The Impact of Water Resource Tax on the Sustainable Development in Water-Intensive Industries: Evidence from Listed Companies

Abstract

China is taking measures to minimize the negative impact of the long-term extensive water use model on the water environment. The large number of zombie enterprises with high energy consumption and low energy efficiency in highly water-consuming industries is one of the important reasons for the water resource governance plight of China. Based on the quasi-natural experiment of water resource tax reform from 2016 to 2020, this paper uses the panel data of listed companies to quantify the impact of water resource tax on the zombification of high water consumption enterprises in this paper. The results show that the zombie degree of high water consumption enterprises after the implementation of the water resource fee to tax reform has become significantly higher, and this conclusion remains stable after a series of tests. The conclusion of this paper has major implications regarding sustainably developing zombie enterprises and highly water-consuming industries in terms of policy.

1. Introduction

Since the reform and opening up, industries in China have long been relying on the resource-factor-driven production method of high input, high consumption, and high pollution to rapidly accumulate output value and promote high economic growth under the strategic orientation of prioritizing the development of heavy industry. Local government officials give excessive support to water-consuming industries for promoting economic growth. Excessive capital allocation to high water consumption in steel and other industries result in the long-term high load on water resource carrying capacity [1]. At the same time, such extensive water consumption makes it difficult to turn the resource-intensive production mode into a technology-intensive one since some enterprises are inclined to use existing capital and technology for production without innovation, making them prone to become zombies with low efficiency, high debt, and continuous loss under the pressure imposed by the market and environment, giving rise to market competition imbalance and overcapacity and further exacerbating water resource scarcity and water environment pollution. Water resources are not the only basic natural resource that supports economic and social development; it is also a strategic economic resource for human survival. Rough water use is not conducive to the construction of ecological civilization or promotion of the sustainable development of the environment and the economy, and highly water-consuming industries should change their inefficient methods of using water and improve energy saving and reduce energy consumption, thereby reducing the pressure on water resources.

Energy development must follow a sustainable path in order to reconcile economic growth, social development, and environmental protection [2]. Over the past few decades, China’s economic growth has been among the highest in the world [3]. And China’s policies on environmental protection and sustainable development are of great significance to both China and the world [4]. In order to strengthen the construction of ecological civilization and accelerate the modernization of harmonious coexistence between human beings and nature, the central government has put forward the guideline that “green mountains and green water are mountains of gold and silver”, and pointed out that a good ecology contains infinite economic value, can constantly create comprehensive benefits, and achieve sustainable economic and social development. In this context, various regions and departments have been vigorously promoting water conservation, and have made remarkable achievements.

However, under the GDP-centered promotion model, it is still difficult to alleviate the problem of “over-allocation of capital to high-input, high-consumption, high-pollution enterprises” through the government’s resource control alone. In order to resolve the mismatch between China’s high water-consumption industrial water use and the country’s water resource stock and environmental carrying capacity, major funds and efforts should be focused on water conservation technology development, eliminating outdated production capacity in the industry as soon as possible, and taking the path of sustainable development. The central government has gradually adopted a market-based and government-supported intervention strategy, in which the pilot water resource tax reform relies on tax leverage to give full play to the decisive role of the market mechanism in resource allocation, and has attracted extensive attention from the academic community.

Water resources have significant impact on environmental issues [5], and it is therefore important for governments to place certain restrictions on their use. Before carrying out water resource tax reform, the main economic leverage for getting water in China was water resource fees, whose collection, use, and management were supervised by the higher water administration departments, and the powers and responsibilities of local water departments and governments were not clearly delineated. As a result, the imposition of water resource fee faces such problems as low collection standards, imperfect collection mechanism, and irregular management, which imposed limited restriction on highly water-consuming fields, did not reflect the scarcity of water resources, and made it difficult to effectively regulate the water supply and demand market.

Water agencies in some areas supply water to industrial enterprises at subsidized prices, resulting in extravagance and wasted water by some enterprises, making it difficult to ensure water consumption and water quality, and leading to lower production efficiency and product quality, and even a reduction in production. However, rushing to close down these enterprises will give rise to many economic and social problems, and the government will choose to postpone the disposition by using subsidies and tax incentives for extending the life of these enterprises. This will make some enterprises become dependent and turn to readily available subsidies for maintaining their operations, thus forming a vicious circle of “getting into trouble with business—taking in subsidies—being short of momentum for business—getting into trouble with business”. Ultimately, a large number of zombie enterprises appear in highly water-consuming industries. The five industries with the largest proportion of zombie enterprises are: water production and supply (25.99%), electricity and heat production and supply (19.14%), chemical fiber manufacturing (19.10%), ferrous metal smelting and rolling processing (15.00%), and petroleum processing, coking, and nuclear fuel processing (14.46%) [6]).

Alleviating structural overcapacity is the key to realizing the transformation of the resource-driven model to innovation-driven economic development [7]. This problem can be solved by strengthening the strict constraints on water resources and curbing the disorderly expansion of enterprises with high water consumption. Water resource tax guides help enterprises optimize their own water resource utilization and improve the efficiency of water resource utilization. At the same time, in this paper, we will clarify the environmental interests of local governments. The regulation causing the water resource income to be directly included in the local disposable income can improve the enthusiasm of local officials in environmental governance. That is, the water resource tax has internalized the environmental costs, changed the local government decision-making trade-offs that mainly focus on GDP growth, and shifted to the environment and the economy in a two-handed manner. Therefore, a water resource tax can break the ice of industrial growth linked to water resource abundance and zombie enterprises relying on water resources to achieve inefficient growth, realize rapid and active growth of the industry, and gradually form a benign and sustainable development model.

At present, the literature on water resource tax reform mainly focuses on the following aspects: interpreting policies [5] and evaluating the effectiveness and impact of pilot projects [8,9,10,11]. A water resource tax can help alleviate water poverty, but the effect varies across industries and regions [12,13]. After the implementation of the water resource tax, the use and protection of water resources are in a relatively stable state, achieving the dual objectives of sustainable economic development and sustainable use of water resources [14]. In addition, the water resource tax will also have some negative effects [9,15], even while effectively reducing water consumption and improving water use efficiency. Most of the above studies focused on the policy itself and its subsequent impact, but did not regard it as a solution to the problem of China’s water environment and zombie enterprises in highly water-consuming industries, nor did they explore the mechanism.

The impact of the water resource tax on the zombification of highly water-consuming enterprises will be analyzed in this paper from the perspective of the living environment for Chinese enterprises and in combination with the role played by their water consumption. When calculating the degree of enterprise zombification, this paper selects the relevant variables (the characteristic indexes of zombie enterprises) that may deepen the degree of enterprise zombie transformation, and includes them in the binary Logit model to estimate the coefficient of each variable in the enterprise zombification index. This step makes the zombie enterprise no longer a simple 0–1 variable, but a continuous variable related to a series of zombie enterprise characteristics. Compared with virtual variables, the degree of enterprise zombification contains more information. Unlike the existing research, which mostly focuses on whether the enterprise is a zombie enterprise or not, this paper selects the relevant characteristic variables to measure the zombification index of the enterprise, which makes the research more relevant to the real-world situations. This is one of the main contributions and innovations of this paper.

This paper may make marginal contributions as follows: ① This paper combs through the causes of zombie enterprises in highly water-consuming industries, illuminates the relationship between the causes of zombie enterprises and the environment, and provides new research perspectives and entry points for articles studying zombie enterprises. ② This paper enriches the literature in related fields by analyzing the effect of water resource tax on the zombification of highly water-consuming enterprises from the correlation between their ways of using water and the causes of zombie enterprises. ③ Different from previous studies that distinguish zombie enterprises only by dummy variables, this paper brings enterprise zombification indexes in to fully measure the factors that impact the zombification of enterprises, describes relevant features in detail, and is more rigorous and practically significant compared with 0–1 categorical variables based on the quasi-natural experiment of water resource tax reform.

Other parts of the paper are arranged as follows: The institutional background and theoretical hypothesis of this paper are introduced in the second part; The data source, research design, measurement model, and index measurement method are introduced in the third part. The empirical results are presented and analyzed; robust tests, endogeneity test, and heterogeneity analysis are carried out in the fourth part. The mechanism is tested in the fifth part. The last part is the conclusion and policy proposals. At the end of the paper, the results are discussed and analyzed.

2. Theoretical Hypothesis

The change of environmental protection fees to taxes will send a stronger signal to polluting enterprises to reduce pollution and lower productivity. At the same time, the environmental protection fee to tax will further internalize the social cost of polluting enterprises’ emissions, i.e., the environmental protection fee to tax, by raising its environmental pollution product input cost of factors to force enterprises to participate in environmental governance [16]. Polluting enterprises are more strictly investigated and punished by environmental departments or tax authorities of local governments [17], forcing polluting enterprises to take responsibility for environmental governance. However, if the design of the environmental tax system is unreasonable, it will be difficult for the introduction of environmental tax to play a positive role [18,19], and it may even seriously hinder the production and operation of enterprises when the environmental tax rate is higher or lower than the marginal pollution control cost of polluting enterprises.

Water resource tax can reduce the dependence of water-consuming enterprises on resource factors and encourage them to develop patents for green inventions and improve their environmental performance [20]. A water resource tax is a regulatory tax based on market mechanism and economic incentives, and a differentiated tax on different industries and regions. The external cost of taxation closer to the target of taxation will have an impact on highly water-consuming industries with a large number of zombie companies. Water resources are uniformly supplied by the state. After the imposition of a water resource tax, the enterprises’ cost of water resources will rise, and the operating cost of enterprises in different industries will also rise to different degrees, which will have two effects on corporate taxpayers. First of all, some enterprises that are highly conscious of innovation will expand sources for profits by improving mining efficiency and resource utilization rate, or will reduce the use of water resources via technological innovation and improving the mode of production to minimize the negative impact of tax burden rising and continue to survive and develop when it is impossible to increase profits by expanding resource investment. Secondly, some enterprises that are not conscious enough of innovation or lack of production capital can hardly realize technological upgrades but can continue their production by investing extensive resources. And it is difficult for them to improve resource utilization and production efficiency. Such enterprises will gradually find it difficult to bear the high cost of water resource use brought by the water resource tax, and could only develop slowly until they withdraw from the market; thus, the mechanism of natural selection can be established for the sake of phasing out the outdated production capacity of the industry.

Therefore, water resource tax can make it more difficult for highly water-consuming enterprises to survive, and further their zombification. On the one hand, a water resource tax would increase the water cost of highly water-consuming enterprises, reduce their profit margins, add to their operating pressure, force them to reduce water consumption, and phase out outdated production capacity in the industry by stimulating their production activeness, carrying out water-saving technological innovation, furthering their zombification, and accelerating their withdrawal from the market. Enterprises that have been in a state of loss for a long time and rely on the support of the government or financial institutions often lack the motivation and ability for technological innovation and management improvement, and it is difficult for them to improve water efficiency and reduce water costs. As a result, a water resource tax would increase the operational difficulties of these enterprises, making it more difficult for them to rise out of zombification. Therefore, the fee-to-tax reform on water resources can fundamentally change the water consumption behaviors of enterprises, transform the growth mode from resource-based to technology-based, and completely put an end to the extensive industrial development in China.

On the other hand, a water resource tax can change local governments’ development goals with GDP as the core, and change their supportive attitudes to highly water-consuming enterprises, furthering their zombification from the perspective of local government’s policy orientation. For a long time, local governments often give such preferences as financial subsidies, credit guarantees, and land supply to zombie enterprises in highly water-consuming industries for GDP growth and maintaining social stability. These measures not only caused the loss of local fiscal revenue, but also delayed the withdrawal of zombie enterprises and the resolution of overcapacity. After the reform, water resource taxes are levied by tax authorities instead of local water conservancy administrative agencies, which can effectively prevent local governments from blindly interfering in the collection of water resource fees for GDP. At the same time, the tax reform on water resources will also affect the indicators for appraising the performance of local governments, making them move from simply pursuing GDP growth to comprehensively considering economic, social, and environmental factors. In this way, the goals of local governments and zombie enterprises can be separated. The incentives for local governments to provide assistance can be minimized. The subsidies for zombie enterprises can be cut off. The institutional environment for the development of zombie enterprises can be worsened. The zombification degree of zombie enterprises can be furthered, and the zombie enterprises can withdraw from the market more quickly.

The adjustment of the water resource tax system mainly aims at enterprises, replaces the previous water resource fee, and links the water tax rate with the region and usage. A water resource tax forces highly water-consuming enterprises to reduce water consumption and phases out outdated production capacity in the industry by stimulating their production activeness, carrying out water-saving technological innovation, furthering their zombification, and accelerating their withdrawal from the market. A large number of studies currently available have explored how water resource tax reform forces enterprises to actively transform production processes based on the double dividend hypothesis. However, the imposition of a water resource tax will have a negative economic impact on some highly water-consuming enterprises, especially those enterprises that fail to effectively cope with tax increases and cannot get out of the trouble by themselves. For them, the increase in tax and the reduction of subsidies may lead to the deterioration of a series of indicators, such as their cash flow and operating state, further their zombification, and thus accelerate the withdrawal of zombie enterprises from the market. The policy effect of water resource tax reform on the zombification degree of zombie enterprises with high water consumption is explored by taking highly water-consuming industries with a high proportion of zombie enterprises as the research objects to explore the practical impact of water resource tax reform on zombie enterprises with low operational performance.

First of all, highly water-consuming enterprises tend to use water intensively, and a water resource tax will directly increase their production costs. For zombie companies in highly water-consuming industries, water resource tax will make them harder to survive, and force them to innovate technologically or withdraw gradually. Water resource tax extends water resource fees in the scope of taxation, and increases the water consumption cost and production cost of enterprises through “fee-to-tax conversion”. With heavier tax burden, enterprises will be confronted with such financial difficulties as profit falling, investment momentum waning, and capital adjustment difficulty, and this will further cut down the operating performance of these enterprises and further their zombification. The tax reform on water resources will directly affect the profits of highly water-consuming enterprises. For those highly water-consuming enterprises with low production efficiency and added value of products, the increase in taxes will further bring down their profits, and the enterprises may face problems such as fund shortage and reduced investment momentum. Operational difficulties make it difficult for enterprises to innovate technologically, and further limit the ability of enterprises to improve the efficiency of water use. Highly water-consuming enterprises may also get their burden of tax increase relieved by adjusting their production modes. However, such adjustment requires a significant investment transformation cost and requires enterprises to have relatively rich capital reserves. Enterprises that cannot afford these costs can only respond to the policy by lowering their operational efficiency and cutting back their production and other means, thus exacerbating their zombification.

And beyond that, water taxation may further exacerbate the zombification of water-intensive firms by increasing environmental uncertainty. Environmental uncertainty arises when managers perceive their business environment or one of its components to be unpredictable [21]. Managers may be uncertain about the direction of future technology, changing consumer preferences and social norms, or the impact of changing regulations on operations [22]. Firms always carry out their manufacturing activities in a specific market environment, and the market situation has a significant impact on their business performance and risk. And unpredictability may encourage firms to invest in introducing deep changes [23,24].

Increased environmental uncertainty will raise the stake of strategic failure, making it difficult for enterprises to calculate the costs and probabilities associated with various strategic options. In a good market situation, the enterprises facing low environmental uncertainty can obtain data about the change in the market situation over time and make the right decision to ensure performance stability. However, when the market environment is fluctuating greatly, it is more difficult for enterprises to make predictions, and the environment becomes more uncertain accordingly. Information asymmetry is not conducive for enterprises to making the best choice, and causes the enterprises to make production decisions via passive reaction instead of active prediction, and struggle to cope with the situation.

As the levying standard of water resource tax is affected by environmental conditions, economic situations, and other external factors, it is difficult for enterprises to accurately predict the adjustment of water resource tax policy, which may make enterprises conservative in future business decision-making and inhibit their investment and development momentum. Besides the policy itself, a water resource tax may also affect the zombification of highly water-consuming enterprises and environmental uncertainty through market demand, environmental pressure, industry competition. Water resource tax will lead to fluctuations in the price and market demand of highly water-consuming products, and market uncertainty will lead to greater competitive pressure, which will make it difficult for highly water-consuming enterprises to predict the market reaction, and thus increase their production uncertainty. The intensified competition between enterprises will increase the uncertainty of enterprises, especially those with overcapacity and serious product homogeneity, because they may face greater competition for market share and profit squeeze. Moreover, water resource tax reflects the attention of the government to water conservation, and more environmental protection policies may be unveiled in the future. Enterprises with high water consumption may face more stringent environmental standards and restrictions, which may impose greater environmental protection pressure on enterprises, and increase the difficulty and uncertainty of their transformation and upgrade.

It is also possible that the water tax, while promoting some high-consumption zombie companies to withdraw from the market, may also lead some low-consumption zombie companies to survive, and the number of these companies will even increase. This is because the water tax may have reduced the total supply of water resources, thus raising the price of water resources, allowing some zombie companies with low water consumption to maintain revenue by raising the price of their products without the need to improve productivity or innovation. Another possibility is that the water tax, while promoting some zombie companies to exit the market, may also hinder the development of some potential enterprises, thus reducing the competitiveness and vitality of the market. This is because the water resource tax may increase the cost and burden of enterprises, making it difficult for some potential enterprises to obtain sufficient capital and credit support, or to bear the fluctuation risk of the water resource tax, thus affecting their investment and innovation decisions. However, due to the limited data acquisition, the data of the above enterprises cannot be obtained in this paper, so the effect of the water resource fee to tax reform policy estimated in this paper may be small compared with the actual effect.

Based on the above analysis, the following hypotheses are proposed in this paper:

Hypothesis 1 (H1).

The imposition of water resource tax would further the zombification of highly water-consuming enterprises.

Hypothesis 2 (H2).

The water resource tax deepens the zombification of highly water-consuming enterprises by increasing their financial pressure and the environmental uncertainty of the enterprises where they are located.

3. Research Design

3.1. Data Source and Sample Selection

The data of A-share listed companies from 2011 to 2021 are used in this paper. Most listed company and industry related variables are derived from CSMAR, and some are from the annual reports of listed companies; Identification data of zombie enterprise come from database Wind; “Business Code—Year” match was performed when combining the data, and businesses missing a large number of variables were eliminated.

3.2. Baseline Demonstration Model

The double difference method (DID) is selected as the benchmark model in this paper. However, the dual difference method does have some limitations. For example, it may ignore the heterogeneity between the experimental and control groups, such as the differences in treatment intensity and treatment effects. This heterogeneity may make the estimated coefficient of the double difference method not a true average treatment effect, but a weighted average or specific subgroup treatment effect, which may make the regression results disturbed by other variables or events, such as common trends, selective bias, and policy endogeneity. These disturbances may make the identification strategy of the double difference method ineffective or inaccurate and require diagnostic testing or correction. In the latter endogeneity test, we tested the benchmark regression results.

As a statistical method used to assess the policy effect, the net effect of the policy was estimated by comparing the difference between the “experimental group” and the “control group” before and after the policy implementation. The pilot reform of water resource tax is a natural experiment that changes some provinces from collecting water resource fees to collecting water resource tax in order to promote the conservation and protection of water resources. This policy change meets the applicable conditions of the double difference method, that is, there is clear time and area of policy implementation, and the samples can be divided into the experimental group and the control group; the experimental group has the same or similar development trend before the policy implementation, that is, the parallel trend assumption is established; this also ensures that the change from the policy is mainly affected by the policy, rather than the interference of other external factors.

Combined with the discussion of the institutional background, China’s water resource tax reform is taken as a quasi-natural experiment to build the DID model and empirically analyze the impact of water resource tax reform on enterprise zombification in this paper:

$${zr}_{it}={\beta }_0+{\beta }_1{DID}_{it}+{\beta }_2{Control}_{it}+{\mu }_{i }+{Year}_t+{\epsilon }_{it}$$

(1)

In Formula (1), ${zr}_{it}$ is the explained variable, degree of enterprise zombification; ${DID}_{it}$ is the core explanatory variable of this paper, double differences; ${Control}_{it}$ stands for a series of control variables at the enterprise level; ${\mu }_i$ is the entity fixed effect; ${Year}_t$ is the year fixed effect; ${\epsilon }_{it}$ is a random error term.

Considering that all enterprises are the target of water resource tax, regression for fully sampled enterprises will also be carried out in this paper to compare the possible differences in policy effects between highly water-consuming enterprises and non-highly water-consuming enterprises.

3.3. Explained Variable

Most of the existing studies only focus on whether enterprises are zombie enterprises, and little attention has been paid to the degree of zombification. Zombification index is introduced as the explained variable in this paper to accurately measure the stage of the enterprise in the process. Corporate zombification refers to the degree to which an enterprise can maintain its operation despite experiencing problems due to low efficiency, loss, and lack of competitiveness. Enterprises in the process of zombification usually have such features as suffering long-run loss, unhealthy balance sheets, and inactive operation. And the degree can be measured by a series of indexes and financial data including, but not limited to, corporate profitability, debt ratio, cash flow status, and government subsidies.

When calculating the degree of enterprise zombification, this paper selects relevant variables (characteristic indicators of zombie enterprises) that may deepen the degree of enterprise zombie transformation and includes them in the binary Logit model, then estimates the enterprise zombification index. The specific method of estimation is as follows: find the corresponding weight coefficient by including the characteristic variable in Formula (2) for regression according to the properties of the zombie enterprises’ continuous operating loss, insolvency, and survival on “evergreen loan” subsidies, and add the dependent coefficients together as the enterprise’s zombification index. This paper does not limit the enterprise’s zombification index to 0–1 for measuring the zombified state of enterprises more accurately. The larger the variable, the higher the degree of corporate zombification.

$${Zombie}_{it}={\beta }_0+{\beta }_1{gap}_{it}+{\beta }_2{perform}_{it}+{ \beta }_3{ROI}_{it}+{\beta }_4{leverage}_{it}+{\beta }_5{tfp}_{it }+{\epsilon }_{it}$$

(2)

In Formula (2), dependent variable ${Zombie}_{it}$ is the dummy variable of zombie enterprise. In this paper, the “Standard” proposed and improved on CHK method [25] by the National Academy of Development and Strategy, RUC is adopted: if an enterprise is identified as a zombie enterprise both in a certain year and the year before that year by FN-CHK method [26], it should be identified as a zombie enterprise in that year; if the enterprise is identified as a zombie enterprise, the value is 1; otherwise, the value is 0. The FN-CHK method can identify zombie companies as follows: First, calculate the minimum interest the company has to pay under normal operation. Second, estimate the company’s interest income. Finally, calculate and divide the interest difference (interest expense minus interest income) by the loan amount. If the interest difference is less than 0, then the business is marked as a zombie business. The explained variable ${dr}_{it}$ (zombification degree) is the sum of coefficients ${\beta }_0~{\beta }_5$ in Formula (2) in this paper; ${gap}_{it}$ is the government subsidy received by the enterprise in the year for measuring the degree of “blood transfusion” obtained by the enterprise; ${perform}_{it}$ is the output value growth rate of enterprise sales used to measure the enterprise’s production and operation performance; ${ROI}_{it }$ is the return on investment of enterprises used to measure the investment and financing behavior of enterprises; ${leverage}_{it}$ is the corporate leverage ratio used to evaluate whether the enterprise has the solvency to avoid problems with “evergreen loans”; ${tfp}_{it}$, the total factor productivity of an enterprise, is used to evaluate whether an enterprise has the ability to overcome short-term business difficulties and turn losses into profits. TFP reflects the essence of productivity as an economic concept [27]. In this paper, OP [28], LP [29], and the OLS methods are adopted to calculate the total factor productivity of an enterprise. The OP method estimates TFP as follows: first, the inverse function of the investment function is used to express productivity, then the non-parametric method is used to estimate the labor coefficient and the productivity function in the production function. The second step constructs the first-order Markov process of productivity, using the least squares method to estimate the capital coefficient in the production function. Finally, the labor coefficient and productivity function in the production function are reestima-ted to obtain the TFP. The LP estimation of TFP is determined by using the inverse function of the demand function of the intermediate input and the same OP method. For the OLS method, first, take the Cobb–Douglas production function logarithmically to find the linear function. Next, estimate the values of each parameter in the linear function. Finally, the random error term of the linear function obtained in the first step, which is TFP, is used. ${\epsilon }_{it}$ is the random error term.

3.4. Core Explanatory Variable

The DID construction of this paper is set according to the time implemented in the water resources “fee to tax” policy document. The reform process of the water resource tax is as follows: (1) In 2016, Hebei Province, as the first pilot area, implemented the water resource tax within the province. (2) In 2017, China expanded the scope of the pilot program to include nine provinces (autonomous regions, municipalities): Beijing, Tianjin, Shanxi, Inner Mongolia, Shandong, Henan, Sichuan, Shaanxi and Ningxia, as pilot areas to carry out water resources tax reform in these nine provinces. (3) In 2020, after the Water Resources Tax Law of the People’s Republic of China was officially issued, the water resources tax was written into law as a national policy, and the remaining provinces and regions in the country began to implement the water resources tax.

In this paper, the core explanatory variable is the DID, which is 1 for enterprises located in Hebei Province in 2016 and later, the remaining nine pilot areas in 2017 and later, and all other areas in China after the official introduction of the Law on Water Resource Tax of the People’s Republic of China in 2020. While evaluating the impact of water resource tax on highly water-consuming enterprises, references are made to the standards by the Ministry of Water Resources of the People’s Republic of China (2013) in this paper. And eight categories of industries are defined as highly water-consuming industries, including the utility, chemical, steel, non-metallic mineral products (mainly coal), petroleum and petrochemical, food, paper, and textile industries. According to the industry classification standard of the National Bureau of Statistics, the value is 1 for enterprises belonging to the above highly water-consuming industries and 0 for enterprises not belonging to the above highly water-consuming industries.

3.5. Core Explanatory Variable

After including the above variables into the calculation of the corporate zombification degree, the following enterprise-level control variables that may affect the corporate zombification degree are selected: TTM (age), the natural logarithm of the enterprise age plus one; Enterprise size, the natural logarithm of the enterprise’s total assets at the end of the year; Tobin’s Q value, used to measure the growth and cash arbitrage motivation of listed enterprises; R & D Personnel Ratio (RDPR), measured by dividing the number of R & D personnel by the number of employees.

3.6. Descriptive Statistics

In this article, the sample size, mean, standard deviation, minimum and maximum variables, and descriptive statistics of the observed variables are shown in

Table 1. Descriptive statistics.

	(1)	(2)	(3)	(4)	(5)
Variable Name	N	Mean	sd	Min	Max
profitability	31,213	0.0328	0.658	−30.69	108.4
leverage	31,214	3.185	48.65	−4.970	6270
return on total assets	31,213	0.0494	0.657	−29.02	108.4
total debt ratio	31,213	0.447	0.367	−0.195	31.47
bank loan	31,214	0.233	0.234	0	1.131
return on assets	31,180	3.574	348.7	−28.59	59,412
zrop	31,214	0.0601	0.0513	−0.247	3.360
zrlp	31,214	0.0600	0.0538	−0.251	3.571
zrols	31,214	0.0599	0.0541	−0.317	3.529
percentage of R&D staff	31,214	10.08	13.00	0	94.49
Tobin’s Q value	30,563	2.320	11.35	0.609	1753
enterprise size	31,213	22.16	1.342	13.76	28.64
age of business	31,214	2.183	0.760	0.693	3.466
Number of id	3861	3861	3861	3861	3861

. The values in the table have been reduced by 1%.

The sample time span used in this paper is 2010–2021, and the enterprise zombification index calculated by using the total factor productivity of the three methods are named zrop, zrlp and zrols, respectively. The mean difference of the three corporate zombification indexes is small, and the standard deviation, maximum value and minimum value of zrolls are the largest among the three calculation methods. Therefore, the regression results of zrolls may have a large error compared with the other two zombification indices.

All other variables in the table are control variables, and the control variables were used to calculate the enterprise zombification index. In total, 10 to 14 behavior the control variables were used in the benchmark regression model. The profit margin on total assets (return) and return on assets will be replaced in the robustness test later. Leverage ratio refers to the ratio of equity capital to the total assets in the balance sheet. It is an indicator to measure the risk of a company’s liabilities and reflects the company’s repayment ability from the side. Bank loan refers to the monetary funds lent by a financial institution and approved by the banking supervision institution of The State Council, with the public as the service object and the repayment of principal and interest. The amount of a bank loan is an important indicator to measure whether an enterprise is a zombie enterprise. The total debt ratio refers to the ratio of corporate debt in the total assets. Profit margin refers to the ratio of corporate profits in the total assets of the current year. The percentage of R & D personnel refers to the ratio of the number of scientific and technological personnel to the total number of employees. Scientific and technological personnel refer to the personnel who are directly engaged in research and development and related technological innovation activities. The Tobin’s Q value is one of the items that affect monetary policy and is defined as the ratio of the market value of an asset to its replacement value.

4. Empirical Result Analysis

4.1. Benchmark Regression Results

The core research of this paper is the impact of water resource tax reform on the zombification degree of highly water-consuming enterprises.

Table 2. Baseline regression results.

	(1)	(2)
	Highly Water-Consuming Industries	Full Sample
DID	0.0044 ***	0.0007
	(0.0010)	(0.0009)
Treat	0.0028 *
	(0.0017)
Post1	−0.0316 ***	−0.0311 ***
	(0.0016)	(0.0018)
age1	0.0123 ***	0.0123 ***
	(0.0010)	(0.0010)
size	−0.0005	−0.0007
	(0.0005)	(0.0005)
Tobin’s Q value	−0.0000 *	−0.0000 *
	(0.0000)	(0.0000)
RDPersonRatio	−0.0001 ***	−0.0001 ***
	(0.0000)	(0.0000)
Constant	0.0623 ***	0.0671 ***
	(0.0098)	(0.0097)
Observations	18,190	30,563
R-squared	0.0332	0.0272
Number of id	3614	3898
Year FE	YES	YES
Individual FE	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, * p < 0.1.

lists the empirical results of baseline regression in this paper. The results show that the introduction of water tax significantly increases the zombification index of highly water-consuming firms, while the change in the zombification index of the sample as a whole is not significant. One possible reason is that, after the imposition of water resource tax, non-highly water-consuming enterprises can adapt to the changes brought by the policy faster than highly water-consuming enterprises. Non-highly waterconsuming enterprises have relatively low demand for water resources, and are more likely to implement water conservation measures during production. Column (2) of Table 2 presents the results of the regression on the degree of zombification for non-highly water-consuming industries. The coefficient of this result is positive but not significant. It shows that, after the implementation of the water resource fee to tax, the zombification index of enterprises in non-high water consumption industries does not change significantly. One possible reason is that the enterprises in non-highly water-consuming industries are more adaptable to the policy and the increase in financial pressure is small, so the water resources tax reform does not have a significant impact on their degree of zombification. In contrast, the production of highly water-consuming enterprises is highly dependent on water resources, and the tax increase may impose greater financial pressure on them. Moreover, due to the constraints of production pattern, transformation is more difficult for some highly water-consuming enterprises, and they cannot respond to the impact of tax increase in a timely manner, leading to their further zombification. Evidence has been found to support Hypothesis 1.

4.2. Parallel Trend Test

Formula (3) as follow is constructed for parallel trend test in this paper.

$${zr}_{it}= \sum \beta \times {Zombie}_{it} +\lambda \sum X_{it} + { \mu }_{i }+{Year}_t +{\epsilon }_{it}$$

(3)

In Formula (3), ${zr}_{it}$ is the explained variable, degree of enterprise zombification; subscript t represents the t^th year when water resource tax is implemented; current is the dummy variable in the pilot year of water resource tax in Hebei Province, ${Zombie}_{it}$ is the dummy variable in the t^th year before the imposition; and ${Zombie}_{it}$ is the dummy variable in the t^th year after the imposition. Other variables have the same meanings as in Formula (2). The first pilot imposition year of water resource tax is chosen as the base period for parallel trend test in this paper, and this part focuses on coefficient $\beta$.

Figure 1 shows the results of parallel trend tests based on event study method. The hollow points in the figure represent the current year coefficients, and the vertical lines represent 95% confidence intervals.

As Figure 1 shows, in the year before the water resource tax pilot collection, the confidence interval of the coefficient is communicated with the x-axis. The estimated coefficient in the years before the implementation of the policy did not pass the significance test, indicating that the effect of the policy did not show before the water resource tax reform. The parallel trend assumption of this paper holds.

From the beginning of the year of implementation, the confidence interval of the coefficient no longer crosses the x axis, and all the estimated coefficients after the collection of water resource tax are significant, indicating that the degree of zombie improvement of enterprises is indeed caused by the reform of water resource tax.

4.3. Robustness Test

The robustness will be tested from the aspects of changing the calculation method of explained variables, making up policy introduction time, and using a balanced panel in this paper.

4.3.1. Changing the Calculation Method of Total Factor Productivity

In this paper, the OP method is adopted to calculate the total factor productivity in the baseline regression; the LP and OLS methods are used to make comparisons in the robustness test. The explained variable, the degree of enterprise zombification, is recalculated in this paper upon the change of the calculation method for total factor productivity. The results obtained are shown in

Table 3. Replacement of Total Factor Productivity (TFP) calculation methodology.

	(1)	(2)	(3)
	zrop	zrlp	zrols
DID	0.0044 ***	0.0046 ***	0.0038 **
	(0.0016)	(0.0017)	(0.0017)
Observations	18,190	18,190	18,190
R-squared	0.0332	0.0305	0.0343
Number of id	3614	3614	3614
Year FE	YES	YES	YES
Individual FE	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, ** p < 0.05.

, and the basic conclusions of this paper remain robust.

4.3.2. Changing the Return on Assets

In the baseline regression, the return on total assets of listed companies is calculated by using the formula (net profit/average total assets) × 100%. To verify the robustness of the conclusions of this paper, the return on assets is replaced by the return on equity to calculate the degree of enterprise zombification. The results are shown in

Table 4. Replacement of return on assets statistics.

	(1)	(2)	(3)
	zrop2	zrlp2	zrols2
DID	0.0040 **	0.0043 ***	0.0035 **
	(0.0016)	(0.0016)	(0.0016)
Observations	18,190	18,190	18,190
R-squared	0.0270	0.0247	0.0278
Number of id	3614	3614	3614
Year FE	YES	YES	YES
Individual FE	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, ** p < 0.05.

. The basic conclusions of this paper remain robust.

4.3.3. Using Balanced Panel

Unbalanced panel data have certain advantages and disadvantages in measurement. The advantage is that they can reflect the incompleteness and diversity of data in the real world, and the disadvantage is that they may introduce bias and uncertainty. Considering that the data of some entities or periods in the unbalanced panel are missing, and different entities or different periods having different observation data may lead to sample selection bias, a balanced panel is extracted from the data to explore the impact of a water resource tax on the zombification of enterprises to avoid the interference caused by the entry and exit of enterprises to the estimated results and ensure the robustness of the estimated results. The results are shown in

Table 5. Balanced panel results.

	(1)	(2)	(3)
	zrop	zrlp	zrols
DID	0.0049 **	0.0046 **	0.0035 *
	(0.0019)	(0.0020)	(0.0019)
Observations	9477	9477	9477
R-squared	0.0318	0.0290	0.0316
Number of id	1520	1520	1520
Year FE	YES	YES	YES
Individual FE	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where ** p < 0.05, * p < 0.1.

below. The coefficients and significance are basically consistent with the baseline regression, and the basic conclusions of this paper remain robust.

4.3.4. Shifting the Policy Introduction Time Forward

To exclude the possibility of these results being triggered by the factors without policy impact, the policy introduction time is shifted forward to 2013–2015 in this paper, and the regression results are shown in

Table 6. Estimated results of forward policy time.

	(1)	(2)	(3)
	zrop	zrlp	zrols
DID	0.0031	0.0042	0.0041
	(0.0028)	(0.0029)	(0.0029)
Observations	18,190	18,190	18,190
R-squared	0.0288	0.0274	0.0310
Number of id	3614	3614	3614
Year FE	YES	YES	YES
Individual FE	YES	YES	YES

below. The introduction time of fictitious policies is not significant year by year, indicating that there was no increase in the zombification of enterprises before 2016. This proves that the increase of the zombification of high-water consumption enterprises is caused by the policy impact. The basic conclusions of this paper remain robust.

4.4. Endogeneity Test

4.4.1. Sample Selection Bias

The pilot water resource tax areas are not selected randomly, but based on the needs of national economic and social development, comprehensive status of water resource, economic development level, and water access. The first pilot area, Hebei Province, is one of the provinces with the most serious water shortage in China and the only province without a river crossing the border. In recent years, the average water resources per mu (about 666 m²) and per capita water resources in Hebei are about one-seventh of the national averages, and the surface has been seriously short of water resources for a long time. Water resources are inborn deficient.

As an important part of the coordinated development of Beijing, Tianjin, and Hebei Province and an important agricultural and industrial base in China, Hebei Province has a variety of water consumers and industries, which is conducive to exploring the differential taxation policies and preferential policies among different water consumers and industries, and creates conditions for comprehensively promoting the water resource tax system and accumulating experience. Other pilot provinces also experience greater pressure in water resource management and protection, and it is urgent to adjust water resource demand through tax leverage so as to facilitate water resource conservation and protection.

In conclusion, the selection of pilot water resource tax areas is affected by some factors that are difficult to measure, and samples in the control group will introduce selection bias because they no longer represent the whole, resulting in bias in the estimated coefficient. Therefore, there are obvious spatial differences in the time and intensity of water resource tax reform, and the economic effect of reform may also vary from region to region. However, the reform performance of each region under differentiated reform cannot be evaluated by DID. To solve this problem, the synthetic DID proposed by Arkhangelsky, et al. [30] was adopted in this paper for estimation; the processing effect was estimated by weighting the time to solve the problems of sample selection bias and policy endogeneity, and the processing effect of each pilot area was reckoned.

The estimated results based on synthetic DID in

Table 7. Synthetic DID estimation results.

Variables	zrop
DID	0.0074 ***
	(0.0022)
Observations	15,972
Year FE	YES
Individual FE	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01.

show that, compared with enterprises in the control group, the average treatment effect of water resource tax reform on highly water-consuming enterprises is 0.0074, which is greater than the estimated coefficient of 0.0048 of the baseline regression. The baseline conclusions of this paper remain robust.

The synthetic double difference method can better handle the heterogeneity between the treatment groups and the control groups, using individual weights and time weights to construct a synthetic control group closer to the treatment group. This shows that compared with the traditional multi-stage double difference method, the treatment group synthesized by the synthetic double difference method has a better nature. Compared with the original data, the treatment method can reflect the policy effect of water resource tax reform more truly.

4.4.2. Instrumental Variables

Instrumental variables can eliminate endogenously induced bias and be used to perform consistent estimation. In this paper, the annual precipitation of the region where the enterprise is located is selected as the instrumental variable of the DID, and the two-stage least square method (2SLS) is adopted to ease endogenous problem. Regional annual precipitation can be regarded as the status of regional water resources that affect the selection of pilot areas, and it follows the principle of correlation. The degree of corporate zombification is more related to the financial situation and profitability of the enterprise, rather than the precipitation of the region where it is located. Therefore, it follows the principle of exogeneity. The regression results of instrumental variables are shown in columns (1)–(4) in

Table 8. Instrumental variables & triple difference method estimation results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Initial Step	Second Step			Triple Difference
	Rain	zrop	zrlp	zrols	zrop	zrlp	zrols
DID	0.0256 ***
	(0.0061)
rain		0.0257 ***	0.0263 ***	0.0257 ***
		(0.0030)	(0.0033)	(0.0032)
DID*rain					0.0009 ***	0.0008 ***	0.0007 ***
					(0.0002)	(0.0002)	(0.0002)
Observations	11,658	10,575	10,575	10,575	11,658	11,658	11,658
R-squared	0.1670	0.0210	0.0156	0.0170	0.0150	0.0094	0.0095
Number of id	2972	2868	2868	2868	2972	2972	2972
Year FE	YES	YES	YES	YES	YES	YES	YES
Individual FE	YES	YES	YES	YES	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01.

. Upon DID replacement with the lagged variable of regional annual precipitation, the regression coefficient of the main explanatory variable is still significantly positive, indicating that water resource tax furthers the zombification of highly water-consuming enterprises. Triple differences of “DID*rain” was also introduced in this paper, and the regression results are shown in columns (5)–(7) of Table 8; the baseline conclusions of this paper remain robust.

5. Heterogeneity Test

5.1. The Heterogeneity of Enterprise Property Rights

The heterogeneity of enterprise property rights may affect the degree of response of enterprises to policies, and then affect the degree of zombification of enterprises. From the political point of view, state-owned enterprises are generally strictly controlled by the government and assume corresponding social responsibilities, which have dual economic and political attributes. State-owned enterprises may be subject to stricter environmental supervision pressure and market regulation, and are more affected by policies. However, state-owned enterprises benefit from their own political attributes and the implicit guarantee of the government, and usually have strong government support and resource allocation advantages that make them more resistant to external shocks. Therefore, state-owned enterprises tend not to encounter problems affecting the production and financial operation constraints, meaning that the impact of the reform is not obvious. Generally, non-SOEs are relatively more market-oriented and more susceptible to market fluctuations, but they are also more flexible in responding to reforms. This nature may make the reform effect of non-SOEs less effective than SOEs.

In this paper, further discussion is introduced regarding the heterogeneous impact of property rights attributes on the relationship between water resources tax and the degree of zombification of enterprises. Sample enterprises are divided into local state-owned enterprises, central state-owned enterprises, and non-state-owned enterprises according to the property rights attributes of the actual controllers. Model (2) is re-estimated, and the results are shown in columns (1)–(3) of

Table 9. Heterogeneity estimation results.

	(1)	(2)	(3)	(4)	(5)	(6)
	Property Rights Heterogeneity			Locational Heterogeneity
	Local State Enterprise	Central State-Owned Enterprises	Non-State Enterprise	Eastern Part	Central Section	Western Part
DID	0.0060 *	0.0104 **	0.0022	0.0034 ***	0.0052 **	0.0038
	(0.0036)	(0.0049)	(0.0020)	(0.0013)	(0.0023)	(0.0028)
Observations	3524	1494	12,718	21,107	5043	4408
R-squared	0.0316	0.0397	0.0382	0.0061	0.0103	0.0065
Number of id	790	296	2727	2842	610	514
Year FE	YES	YES	YES	YES	YES	YES
Individual FE	YES	YES	YES	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, ** p < 0.05, * p < 0.1.

. The results show that the zombification degree of state-owned enterprises with high water consumption has increased significantly, indicating that the water resource tax policy has had a negative impact on the operation of state-owned enterprises, making them more likely to fall into a zombie state. In addition, the zombification of local SOEs with high water consumption is significantly higher compared to central SOEs, possibly due to differences in policy implementation. The water tax reform pays the water tax to local governments, which involves a larger amount of tax revenue compared to water resource fees, and therefore local governments will be more motivated to collect the tax than before the reform. Therefore, local SOEs, which are directly under the jurisdiction of local governments, are subject to stricter policy regulation and enforcement. Central SOEs, on the other hand, may have some flexibility in policy implementation. The increase in the zombification of non-state-owned enterprises was not significant.

5.2. The Heterogeneity of Enterprise Locations

The production environment of enterprises in Eastern, Western and Central China differs in the economic development level, market size, infrastructure, human resource, policy support, degree of openness, etc. Therefore, the heterogeneity of locations can be regarded as the heterogeneity of the objective living environment for enterprises. As the most economically developed region in China, the eastern region holds more than half of the country’s total GDP, has a large number of enterprises and mature production environment, and faces fierce competition. Enterprises in the central region enjoy relatively stable production environment, and develop in a more balanced way. Enterprises in the western region have more diversified production environment and huge development potential. To compare the impact of the water resource tax on the zombification of enterprise in different regions, the locations of enterprises are differentiated according to the positions of their locating provinces in China, and the regression results are shown in columns (4)–(6) of Table 9. Water resource tax significantly furthered the zombification of highly water-consuming enterprises in Eastern and Central China, but had no significant impact on western China.

The possible reasons are that the concentration of enterprises and stricter environmental control in the eastern region amplify the effects of the policy. Some highly water-consuming enterprises in the central region have certain competitive advantages in the industry, but the water resource tax may make it more difficult for these enterprises to operate and easier for them to enter the state of zombification. Compared with the central and eastern regions, the western region may have different industrial characteristics and water resource utilization, which makes the impact of water resource tax reform on the zombification of its enterprises insignificant.

Taking into account possible location heterogeneity, this paper examines whether there are ex ante differences between the subgroups based on “whether the firm is a state-owned enterprise” and “whether the firm is located in the pilot region”. The results in

Table 10. Test of the heterogeneity between the groups.

	(1)
	zrop
0b.place#0b.guoqi	0.0000
	(0.0000)
0b.place#1.guoqi	−0.0025
	(0.0022)
1.place#0b.guoqi	0.0018
	(0.0012)
1.place#1.guoqi	−0.0003
	(0.0023)
Observations	18,190
R-squared	0.0290
Number of id	3614
Year FE	YES
Individual FE	YES

show that the coefficient of soe*place is not significantly different between the two groups. It further shows that the result difference of the heterogeneity part of this paper is caused by the reform of water resource tax.

6. Mechanism Test

6.1. Mechanism of Financial Pressure

The rising financial stress is an important cause of the zombification of companies. This is a conclusion confirmed by numerous studies. Companies with higher financial stress may face production difficulties, such as heavier debt burdens, limited investment, brain drain, and broken capital chains. Water taxes could impose heavier financial pressure on water-hungry enterprises by raising their production costs, leading to a deeper zombification of them. In order to verify this impact mechanism, the values of enterprises’ current debt ratio and operating debt ratio lagged by one period are used in this paper to measure the financial pressure of enterprises and analyze the impact of water tax on the financial pressure of enterprises with high water consumption.

As shown in

Table 11. Financial stress mechanism test.

	(1)	(2)	(3)	(4)
	L. Operating Debt Ratio	L. Current Debt Ratio	zrop	zrop
DID	0.0203 ***	0.0122 **
	(0.0067)	(0.0056)
DID*L. Operating debt ratio			0.0085 ***
			(0.0019)
DID*L. Current debt ratio				0.0064 **
				(0.0026)
Observations	16,527	16,527	16,527	16,527
R-squared	0.0331	0.0211	0.0055	0.0044
Number of id	3225	3225	3225	3225
Year FE	YES	YES	YES	YES
Individual FE	YES	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, ** p < 0.05.

(1)–(2) below, the water tax significantly increases the current debt ratio and operating debt ratio of the enterprise. The first half of the paper was verified. In columns (3)–(4), we return the lag item with the DID to the degree of zombification of the enterprise. The result of the interaction item is significantly positive, indicating that the water tax reform leads to the increase of zombification of the enterprise by leading to higher debt and current debt ratio.

6.2. Mechanism of Environmental Uncertainty

Environmental uncertainty refers to the situation in which the future expectations of an enterprise become unstable and uncertain regarding the change of the external production environment. It is an indicator that reflects the change degree and impact of the external environment faced by an enterprise and is used to evaluate the operational risk, strategic choice and adaptability of an enterprise. Generally speaking, the higher the uncertainty of the environment, the greater the difficulty of decision-making and failure probability of the enterprise. Therefore, the higher the probability of conversion to zombie enterprises, the more flexible and innovative strategies are needed to deal with it. This may be one of the ways that the water resource tax will affect the zombification of enterprises.

After the imposition of the water resource tax, the operating environment of enterprises has changed notably, water cost increases, and the market demand fluctuates. These changes will lead to greater environmental uncertainty for enterprises. For highly water-consuming enterprises, the imposition of a water resource tax will inevitably increase their production costs, affect their profitability and market competitiveness, and let them face a more stringent operating environment. Greater environmental uncertainty may affect their operating decision, strategic choice, and zombification. In the face of greater environmental uncertainty, enterprises will be more cautious in investment and expansion, may reduce the start-up of new projects and market development, and choose relatively conservative operating strategies to cope with uncertainties and avoid the market risk of investment failure. And these behaviors may further the zombification of enterprises, raise the number of enterprises falling into a state of low profit and even loss.

In order to verify the impact of environmental uncertainty on enterprise zombification, this paper adopts the method of Shen et al. [31], taking the difference between the environmental uncertainty of the current year and the previous year as the change value of environmental uncertainty. Measure the relevant data from the annual report of the listed companies, and calculate the environmental uncertainty without the industry adjustment and the environmental uncertainty after the industry adjustment, as the intermediary variable of the enterprise zombification. The method is: calculate the standard deviation of the abnormal sales revenue in the past five years and divide the average sales revenue in the past five years as the alternative variable of environmental uncertainty, thereby obtaining the environmental uncertainty without industry adjustment. Subsequently, the median of the unadjusted environmental uncertainty of all companies in the same industry j in the same year is taken as the industry environment uncertainty in the same year. The standard deviation of the five-year residuals is obtained, divided by the average of the five-year sales revenue, and then adjusted by the industry median.

As shown in

Table 12. Environmental uncertainty mechanism tests.

	(1)	(2)	(3)	(4)
	Not Adjusted by the Industry	After Industry Adjustment	zrop	zrop
DID	0.0101 **	0.1395 ***
	(0.0043)	(0.0406)
DID*Not adjusted by the industry			0.0177 ***
			(0.0052)
DID*After industry adjustment				0.0017 ***
				(0.0006)
Observations	22,365	22,365	19,649	19,649
R-squared	0.0151	0.0177	0.0231	0.0229
Number of id	3016	3016	2853	2853
Year FE	YES	YES	YES	YES
Individual FE	YES	YES	YES	YES

Note: Robust standard errors are in parentheses, and “*” denotes significance level, where *** p < 0.01, ** p < 0.05.

, Sections (1) and (2) below, the implementation of the water resource tax reform significantly increases the environmental uncertainty of enterprises. We also added the interaction term between enterprise environmental uncertainty and DID. As shown in Table 12, Sections (3) and (4) below, the results in the table are significantly positive at 1%, indicating that the water resource tax improves the zombification index of enterprises by improving the environmental uncertainty of enterprises. Hypothesis 2 of this paper is verified.

7. Conclusions and Policy Proposals

Water resource is a basic and strategic resource, and the contradiction between its supply and demand goes against the construction of ecological civilization and the promotion of green development. Water resource tax, as a tax instrument for improving local tax system and facilitating the efficient use of water resources, can increase enterprises’ cost of water intake and guide enterprises to take the initiative to adopt water-saving measures and regulate their water consumption, which also provides a possible solution to the management of a large number of zombie enterprises in the highly water-consuming industries. The results show that the water resource tax promotes the zombification of enterprises with high water consumption, and there is heterogeneity between the property rights of enterprises and regions. In view of the above conclusions, the following policy proposals are made:

Improve the management mechanism for environmental uncertainties and strengthen the assessment of enterprise investment risks. Water resource tax reform has increased environmental uncertainty and accelerated the growth of net cash flows generated by corporate investment, implying an accelerated decline of enterprises in their life cycle. At the same time, the government can establish a more stable and transparent environmental management mechanism, provide clearer policy guidance, and reduce the uncertainty of objective environment for enterprises. Simultaneously, scientific and efficient collection and management system can ensure the efficiency and fairness of water resource tax to the maximum extent, and is a necessary prerequisite for the full imposition of water resource tax. Enterprises should actively build their anti-risk ability to deal with various risks brought by the increase of environmental uncertainty while perfecting the water resource tax system and supporting measures. Therefore, the government can also encourage enterprises to conduct a more comprehensive investment risk assessment to reduce the possible adverse consequences under the impact of water resource tax policy while improving the environmental uncertainty management mechanism.

Improve the exit mechanism for enterprises. Zombie enterprises remaining in the market will not only break the market order and squeeze the living space of other enterprises, their production mode of high energy consumption and low energy efficiency will also impose irreversible impact on the environment. Highly water-consuming enterprises with larger zombification degree are more likely to be zombie enterprises, but are still alive in the market due to the high cost of bankruptcy, the inability to support laid-off employees, the existence of legal risks, and other factors. As a result, measures should be taken to accelerate the exit of these enterprises. The enterprise exit mechanism can be improved through the following aspects: ① Adhere to the principles of marketization and legalization, do not engage in “one-size-fits-all” activities, and adopt different exit methods for different zombie enterprises, including self-liquidation, compulsory judicial liquidation, and bankruptcy liquidation. ② Strengthen policy guidance and support, improve the exit system, smooth the exit channels for main market players, reduce their exit costs, and vitalize them for competition. ③ Establish and improve risk prevention and handling mechanisms, strengthen the supervision and audit of zombie enterprises, and prevent zombie enterprises from surviving on the illegal provision of government subsidies, loans, and other means to ward off systemic risks.

Strengthen innovation and capacity-building in enterprise water resource management. The government should rationally plan the industrial structure and scale of production in accordance with regional water resource endowments and the carrying capacity of the water environment. For example, it should carry out water-use audits, water-use efficiency benchmarking, and water-saving reforms. Priority should be given to the development of low water-consuming and high-output industries, and the prior use of water by high water-consuming industries should be strictly controlled to avoid the waste and over-exploitation of water resources. At the same time, the efficiency and effectiveness of water use will be improved, the recycling of industrial water will be promoted, and the goal of efficient use of water resources will be realized. The government should also improve the efficiency and effectiveness of water use, promote the recycling of industrial water, and realize the multiple use of water and its gradual utilization. At the same time, enterprises should also strengthen the top-level design of water resource management, formulate water resource management strategies and plans in line with their own characteristics and development goals, clarify the objectives, tasks, measures and indicators of water resource management, establish and improve water resource management institutions and responsibility system, and implement the findings. In addition, enterprises should enhance their scientific and technological support, strengthen the research and development of key common technologies and equipment, promote the popularization of advanced and applicable technologies and techniques, improve the technical level and equipment level of water resource management, make use of the market mechanism, incentivize water-saving behaviors by means of pricing, taxation, subsidies, etc., to reduce the cost of water consumption, improve the competitiveness of water consumption, and enhance the vitality of the enterprise and the capacity for sustainable development.

8. Discussion

Based on the empirical data of listed companies of China water resource tax reform, this paper analyzed the impact of the water resources tax on the zombie of water consumption enterprises, and found that the heterogeneity between enterprise property rights and region, namely state-owned enterprises and groundwater in high water consumption enterprises, are more vulnerable to the negative impact of the water resources tax, and private enterprises and high water consumption enterprises can reduce the burden of the water resources tax by adjusting the water structure and improving water efficiency. This research result shows that resource tax can become a tool to indirectly contribute to the sustainable development goals, which has certain enlightenment and reference significance for water resource tax reform and zombie corporate governance in other regions or countries.

The question studied in this paper is how the water resource tax promotes the degree of enterprise zombification, but it does not discuss how the zombification process will end. This paper proposes a continuous variable that measures the degree of enterprise zombification. The following research is how to make up for the possible shortcomings of this paper, and how to better measure this index? This can be discussed in combination with the theory of enterprise life cycle from the direction of proposing the prediction model. And there are some shortcomings and limitations. For example, the time span of the water resource tax reform is relatively short. This paper uses public company data, with few samples. There are many factors affecting water resource tax reform, it is difficult to control water-related variables, and so on. Therefore, the future research needs to further expand the data sources, improve the accuracy of the enterprise zombification degree index, improve the quality of research, and deeply explore the mechanism, path, and effect of water resource tax reform to provide stronger support for the theoretical innovation and practical improvement of water resource tax reform.

The reform of water resource tax can learn from China’s experience and establish a more market-oriented, law-based and differentiated water price formation mechanism, so as to give full play to the price signal role of water resource tax. We will guide enterprises and residents to conserve water and rationally develop and utilize water resources. At the same time, we will avoid excessive intervention and administration, maintain the stability and predictability of tax policies, and reduce the uncertainty and institutional costs of enterprises. In addition, other countries can promote the efficient allocation of water resources through the reform of water resources tax, and realize the unification of the fairness and efficiency of water resources. For example, tiered water price and two-part water price are used to price different water sources, water quality, water, uses and users. This reflecting the scarcity and externality of water resources.