavy Polluting Enterprises
Kim believes that green innovation is an important means to achieve environmental protection goals and sustainable development goals, and it is the main approach to address the contradiction between economic growth and environmental sustainability [1] .This article also argues that the implementation of environmental taxes is beneficial for businesses to engage in green technological innovation. Based on the "stakeholder theory," the motivation for corporate green innovation may be related to internal and external interests. First, in terms of external interests, investors of publicly traded companies pay more attention to the information disclosed by the company. The information on environmental tax collection is also disclosed, which can influence investors' decision-making. Green technological innovation can enhance investors' valuation of the company and reduce negative expectations arising from environmental pollution issues. At the same time, by reducing pollutant emissions through green technological innovation, companies actively undertake social responsibility and enhance their reputation [2] . Additionally, companies that engage in green technological innovation and produce green products contribute to the promotion of green consumption, aligning with the national call and enhancing the company's social impact. Second, in terms of internal interests, the impact of environmental taxes on corporate green technological innovation is primarily achieved through cost incentives [3] . The imposition of environmental taxes internalizes the external costs of environmental pollution, resulting in increased environmental costs that decrease a company's market competitiveness and affect its resources, funding, profits, economic efficiency, and long-term development [4] . Therefore, companies have a reason to take measures to address the cost issues associated with taxation, and innovation can effectively improve a company's level of competitiveness. Green innovation balances both environmental protection and competitive benefits, making it a long-term beneficial decision for companies. Furthermore, green innovation has spillover effects [5] . The imposition of environmental taxes leads some companies to engage in green technological innovation, which, in turn, stimulates surrounding businesses to pursue green innovation. Based on this, this article proposes the theoretical hypothesis: Theoretical Hypothesis H1: The implementation of environmental tax policies can significantly enhance a company's level of green technological innovation.

Restrictions on corporate financing and green innovation of heavily polluting enterprises
Resources are a significant factor influencing corporate decision-making, and financial capital is an essential resource for businesses. Green technological innovation particularly requires financial support. Scholars such as Zhang Jinsong and Wu have confirmed the substantial impact of implementing green credit policies on enhancing corporate green innovation [6] . When faced with additional taxes resulting from environmental pollution issues, companies with different financial conditions may respond differently. If a company faces higher financial constraints and experiences a shortage of financial cash flow, the imposition of environmental taxes will further exacerbate the company's financial scarcity. Consequently, the company may find it challenging to allocate additional funds for technological innovation and research and development, thus leaning towards decisions to reduce the scale of nongreen products rather than pursuing green innovation. On the other hand, when a company enjoys relatively relaxed financial conditions and strong financing capabilities, it tends to increase its investment in research and development funding and personnel to mitigate the weakening competitive effects caused by rising costs. This tendency is particularly evident in highly competitive markets. Given sufficient cash flow, companies are more likely to allocate greater resources to research and development. Based on these observations, the following theoretical hypothesis is proposed: Theoretical Hypothesis H2: The implementation of environmental taxes has a heterogeneous effect on corporate green innovation. Compared to companies facing financing constraints, companies without financing constraints exhibit more pronounced green technological innovation under the incentive of environmental taxes.

Enterprise ownership, economic development, and green innovation of heavily polluting enterprises
Due to the existence of ownership discrimination, it can be argued that state-owned enterprises have relatively lower financing costs compared to non-state-owned enterprises. State-owned enterprises are typically funded or controlled by the national or local government, which provides them with evident resource advantages. Moreover, the special status of state-owned enterprises in China implies a greater propensity to actively respond to national policies. Additionally, there are significant disparities between Chinese cities in terms of economic foundations, industrial structures, institutional environments, and other aspects, indicating that heterogeneous levels of green technological innovation exist between first-tier and non-first-tier cities [7] . Based on these premises, this paper proposes the following hypotheses: Theoretical Hypothesis H3a: State-owned enterprises exhibit a more pronounced level of green technological innovation compared to non-state-owned enterprises.
Theoretical Hypothesis H3b: Non-first-tier cities demonstrate a more significant level of green technological innovation compared to first-tier cities 3 Model construction and data source processing

Data sources
This study utilizes patent data and financial data of Chinese listed companies from 2003 to 2019. The patent data of listed companies is sourced from the National Intellectual Property Administration of the People's Republic of China, while the financial data of companies is obtained from CSMAR.
For the selection of green patents among listed companies, the determination of green patents is based on the "International Patent Classification Green Inventory" published by the World Intellectual Property Organization (WIPO) in 2010, combined with the seven classification criteria of the United Nations Framework Convention on Climate Change. To examine the innovativeness of different patent types, green patents are further categorized as green utility patents and green invention patents.
Regarding the selection of heavy-polluting enterprises, this study screens companies using data from CSMAR, the "Catalogue of Industry Classification Management for Environmental Check of Listed Companies" established by the Ministry of Environmental Protection in 2008, the "Guidelines for Environmental Information Disclosure of Listed Companies," and the 2012 industry classification by the China Securities Regulatory Commission. Heavypolluting enterprises are chosen as the experimental group, while non-heavy-polluting enterprises are selected as the control group. After merging the collected company data, the following procedures are conducted: (1) deletion of samples with missing variables and samples from the financial industry, (2) removal of samples of companies with "ST" and "*ST" prefixes, and (3) trimming of control variables at the 1% significance level.

3.2.1Model Construction
To examine the impact of the implementation of environmental protection taxes on green innovation in heavily polluting enterprises, this study constructs the following difference-in-differences model:

3.2.2Variable Definitions
(1) Dependent Variable: Green Innovation (Sumpat) Sumpat represents the number of green patents applied by a listed company i in year t, following the indicator construction method of Chen [8] . In subsequent analyses, the natural logarithm of the number of green patent applications (lnsumpat) is used to measure green innovation. Furthermore, the number of green patents is divided into green invention patents (lninvpat) and green utility model patents (lnutipat), both transformed using logarithm. The study examines the differential effects of environmental protection tax policies on invention patents and utility model patents within the domain of green patents.
(2) Independent Variables: Treat * Post Treat represents a dummy variable for heavily polluting enterprises, taking the value of 1 if the enterprise belongs to a heavily polluting industry and 0 otherwise. Post represents a dummy variable for the pre-and postimplementation of environmental protection taxes, taking the value of 1 after the implementation (after 2018) and 0 for pre-implementation years.
(3) Control Variables: X X represents a series of control variables, following the practices in existing literature [9] . The following control variables are selected: Firm Size (lnsize): Firm size has an impact on innovation, with larger firms allocating more funds to green innovation [10] . It is represented by the natural logarithm of total assets. Firm Age (lnage): The developmental stage of a firm is an important factor influencing innovation. It is represented by the natural logarithm of firm age. Debt Level (lnbanklev): Appropriate levels of debt provide firms with sufficient funds for innovation. It is represented by the natural logarithm of the ratio of bank loans to total assets. Capital Intensity (lncapity): Higher capital intensity indicates stronger innovation capability. It is represented by the natural logarithm of the ratio of total assets to operating income.
Return on Assets (lnroa): Firm's financial performance has a significant impact on innovation. It is represented by the natural logarithm of the ratio of net profit to total assets. All control variables are shown in Table 1. The natural logarithm of the firm's age, reflecting the developmental stage of the firm.
Debt Level (lnbanklev) The natural logarithm of the ratio of bank loans to total assets, indicating the level of debt for the firm. Capital Intensity (lncapity) The natural logarithm of the ratio of total assets to operating income, measuring the capital intensity of the firm. Return on Assets (lnroa) The natural logarithm of the ratio of net profit to total assets, capturing the firm's profitability. Table 2 presents the descriptive statistics of the main variables. As shown in Table 2, the average of the natural logarithm of the number of green patent applications by Chinese listed firms is 0.2050. Among them, the value for green invention patents is higher than that for green utility model patents, indicating that in the application of green patents, there is a greater proportion of green invention patents with higher value in terms of green technology transformation and research and development.

4.1.1Propensity Score Matching (PSM) Results
In this study, the propensity score matching method was employed using the nearest neighbor matching approach.
A Logit model was used to estimate the propensity scores, and observable continuous variables from the control variables were utilized as covariates for matching. This was done to examine the significance differences between variables. The matching results are presented in Table 3 below. The standardized biases between the experimental group and the control group were significantly reduced, and the mean differences of the coefficients were not statistically significant, indicating that the matching requirements were met.

4.1.2Results of the Difference-in-Differences Model
Based on the constructed Model (1) and empirical analysis of the impact of the implementation of environmental protection tax policy on green innovation in heavily polluting enterprises, as shown in Table 4, the first column of Table  4 represents the logarithmic form of the total number of green patent applications, the second column represents the logarithmic form of the total number of green invention patent applications in heavily polluting enterprises, and the third column represents the logarithmic form of the total number of green utility model patent applications in heavily polluting enterprises. In columns (1)-(3), this study employs a double fixed-effect model to control for individual and time variations, and robust standard errors are considered in the regression analysis to address potential autocorrelation issues.
The regression results in Table 4 indicate that the environmental protection tax policy has a positive effect on green innovation in heavily polluting enterprises. According to Table 4, the coefficient of the * interaction term in the first column is positive and significant at the 5% level, after controlling for the double fixed effects. Furthermore, analyzing the impact mechanisms of different types of green patents on corporate green innovation, the * coefficient in the second column is significant at the 10% level, while the * coefficient in the third column is significant at the 1% level. This suggests that, under the control of double fixed effects, the positive effect of the environmental protection tax policy on green utility model patents is greater than that on green invention patents. This may be due to the higher innovativeness and difficulty of green invention patents compared to green utility model patents, indicating that the current promotion effect of the environmental protection tax policy on green innovation is still limited. The results of the regression are consistent with the findings of previous studies by Lian Chao et al. [11] . Thus, Hypothesis 1 is confirmed.
Regarding the control variables, the age, size, and indebtedness of enterprises have a certain degree of promotion effect on corporate green innovation, while capital intensity and the ratio of enterprise market value to total capital have a negative impact on green technological innovation. This indicates that the economic characteristics of enterprises do not effectively drive green innovation within the companies, suggesting that external driving forces are more likely to influence corporate green innovation.

4.2.1Heterogeneity Analysis of Ownership Structure
The ownership structure of companies can potentially affect their research and development (R&D) investment and technological innovation. Therefore, based on Model (1), the sample is further divided into state-owned enterprises and non-state-owned enterprises to examine whether there is heterogeneity in the green technological innovation effects of environmental protection tax policy across different ownership types.
The estimation results are presented in Table 5. Columns (1) and (2) show that, after controlling for the control variables, the coefficient of the interaction term for stateowned enterprises is positive and significant at the 5% level, while the coefficient of the interaction term for nonstate-owned enterprises is positive but not significant. This suggests that the environmental protection tax policy has a more pronounced effect on promoting green innovation in state-owned enterprises. One possible explanation is that state-owned enterprises have better resource advantages and government subsidies, making them more conducive to engaging in green innovation activities compared to non-state-owned enterprises.

4.2.2Examining the Heterogeneity of Regional Economic Development
Using the PSM matching model, we investigate the impact of environmental protection tax policy on green innovation in heavily polluting enterprises under the heterogeneity of regional economic development. The results are presented in Table 6. The findings indicate that the implementation of environmental protection tax has a positive and significant effect on green innovation in non-first-tier cities, while it is not significant for enterprises in first-tier cities. This suggests that the implementation of environmental protection tax policy has a more pronounced impact on green innovation in heavily polluting enterprises in non-first-tier cities, promoting their innovation efficiency improvement

4.2.3Examining Heterogeneity in Financing Constraints
The implementation of environmental protection tax policy encompasses innovative policies in green finance, such as green credit and green bonds. For heavily-polluting enterprises that may face significant financing constraints in their green innovation efforts, these green financial policies can alleviate the financial pressure and encourage active engagement in green technological innovation. This study measures financing constraints using the SA index, with the median value as the threshold. A value of 1 represents high financing constraints for enterprises whose SA index exceeds the median, while a value of 0 represents low financing constraints for enterprises whose SA index is below the median. As shown in Table 7, the results indicate that after the implementation of environmental protection tax policy, financing constraints are no longer an impediment for heavily-polluting enterprises in pursuing green innovation. The possible reason is that environmental protection tax, as a nationwide robust policy, exerts a significant positive driving force on green innovation in heavily-polluting enterprises.

Placebo Test: Early Implementation of the Policy
By advancing the implementation time of the environmental protection tax policy to 2013 and 2014, the impact of the environmental protection tax on green innovation in heavily polluting enterprises was analyzed. The results are shown in Table 8. The estimated coefficients of the interaction terms in columns (1) and (2) are not significant, indicating that advancing the implementation time of the environmental protection tax does not have a significant effect on green innovation in heavily polluting enterprises. This confirms the robustness of the baseline regression.

Lag analysis
Considering that the formation of corporate green innovation requires some time, the effect of environmental protection tax policy may have a certain lag. By lagging the number of green patent applications by one period and two periods, the impact of environmental protection tax policy on green innovation of heavily polluting enterprises was examined. The results, as shown in Table 9, indicate that the coefficients of the double-difference terms are positive and significant at the 10% level. This is consistent with the main conclusions stated earlier. Based on data from Chinese listed companies from 2003 to 2019, this study examined the impact and mechanisms of environmental protection tax on firm's green innovation and obtained the following conclusions: Environmental protection tax policy has a positive and significant impact on green innovation in heavily polluting enterprises, contributing to their improvement in green innovation.
The effect of environmental protection tax policy on different types of green innovation in heavily polluting enterprises is heterogeneous. The policy has a higher significant positive effect on green utility patents compared to green invention patents.
There is heterogeneity in the impact of environmental protection tax on green innovation among different types of ownership in heavily polluting enterprises. The positive effect of the policy on green innovation is mainly observed in state-owned enterprises, which have better access to research and development funding. There is regional heterogeneity in the impact of environmental protection tax on green innovation in heavily polluting enterprises. The policy has a significant positive effect on green innovation in non-first-tier cities. This may be because non-first-tier cities have greater room for improvement in green innovation, and the policy has a larger impact on them.

Policy Recommendations
Continuously optimize the environmental protection tax system and promote green development. As stated in the 19th National Congress of the Communist Party of China, it is important to accelerate ecological civilization system reform, build a beautiful China, and establish a marketoriented green technology innovation system. The findings of this study suggest that the environmental protection tax policy plays a role in promoting firm's green innovation, aligning with the goal of promoting economic development through green innovation. Therefore, it is necessary for China to persist in implementing and improving the environmental protection tax policy, ensuring its effective implementation and positive impact on firm's green innovation.
The transformation of heavily polluting enterprises towards ultra-low emissions should be a key focus of implementing the environmental protection tax. This study's heterogeneity analysis indicates that the environmental protection tax policy is more effective in promoting green innovation in heavily polluting enterprises compared to non-polluting industries. However, the promotion of green invention patents needs to be enhanced, as the current focus is mainly on green utility patents. While green utility patents contribute to short-term green innovation in heavily polluting enterprises, green invention patents with higher value are crucial for long-term promotion of green innovation. Therefore, companies should formulate green innovation strategies based on their own structure and current situation. Government departments should consider the heterogeneity of firm attributes and regional economic development, as the environmental protection tax policy has a significant positive effect on non-first-tier cities. The government should implement differentiated environmental tax policies tailored to different regions and implement specific measures to stimulate firm's green innovation.
Compared to state-owned enterprises, the role of environmental protection tax policy is weaker in non-stateowned enterprises. The government should develop targeted policies to enhance green innovation in non-stateowned enterprises, such as providing more tax incentives, in order to fully leverage the positive effects of the environmental protection tax.
This study examined the impact of environmental protection tax policy on green innovation in heavily polluting enterprises from different perspectives. However, limitations exist due to data limitations and the relatively short implementation period of the policy. With the improvement of the environmental protection tax policy and the availability of more enterprise data, future research can delve deeper into the policy's effects.