Water Resource Tax and Green Industrial Development: Reform from the Largest Emerging Economy

Abstract

The global challenge of water resource management presents a policy dilemma: while water resource tax aims to foster green development, it may hinder the economic potential of micro entities. This paper evaluates the efficacy of a trial of water resource tax reform in China regarding the green total factor productivity of listed Chinese industrial enterprises over the period spanning 2012–2019 by employing a quasi-natural experiment. This study utilizes multi-period Difference-in-Differences (DID) and propensity score matching methodologies to deal with the self-selection bias inherent in choosing pilot areas. The findings illustrate that the reform exerted a crucial beneficial impact on the GTFP of industrial enterprises. The main takeaway of this study is that the phased reform, integrating water resource taxes with the adaptation of micro entities, offers a pathway for economies to balance resource restrictions with sustainable development.

1. Introduction

There is no doubt that the global issues of groundwater overexploitation and inadequate water resource management have become increasingly pressing [1]. Implementing resource taxes has been proven to tackle the above concerns effectively [2]. However, a debate has broken out regarding the distortionary effects of water resource taxes on the economically sustainable growth of enterprises [3]. As key players in the market, enterprises face soaring compliance costs and economic burdens due to the imposition of water resource tax policies, leading to a reduction in productive efficiency [4]. Conversely, another argument has been proposed that water resource taxes can promote economically sustainable development by driving technological innovation and improving production processes in enterprises [5]. Notably, emerging markets and developing economies face particularly prominent challenges, including groundwater overexploitation, extensive patterns of water management, and the pressing need for improved economic efficiency [6]. Thus, gaining a comprehensive understanding of the role of water resource tax reform in an emerging economy, which possesses the most comprehensive industrial systems and is the fastest-growing economy, is significant for mitigating the conflict between water resource preservation and economic development worldwide.

As shown in Table 1, many countries, primarily advanced economies, have implemented water resource taxes to promote sustainable water management. For instance, Portugal, Italy, France, and Cyprus levy taxes on water extraction, primarily targeting direct users of water resources [7], which is similar to China’s practice of designating direct water users as taxpayers. However, several institutional differences remain between China’s reform and international practices. Regarding the responsible authority, countries such as France and Denmark assign the collection and administration of water taxes to regional governments or river basin organizations, while in China, these functions are centralized under national tax authorities. In terms of policy emphasis, Israel places particular focus on agricultural water use, whereas China exempts most agricultural users and prioritizes the regulation of industrial water consumption. In the taxation approach for water used in power generation, Spain links water resource taxes to the economic value of hydropower output, while China applies a “water follows electricity” principle, basing taxation on the actual volume of water used to generate power. In terms of institutional trajectory, most countries introduced water resource taxes as standalone fiscal measures, while China pursued a gradual “fee-to-tax” transition, shifting its water resources economic leverage from administrative charges to a more binding and compulsory tax instrument.

Although water resource tax policies differ across countries in their institutional design, they have at least shown effectiveness in achieving water management outcomes. Porcher analyzed France’s water tax and found that it effectively reduced water consumption [8]. Svendsen and Oh compared the cases of California and Denmark, highlighting Denmark’s success in reducing per capita water use through taxation [9]. Berbel et al. examined several EU countries and concluded that water resource taxes play a significant role in water conservation and the internalization of resource costs [10]. In China, Su and Cheng found that the reform improved regional total factor water efficiency [11], while Mu et al. reported that it helped alleviate water poverty in different areas [12].

Table 1. Water resource tax policies in different economies.

Groups in the World Economic Outlook Classification	Country Name	Instrument Name	Effective First Year	Intervention
Advanced Economies	Belgium	Groundwater tax	1997	Use
	Cyprus	Water use tax	2017	Use
	Denmark	Duty on piped water	1994	Use
	France	Charge on water abstraction	1964	Use
	Israel	Extra-use water levy	2009	Use
	Japan	Water and land utilization tax	1989	Use
	Latvia	Tax on the use of water resources for production of electricity	2014	Production
	Spain	Tax on the use of continental waters for the production of electric energy	2013	Production
	United States	Public community water system tax	1984	Use
Emerging Market and Developing Economies	China	Water resources tax	2016	Production
	Costa Rica	Water use Levy	1999	Use
	Mongolia	Tax on water use		Use
	Morocco	Tax on mineral and drinking water		Production

Note: data are from the PINE database and manual collation. “Emerging Market and Developing Economies” refers to “middle-income economies that are neither developed nor low-income economies” [13].

Table 1. Water resource tax policies in different economies.

Groups in the World Economic Outlook Classification	Country Name	Instrument Name	Effective First Year	Intervention
Advanced Economies	Belgium	Groundwater tax	1997	Use
	Cyprus	Water use tax	2017	Use
	Denmark	Duty on piped water	1994	Use
	France	Charge on water abstraction	1964	Use
	Israel	Extra-use water levy	2009	Use
	Japan	Water and land utilization tax	1989	Use
	Latvia	Tax on the use of water resources for production of electricity	2014	Production
	Spain	Tax on the use of continental waters for the production of electric energy	2013	Production
	United States	Public community water system tax	1984	Use
Emerging Market and Developing Economies	China	Water resources tax	2016	Production
	Costa Rica	Water use Levy	1999	Use
	Mongolia	Tax on water use		Use
	Morocco	Tax on mineral and drinking water		Production

Note: data are from the PINE database and manual collation. “Emerging Market and Developing Economies” refers to “middle-income economies that are neither developed nor low-income economies” [13].

This study focuses on responding to the ongoing debate by analyzing the influences of water tax reform in China. The water resource fee, once China’s primary economic lever for managing water use, was undermined by low collection standards, fragmented collection responsibilities, and irregular management. The reform adopted a more compulsory tax-based lever to replace the water resource fee, shifted institutional responsibility from fragmented departmental management to centralized tax authorities, and rendered water governance more binding and rigid in its regulatory effect, and it has been advanced through a phased, multi-batch pilot strategy to gradually reconcile the tension between resource conservation and economic development. It aims to alleviate groundwater overexploitation, improve water resource utilization efficiency, and promote the development of green industries. Despite still being guided by a development-oriented mindset in decision-making [14], China has nevertheless transitioned from administrative fees to a more stringent taxation system in water resource management. The world faces similar predicaments—rising water consumption rapidly contributing to groundwater overexploitation, increased water prices that impede enterprise growth [6], and inefficient water use in the industrial sector of emerging economies [15]. This ‘Chinese solution’ underscores the potential for emerging economies to implement stricter water resource management frameworks, even as they balance the competing demands of economic development and environmental protection. Examining this reform provides valuable insights for other economies facing comparable challenges.

This study examines whether the pilot policy has positively influenced the green development of micro entities, explicitly targeting industrial enterprises. This question has yet to be satisfactorily answered in the literature. Much of the literature centers around the effectiveness at the macro level, concluding that the reform has had positive influences. Ouyang et al. assessed the effectiveness of a pilot water resources tax policy in China and found that it improved water use efficiency and optimized water use structure [16]. Wang et al. employed a DSGE model to evaluate China’s water resources tax policy in Hubei Province, demonstrating its effectiveness in reducing water consumption and improving utilization efficiency through behavioral incentives [17]. Compared with the above studies, this paper contributes by examining the reform’s effects at the micro level, applying a quantitative empirical design and using a Difference-in-Differences approach to better estimate the policy’s net effect. Related studies have also examined the reform from a micro-level perspective. Liu [18] found that the reform enhances the environmental performance of businesses. Zhu and Zhang [19] scrutinized the mechanisms through which the water resource tax reform drives green technological innovation of the listed enterprises. Xu et al. [20] focused on high-water-consumption companies, a topic closer in object to this study. Xu et al. [21] examined productivity and innovation in high-water-consumption enterprises from 2007–21 and concluded that the water resource tax reform has had positive effects on enterprise productivity. Their research, however, does not extend to the industrial sector, nor does it examine whether the reform achieved the intended goal of fostering green development in enterprises. Furthermore, the findings in this paper differ from the current literature on water consumption enterprises. On account of these shortcomings, these studies fail to convincingly demonstrate whether the water resource reform has advanced the green development of industrial enterprises. The contribution of this paper is to provide a robust design that reveals the impact of the reform in China concerning the green total factor productivity of listed industrial companies, as well as offering a micro-level insight to address the global dilemma—balancing implementing taxation policies to restrict resource consumption with maintaining economic growth—especially in emerging economies.

Green total factor productivity is an effective indicator for measuring green development [22]. By incorporating environmental factors into the TFP index system, the goal is to minimize environmental pollution while maximizing input and output [23]. Because it considers the balance between economic development and environmental protection, it can reflect the quality of development [24]. This paper constructs a green total factor productivity index system for listed industrial enterprises to analyze the effect of the reform. Economic and environmental data from 1239 listed enterprises in China over the period spanning 2011–2017 comprise this index system. The calculated outcomes from the index system were designed to examine the impacts of the water tax reform in 2016 and 2017, respectively, by employing a quasi-natural experiment.

There is possible self-selection bias in this experiment, since provinces implementing the pilot policy of water resource taxation are not random. Besides employing multi-period DID, this paper utilizes propensity score matching to mitigate this bias. This method matches similar characteristics between individuals in the treatment and control groups to simulate dummy untreated samples and constructs a counterfactual framework. Then, the matched weight variables are added to the regression model to assess the influences of the reform. Additionally, this paper offers adequate tests to make sure the outcomes are convincing.

This study complements the literature with three marginal contributions as follows: First, unlike current studies that focus on water resource taxes in advanced economies, this paper offers new insights into exploring water resource taxation management tools in emerging economies, for which such research is scarce. Second, the theoretical analysis of previous studies about this reform centers on the policy itself [25,26], while this research introduces theories from the environmental taxation and organizational management fields to formulate a broader theoretical framework linking the macro level with the micro level. Third, an index system is established to measure green total factor productivity (donated by GTFP), allowing for a direct observation of how industrial enterprises adapt to the tax policy.

In public policy analysis, it is often difficult to ensure a random distribution of samples due to the inherent differences between groups before policy intervention. A quasi-natural experiment is a research design that simulates a natural experiment when the sample is not randomly assigned [27]. Because the research subjects usually do not influence the implementation of policies, public policies can be considered as exogenous shocks to the research subjects. Based on this, the implementation of China’s water resource tax policy can be regarded as a quasi-natural experiment.

Overall, this research provides credible evidence that China’s water tax policy advances the GTFP of A-share listed industrial enterprises. The empirical results indicate that systematically implementing water resource taxation policies is able to balance improvements in both water use efficiency and the economic efficiency of micro entities, thereby promoting green development.

In what follows, Section 2 introduces the policy background, and the theoretical framework is developed. Section 3 outlines the data used in the analysis and the empirical methods employed. Section 4 provides an in-depth examination of the empirical findings, accompanied by robustness tests to validate the results. Next, Section 5 discusses the novel insights from this study, compares them with previous research, and emphasizes this paper’s contributions to the field. Finally, Section 6 concludes the main findings and offers policy advice.

2. Policy Background and Theoretical Analysis

2.1. Policy Background

China’s fast development has resulted in a surge in water requirements, posing severe challenges to water resource management [28]. By 2050, the proportion of industrial water usage in China is projected to increase steadily [29]. However, water resource utilization efficiency is low, and it hinders the advancement of the green economy [30].

The Chinese government considers tackling water resource issues as a national strategic priority [31]. In 1988, China enacted and implemented the Water Law of the PRC [32] Subsequently, the Regulation on Water Extraction Licenses and Collection of Water Resource Fees was introduced by the MWR in 2006 [33]. However, due to low overall fee standards, arbitrary selection of collection subjects, and inconsistent enforcement, the fee system faced several limitations. Fee rates were often determined based on local economic conditions rather than water resource scarcity, which, combined with rent-seeking behavior by local governments pursuing economic growth, weakened the effectiveness of the fee system in curbing excessive water consumption. Additionally, the fee system lacked sufficient legal rigidity, resulting in low collection rates and limited enforcement. These shortcomings highlighted the need for a more robust and standardized approach to water resource management, prompting the Chinese government to initiate the transition from water resource fees to taxation.

Based on this background, three national departments jointly initiated a pilot policy to shift water resource charges to taxation, with the objectives of mitigating groundwater over-exploitation, encouraging a shift towards green development, and providing a model for implementation in other regions. The reform has two key features. First, the collection authority shifted from multiple administrative departments to a unified tax administration. Under the previous fee system, the collection, management, and supervision of water resource fees were divided among the water resources department, the Ministry of Finance, and the price department, leading to coordination challenges and inefficiencies. In contrast, the water resource tax is uniformly collected by tax authorities, streamlining the process and reducing the potential for rent-seeking behavior. Second, unlike the flexible and compensatory nature of fees, taxes are inherently mandatory and legally binding [6], requiring taxpayers to comply with the law or face penalties. This fundamental difference significantly strengthens the regulatory effectiveness of the tax system compared to the previous fee system. Additionally, the reform’s centralized and legally binding framework increases compliance costs for firms by reducing evasion opportunities and enforcing stricter reporting and monitoring requirements.

Considering that the reform is likely to cause significant impacts, selecting the pilot zones requires considerable caution. According to the Chinese government, in order to fully leverage taxation, those provinces forced to participate were chosen based on their serious water supply–demand imbalances and critical groundwater overexploitation instead of through random selection [34]. What is more, the selection of other pilot zones needs to consider the typical idiosyncrasies of provinces and whether their willingness to take part in the reform is strong. Hence, those treated provinces may have self-selection issues, which could bring challenges to later empirical analysis.

It is worth noting that, according to many relevant documents, the reform does not focus on agricultural production and domestic water usage. Even though agricultural water consumption occupies a significant proportion of overall water usage, the reform still highlights exempting agricultural production and domestic water usage from taxation. In the case of domestic water usage, the tax is not directly imposed on households but on water suppliers [16], and the tax amount is consistent with the previous water resource fee. As a result, households experience no direct impact from the reform [17]. So far, the emphasis of the reform remains on industrial production, which is a critical factor in selecting the research focus for this study.

Anyway, China’s water management has transformed from a fee-based system to a taxation model, shifting from pilot trials to full implementation, and it has a binding character. It is worth highlighting that the current reform imposes minimal regulation on agricultural water usage, and therefore, this study predicts that it will have an insignificant effect. On the other hand, industrial water usage is a primary target of regulation, and the considerable tightening of water resource management is expected to have a substantial impact on industrial production and operations. Thus, the subsequent analysis in this study focuses on evaluating the efficacy of the reform in enhancing the economic efficiency and green development of industrial enterprises, particularly in terms of green total factor productivity, compared with non-pilot areas.

2.2. Theoretical Analysis

Institutional Change Theory proposes that changes in the macro environment can create external incentives to influence business strategies and behaviors [35]. In the past, water was managed through resource fees. And yet, this regulatory system lacked legal coercion, leaving room for companies to evade compliance [36]. The shift from a fee-based system to a tax-based system represents a fundamental institutional change. The previous water resource fee system faced challenges such as fragmented regulation, inconsistent enforcement, and limited policy rigidity. These structural limitations allowed firms to underreport usage or delay payments without significant penalties, reducing their compliance costs. The reform centralizes authority within the taxation department, establishing a unified, legally binding framework. This enhances regulatory rigidity, increasing administrative and operational burdens on firms. By reducing evasion opportunities and forcing firms to internalize more of the cost of water usage, the reform significantly raises compliance costs, creating strong incentives for firms to adapt their strategies and behaviors.

The Double Dividend Hypothesis supports that taxing environmentally harmful behavior can improve environmental quality and enhance economic efficiency [37]. The macro-level policies need to bring positive influences on micro entities. It depends on dynamic capabilities that enterprises reallocate resources to adapt to changes in the external environment [38]. To deal with the challenge that implementing the reform leads to higher compliance costs, enterprises need to dynamically adjust their business strategies, such as optimizing factor allocation to reduce the share of water resources in production costs and alleviate financial pressures. By integrating, constructing, and reconfiguring both internal and external resources in response to the cost pressures induced by the water resource tax, enterprises can create and maintain a competitive position within the industry [39,40,41]. Build on this, the logical framework diagram is presented in Figure 1, and this study proposes the following hypothesis:

Hypothesis 1.

China’s pilot policy of the water resource tax reform has a positive effect on the green total factor productivity of listed industrial enterprises.

The policy effects of the water resource tax reform may vary depending on the degree of water resource dependency among industrial enterprises, leading to differential impacts on their green development. Firms in high-water-consuming industries, due to their strong dependency on water resources in the production process, face limitations in their dynamic adjustment capabilities, making it difficult to significantly reduce water usage through adjustments to production processes. In high-water-consuming industries, the production process is heavily reliant on water resources, which restricts their ability to dynamically adjust and optimize resource allocation in response to the water resource tax reform. As a result, these firms struggle to effectively reduce water consumption, thereby limiting the improvement in their green development. In contrast, firms in non-high-water-consuming industries are less dependent on water resources, granting them greater flexibility and stronger dynamic adjustment capabilities. Under the water resource tax reform, these firms can leverage their dynamic capabilities to more easily adjust production processes and resource allocation, thereby enhancing their green development.

Hypothesis 2.

The positive effect of the water resource tax reform on the green total factor productivity is more pronounced in non-high-water-consuming firms compared to high-water-consuming firms.

The impact of the water resource tax reform on the green total factor productivity of industrial enterprises may vary significantly depending on the level of industry competition. A highly competitive environment enables firms to develop stronger dynamic adjustment capabilities, allowing them to quickly adapt resource allocation and respond to external environmental changes [42]. In highly competitive industries, firms face intense market competition, which drives them to continuously optimize resource allocation and improve efficiency to maintain their competitive edge. In the context of the water resource tax reform, firms in highly competitive industries can respond more rapidly and effectively to policy changes due to their existing abilities to cope with market pressures. To address the cost pressures induced by the reform, these firms have stronger incentives to enhance efficiency and optimize resource allocation. In contrast, firms in less competitive industries lack long-term market competition pressure, resulting in weaker dynamic adjustment capabilities and slower response speeds. Consequently, firms in less competitive industries struggle to quickly adapt their resource allocation in response to the reform, leading to a more limited impact on their GTFP.

Hypothesis 3.

Compared to firms in less competitive industries, the water resource tax reform has a stronger positive effect on the green total productivity of firms in highly competitive industries.

Environmental management is a critical mechanism for enterprises to address environmental challenges and enhance sustainable development capabilities. Firms that participate in voluntary environmental management have already developed comprehensive monitoring, assessment, and adaptive environmental management systems, allowing them to effectively manage resource utilization [43]. Due to their high-level environmental management capabilities, the institutional pressures brought by the water resource tax reform do not exceed their “load-bearing capacity,” resulting in limited impact on their green development. In contrast, firms that do not participate in voluntary environmental management lack mature environmental management systems, making it difficult for them to cope with the external pressures induced by the water resource tax reform. Under the water resource tax reform, these firms need to dynamically adjust their operations to respond to policy changes, thereby enhancing their green development.

Hypothesis 4.

Compared to firms that engage in voluntary environmental management, the water resource tax reform has a more pronounced impact on the green total factor productivity of firms that do not adopt such voluntary practices.

3. Research Design

3.1. Sample Selection and Description of Data

This paper identified industrial sectors—mining and quarrying (B), manufacturing (C), and production and supply of electricity, heat, gas, and tap water (D)—based on the National Economic Industry Classification. This study integrated data from three primary sources: enterprise-level data from the CSMAR database, including net fixed assets for capital input, operating costs for energy input calculation, operating income for expected output, as well as firm size, growth, age, board size, and independent director ratio for control variables; industry-level data from the China Environmental Statistical Yearbook, providing total energy consumption, SO₂ emissions, COD emissions, and solid waste generation for environmental performance measurement; and regional-level data from the China Statistical Yearbook and National Bureau of Statistics, featuring GDP per capita for regional economic development assessment.

This study had an objective limitation regarding the selection of the time window. The primary reason was that China’s statistical statement system for fixed-asset investment prices was abolished in 2020, and the National Bureau of Statistics no longer publishes the fixed-asset investment price index. Considering the initiation of the reform, the final time window selected for analysis was from 2012 to 2019.

3.2. Variables Measurement

3.2.1. Explained Variable

The explained variable is green total factor productivity. The literature generally uses the Data Envelopment Analysis (DEA) methodology to measure green total factor productivity, integrating input factors and output factors [44]. Traditional DEA models are limited by the assumption that inputs and outputs must be scaled proportionally, which prevents them from effectively accounting for slack variables [45]. The approach proposed by Tone and Tsutsui offers a solution through the EBM model [46]. This model relaxes the assumption of proportional input reductions in the radial function, enabling the effective combination of both radial and non-radial approaches [47]. In this study, the EBM model with the GML index was used to quantify green total factor productivity utilizing the MaxDEA software (https://www.maxdea.com/).

Drawing upon the relevant research [35], the paper adopted the following indicators, as shown in

Table 2. Input–output factors.

Indicator Type	Primary Indicator	Secondary Indicator	Unit
Input	Capital investment	Capital stock	RMB
	Labor input	Employee	person
	Energy input	Standard coal consumption	10,000 tons
Output	Expected output	Operating revenue	RMB
	Undesired output	SO2 emissions	ton
		COD emissions	ton
		Solid waste	10,000 tons

, the input–output factors table: The net value of fixed assets, modified to reflect inflation using the fixed-asset investment price index, which is used to define capital input. Labor input, denoted as the total number of employees. To estimate the energy input, the occupation of the firm’s operating costs in the corresponding industry is first calculated and then multiplied by the total energy usage for that industry. Expected output refers to the operating income and adopts the GDP deflator in 2011 to adjust. Undesired output contains three kinds of waste: SO₂ emissions, COD emissions, and solid waste. The approach for calculating undesirable results is consistent with the energy input. Following the approach of Zhong and Li [48], it is assumed that the GTFP in 2011 is 1, and the GTFP in 2012 is the value in 2011 multiplied by the GML index for 2012. The GTFP for subsequent years is then deduced by analogy.

3.2.2. Explanatory Variable

The explanatory variable is the water tax reform (denoted by Policy_time). This variable is defined as the product of a regional dummy variable and a time dummy variable. If a business is located in a pilot province, the value of the regional dummy variable (Policy) is set to 1; otherwise, it is set to 0. If the pilot policy has been implemented in the province where the firm is located, the time dummy variable (time) is also set to 1. In this study, we regard 2016 and 2018 as the policy shock years. The water resource tax reform was implemented in a staggered manner, with the first pilot launched in Hebei Province in May 2016 and extended to nine additional regions in December 2017. Given the late implementation in 2017 and the potential lag effects of the policy, we selected 2018 as the effective year for the second phase of the reform in the nine regions.

3.2.3. Control Variables

The control variables were as follows: Firm size (size): this is represented by the total assets of the firm, which reflects the scale of operation. Leverage (lev): this is described as the proportion of total liabilities to total assets, representing the financial risk level of the enterprise. Growth (growth): this metric reflects the asset’s growth rate, assessing its potential to expand and adjust in response to market fluctuations. The enterprise’s age (lnage): this is measured by the years since its IPO, and this factor indicates the stability and maturity of the business within its industry. Board size (board): this represents the number of members on the board, which impacts the internal decision-making structure of companies. Proportion of independent directors (indep): this represents the occupation of independent directors, which affects the quality of corporate decision-making concerning resource allocation, risk management, and compliance. In addition, given that this study investigates the effectiveness of provincial policy implementation, a variable for regional economic development status (lnpgdp) is included to improve the credibility of the model. This paper employed censoring alongside logarithmic transformations on the pertinent continuous variables.

3.3. Empirical Strategy and Model Set

This study regards the reform as a quasi-natural experiment and applied the DID model, which is used in analyzing kinds of policies by comparing the treatment group with the control groups [49], to measure its effects. In accordance with the reasons mentioned above, to improve the credibility of the empirical results [50], the paper integrated PSM and multi-period DID.

The benchmark model is shown in Equation (1).

$${\mathit{GTFP}}_{\mathit{it}}={\alpha }_0+{\beta }_1{\mathit{Policy\_time}}_{\mathit{it}}+{\beta }_2{X}_{\mathit{kit}}+{\delta }_i+{\mu }_t+{\epsilon }_{\mathit{it}}$$

(1)

where i, t, and k represent the individual industrial enterprises, the year, and the number, respectively. If the policy had a great influence on the listed industrial enterprises, it will be significantly positive. X_kit denotes control variables, ${\delta }_i$ refers to the individual fixed effect, ${\mu }_t$ represents the year fixed effect, and ${\epsilon }_{\mathit{it}}$ is the random error term.

4. Empirical Results

4.1. DID Estimation

Descriptive statistics for the relevant variables are shown in

Table 3. Descriptive statistics.

Statistic	N	Mean	SD	Min	Max
GTFP	9204	1.467	1.062	0.333	7.447
size	9204	22.34	1.245	19.98	26.08
lev	9204	0.432	0.199	0.0560	0.897
growth	9204	0.114	0.212	−0.276	1.168
lnage	9204	2.231	0.723	0	3.219
board	9204	2.149	0.193	1.609	2.708
indep	9204	0.372	0.0520	0.333	0.571
lnpgdp	9204	11.05	0.417	10.14	11.97

Note: The table shows the number of observations, the mean, the standard deviation, and the minimum and maximum values.

.

Table 4. The estimated results of DID.

Variables	(1)	(2)
	GTFP	GTFP
Treat_time	0.207 **	0.194 **
	(2.620)	(2.526)
control variable	NO	YES
fixed effect	YES	YES
N	9198	9198
R2	0.753	0.761

Note: ** denotes the 5% significance level; values in parentheses are the robust standard error clustered at the province level.

describes the average effects derived from the DID benchmark model. Column (1) shows that under the empirical results without any control variables, the coefficient of the expected variable was at the 0.05 significance level. Subsequently, to assess the robustness of the empirical results, control variables were incorporated into the model. After running the software again, the statistical direction and significance of the calculated results shown in column (2) were consistent with those in column (1). Based on this, the hypothesis is supported in Section 2.

4.2. PSM Analysis

This section applies the control variables as matching covariates in order to find industrial enterprises similar to the treatment group. Nearest neighbor matching was chosen with a 1:3 ratio with the logit model to measure the propensity score. Next, this section tests the balance, the common trend hypothesis, and the matching quality in order.

After matching,

Table 5. Balance test of PSM.

Variables	Unmatched/Matched	Mean		T
		Treated	Control
size	U	22.597	22.229	1.49 *
	M	22.597	22.59	1.20 *
lev	U	0.441	0.428	1.09 *
	M	0.441	0.439	1.06
growth	U	0.100	0.120	0.79 *
	M	0.100	0.101	1.10 *
lnage	U	2.279	2.209	0.92 *
	M	2.279	2.295	0.98
board	U	2.181	2.136	1.09 *
	M	2.181	2.176	1.04
indep	U	0.368	0.374	0.89 *
	M	0.368	0.370	0.95
lnpgdp	U	10.998	11.074	1.09 *
	M	10.998	10.986	1.03

Note: * denotes the 10% significance level.

and Figure 2 show that the t-test results were no longer statistically significant, and the standard deviations of all covariates were less than 10%. Figure 3 and Figure 4, demonstrate that, after matching, most observations were located within the range of common support and that the propensity scores were closer than before, respectively. Based on the above results, relevant checks were satisfied, and the matching quality was acceptable.

4.3. PSM-DID Estimation

Table 6. The estimated results of PSM-DID.

	Neighbor Matching	Caliper Matching	Nuclear Matching
	(1)	(2)	(3)
Variables	GTFP	GTFP	GTFP
Treat_time	0.200 **	0.201 **	0.195 **
	(2.131)	(2.055)	(2.531)
control variable	Yes	Yes	Yes
fixed effect	Yes	Yes	Yes
N	6874	3886	9180
R2	0.759	0.785	0.761

Note: ** denotes the 5% significance level; values in parentheses are the robust standard error clustered at the province level.

presents the DID regression results after matching. As can be seen from Table 6, the regression coefficient of Policy_time_it was 0.200, which was significantly positive at the 5% significance level. The results in Table 6 are also consistent with the empirical result in Table 4, which again verifies that the tax reform exerts a significant positive effect on green total factor productivity. Furthermore, additional methods of these estimates were performed using other matching methods, including caliper matching and nuclear matching. As indicated by Table 6, the empirical results from the other methods did not have any statistically significant differences from the nearest neighbor matching, thus strengthening the credibility of the findings.

4.4. Heterogeneity Analysis

4.4.1. Industrial Water Consumption Intensity

This section conducts a deep analysis of the impact of the reform on typical industries by examining the water usage levels of various industrial sectors. The classification of high-water-consuming industries follows that in [51], identifying 22 industries as high-water consumers based on the Guidelines for Industry Classification of Listed Companies. Detailed information is provided in

Table 7. Classification of high-water-consuming industries.

Sector	Industry Name
Mining and quarrying (B)	Mining and washing of coal, extraction of petroleum and natural gas, mining and processing of ferrous metal ores, mining and processing of non-ferrous metal ores, ancillary mining activities
Manufacturing (C)	Processing of food from agricultural products, manufacture of foods, manufacture of wines, beverage and refined tea, manufacture of cigarettes and tobacco, manufacture of textiles, manufacture of textile wearing apparel and ornament, manufacture of leather, fur, feather products, and footwear, manufacture of paper and paper products, printing, reproduction of recording media, manufacture of cultural, educational, arts and crafts, sports, and entertainment products, processing of petroleum, coking, processing of nucleus fuel, manufacture of chemical raw material and chemical products, manufacture and processing of ferrous metals, manufacture and processing of non-ferrous metals, manufacture of fabricated metal products
Production and supply of electricity, heat, gas, and tap water (D)	Production and supply of electric power and heat power, production and distribution of gas

.

The outcomes displayed in fields (1) and (2) of

Table 8. Estimation results of heterogeneity analysis.

	(1)	(2)	(3)	(4)	(5)	(6)
	Non-High-Water Consumption	High-Water Consumption	High Competition	Low Competition	Non-Certified	Certified
Variables	GTFP	GTFP	GTFP	GTFP	GTFP	GTFP
Treat_time	0.208 ***	0.126	0.201 **	0.118	0.224 **	0.079
	(2.995)	(1.313)	(2.242)	(1.102)	(2.316)	(0.921)
control variable	Yes	Yes	Yes	Yes	Yes	Yes
fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	5316	3880	5750	3325	6581	2346
R2	0.839	0.755	0.797	0.813	0.778	0.819

Note: ** and *** denote the 5% and 1% significance levels, respectively; balues in parentheses are the robust standard error clustered at the province level.

reveal that the pilot policy has had no significant effects on the GTFP of high-water-consuming industries, while the results for non-high-water-consuming industries were completely the opposite. This discrepancy in impact may be attributed to differences in the adaptability and flexibility of the kinds of industries in reducing water consumption.

The underlying reason for this impact lies in the characteristics of the production processes within each industry. Industrial businesses in high-water-consuming industries are heavily resource-dependent, and the cost increases resulting from the reform directly and significantly raise their operating costs. Consequently, these industries have limited capacity for dynamic adjustment under the rigidity of the water resource taxation, which restricts their ability to improve the green total factor productivity. In contrast, for other industrial enterprises, water is not a critical constraint in their production processes, allowing them greater flexibility in responding to the water reform. This flexibility is reflected in their ability to quickly and effectively reallocate resources to manage expected water costs, thereby reducing water consumption without compromising productivity.

4.4.2. Degree of Industrial Competition

Given that differences in industry competition may result in disparities in motivations for resource reallocation and dynamic adjustment capability, this paper investigates the differential impact on industries with distinct levels of competition. The Herfindahl–Hirschman Index (HHI) is used to measure industry concentration [52]. The industries are then divided into two categories: those that are highly competitive, where the HHI is lower than the mean, and those that are less competitive, with an HHI higher than the mean.

In Table 8, columns (3) and (4) show that the pilot reform has boosted the GTFP of firms in high-competition industries. The reason is that, with the enhancement of market competition, enterprises are forced to improve efficiency and reduce costs in order to maintain their competitive edge in a fierce market environment [53].

These industries operate in an environment that demands continuous performance monitoring and rapid adaptation. Firms that cannot respond rapidly to changes in policy or other exogenous circumstances face the risk of losing their market share. This constant pressure further heightens the dynamic capabilities of firms in terms of their speed in reconfiguring resources and processes, thus enabling an effective response to cost pressures brought about by the pilot reform. The high-intensity dynamic capability suggests that companies not only have a heightened sensitivity to regulatory changes in competitive industries but have also developed a stronger incentive to transform these regulatory challenges into opportunities for efficiency gains, driven by both internal performance incentives and external competitive threats.

4.4.3. Voluntary Environment Management

To evaluate the interaction between the external water resource tax pilot policy and enterprises’ voluntary environmental management participation, this study uses ISO14001 international standard certification as an indicator of voluntary environmental management [54]. Enterprises are categorized into a non-participation group, with a value of 0 assigned, and a participation group, with a value of 1 assigned, to assess the heterogeneity of the policy effects.

The results in columns (5) and (6) of Table 8 indicate that the water resource tax reform has enhanced the GTFP of non-certified firms, whereas the impact on enterprises engaged in voluntary environmental management is not statistically significant.

The heterogeneity arises because of the different maturity of environmental management practices across enterprises. Enterprises that are, therefore, involved in voluntary environmental regulation, like those with ISO14001 certification [54], represent a proactive management framework at high standards, incorporating environmental responsibility within operations. These enterprises have already developed comprehensive monitoring, assessment, and adaptive environmental management systems to effectively manage resource utilization. The institutional pressures brought by the water resource tax reform have not exceeded the “load-bearing capacity” brought about by strict environmental management. Hence, the regulatory changes have not introduced new challenges, and the green total factor productivity of enterprises has remained stable. In addition, less maturity in robust frameworks characterizes companies that also do not engage in any voluntary attitude concerning environmental protection. This reform puts a greater cost on water usage, hence exerting external pressure beyond their static capacities, which is bound to spark dynamic adjustments.

4.5. Robustness Checks

4.5.1. Parallel Test

To conduct the parallel test, this study adopted an event study analysis [55] to examine the differences between the treated and untreated groups for each year relative to the reference year. Given the relatively late timing of the trial expansion for the water tax, 2018 is considered the first year of policy implementation in the nine regions. Lead and lag terms were created for the years before and after the implementation year to capture the dynamic effects. The following model was constructed to evaluate the dynamic effects of the water resource tax reform:

$${\mathit{GTFP}}_{\mathit{it}}={\alpha }_0+{\Sigma }_{2012}^{2019}{\beta }_1{\mathit{Policy\_time}}_{\mathit{it}}+{\beta }_2{X}_{\mathit{kit}}+{\delta }_i+{\mu }_t+{\epsilon }_{\mathit{it}}$$

(2)

Figure 5 presents the results of the parallel trend assessment. The regression coefficients for the three periods preceding the policy implementation were statistically insignificant and fluctuated around zero, indicating that the control group did not systematically differ from the treatment group before the reform. This supports the parallel trends assumption, a key requirement for the DID approach. After the pilot policy was implemented, the coefficients began to exhibit an upward trend, with the coefficients for the last two years reaching at least 95% statistical significance. These findings indicate a lag influence of the water tax reform on the GTFP of listed industrial companies.

Furthermore, Figure 5 reveals that the effects caused by the policy were particularly positive in the second year after implementation (post_2). This study attributed this phenomenon to the “stressful changes” that enterprises made in production processes and resource allocation during the first year (post_1) following the policy change. These adjustments, driven by external regulatory changes, began to increase GTFP in the year after the policy was launched, with a notable effect in the second year. However, the relatively straightforward and effective measures were largely exhausted during this period, resulting in diminishing marginal benefits from the remaining measures. This phenomenon triggered temporary “policy fatigue” and resulted in a lower impact in the third year (post_3) compared to post_2. Nevertheless, the positive impact of the reform remained statistically significant at the 99% confidence level in post_3, with expectations for continued and gradual impacts in the following years.

In summary, the parallel trend assessment confirms the validity of the model, demonstrating that before the policy, the tendency between the control and the treatment groups was consistent. Thus, the DID model used in this paper satisfies the parallel trends assumption.

4.5.2. Placebo Test

This research conducted a placebo test to determine whether factors other than the pilot reform may have contributed to the improvement in the GTFP of industrial enterprises. If the selection of the control groups is valid, the processing effects from conducting this test are statistically insignificant, and the coefficient ought to hover around 0. Figure 6 indicates that the coefficients followed a normal distribution centered around zero. The red circles represent the estimated coefficients from 1000 placebo replications, while the blue line shows the kernel density of their distribution. This finding demonstrates that the placebo test was successfully performed, thereby confirming the robustness of the study’s conclusions.

5. Discussion

According to prior research, water resource tax reform has been shown to improve GTFP, thereby promoting green development. This effect appears to be short-term and is more marked in industries with lower water consumption and higher levels of competition and among enterprises with well-established environmental management systems.

This study suggests that the shift in water resource management policies has triggered a “temporary” adaptive response from industrial enterprises at the production level [56]. The rigidity of the taxation system has made increased water usage costs inevitable for enterprises. Under production pressures, industrial enterprises have had no option but to adopt dynamic adaptation mechanisms to cope with the new production constraints—such as reducing water-intensive activities and reallocating resources to mitigate the additional expenses resulting from external shocks. As a result, the findings clearly illustrate that after the reform, there was a significant improvement in the GTFP of industrial enterprises.

This study further explored the internal mechanisms through which water resource policies impact enterprises at the micro level. The existing literature generally asserts that the pilot policy motivates enterprises to innovate in response to external regulatory pressures. However, this study indicates that, during the initial phase of policy implementation, these positive changes were more likely due to enterprises optimizing their resource allocation. Given the high investment and long development cycle typically associated with water-saving innovations [57], it is challenging for enterprises to quickly convert R&D investments into practical outcomes in the short term. Furthermore, if industrial enterprises have low water dependency and strong dynamic capabilities, internal resource optimization alone is sufficient to address the pressures from the reform. The contribution of this study is to offer a new perspective: that the success of water resource taxation policies depends not only on achieving long-term environmental goals but also on assisting micro entities participating in the reform with overcoming difficulties and adapting to a new production environment during the early phase of implementation.

First, this study has crucial implications for policymakers aiming to advance effective resource management while simultaneously guiding firms to green development. At least, over a short time span, water taxation enables enterprises to reconfigure their production resources to support green development without immediately necessitating high-risk, high-cost research and development activities.

Moreover, the findings provide valuable insights for calibrating regulatory pressure levels at different stages of policy implementation. Even though the eventual goal of resource taxes is to encourage corporations to promote green innovation and transition, excessive taxation may give rise to high compliance costs, thereby reducing their long-term potential. It is reasonable that policymakers take account of adopting a gradual approach to implementing water resource taxation policies, allowing enterprises to have enough time to make necessary adjustments and establish a green innovation system.

Therefore, for economies in a dilemma—particularly emerging economies—this study highlights a transferable principle: achieving the dual objectives of resource protection and economic sustainability through a phased implementation of water resource taxation. Water taxation should incentivize enterprises to optimize internal resource allocation, followed by a gradual transition toward green innovation. While the findings of this study offer valuable insights for emerging economies grappling with similar challenges, the external validity may be constrained by the specific context of the empirical analysis. Additionally, due to the lack of separate disclosure of employee counts for parent companies and subsidiaries by most firms, this study captures overall firm-level trends through consolidated financial data. Future research could enhance the generalizability of these findings by conducting comparative studies using macro-level, cross-country data to explore the broader applicability of water resource taxation policies, as well as by incorporating more granular firm-level data.

Unfortunately, as noted in the research design of this study, the objective difficulty in obtaining data for indicators has limited the selection of the period of study. In the future, research could tackle this issue by changing more comprehensive samples, particularly by examining the impacts of the reform following a full-scale introduction. Further investigation into the mechanism of the policy’s effects over different time periods would enhance both its applicability and research value.

6. Conclusions and Policy Recommendations

6.1. Conclusions

Although the effectiveness of policy instruments has been proven, the world—particularly emerging market countries—continues to face the challenge of balancing resource conservation with economic growth. Ensuring that water resource tax policies effectively curb resource over-exploitation without creating disincentive effects for micro entities while also promoting the green transition of enterprises is recognized as a pressing challenge for attaining sustainable development. This study examined the institutional pressure that drives enterprises toward green development through dynamic resource allocation, focusing on water resource tax policies, with detailed theoretical analysis and rigorous empirical testing. The principal findings are as follows: (1) The water tax reform has enhanced green total factor productivity. The conclusions are credible, as validated through various robustness checks. (2) The reform has improved green productivity, specifically in non-high-water-consuming enterprises, highly competitive industries, and enterprises that do not implement internal environmental management.

6.2. Policy Recommendations

According to the above analysis, this study provides practical recommendations for similar policy tools in various economies. The specific policy suggestions are as follows: (1) Adopting a phased and progressive approach. Initially, enterprises should be guided to establish dynamic adjustment mechanisms for water resource management while advancing research and experimental development spending. This would assist enterprises in transitioning from simple resource optimization to a deeper green transition. (2) Formulating targeted strategies based on the characteristics of the industry. For example, authorities should provide convenience to enterprises with high water demand by directly offering fund support or presenting policies related to mitigating financial difficulties. For industries lacking competitiveness, water market mechanisms should be employed, such as compulsory water rights trading, to stimulate enterprises to reduce water consumption. (3) Enterprises ought to voluntarily participate in environmental management. They are capable of instituting water usage efficiency criteria and constructing resource utilization frameworks. This existing positive stress will be converted to competitive dominance by shifting from passive compliance to proactive restriction in green production.