Does the water resource ‘fee to tax’ policy alleviate water poverty? Evidence from a quasi-natural experiment Open Access

Since many developing countries are affected by water scarcity and climate crises are expected to worsen this issue, it is crucial to understand how increasing threats to water availability might affect future progress in human development and economic take-off (Moshtagh & Mohsenpour 2020; Prabha et al. 2020). Accordingly, water management has become a priority for policy-makers, and a key component of water management strategies is alleviating water poverty (Fito & Van Hulle 2021). Water poverty is a concept that transcends water scarcity and is multidimensional, focusing specifically on the social and economic aspects of water resource management (Yoon et al. 2021). Indeed, among the various tools proposed to address water poverty, the use of the water poverty index (WPI) is noteworthy and holistic (Koirala et al. 2020). The WPI is a multidimensional tool that includes the assessment of water stress situations based on five components: resources, assessment, capacity, environment, and use. Although WPI requires multiple datasets for its construction, it is an improvement over previous indices because it links the estimates of water availability with socio-economic dimensions (Thakur et al. 2017). Recent endeavors to use the WPI have concentrated on regional comparisons between countries, states or sub-districts and on factors influencing water poverty (Sullivan et al. 2006; Jemmali 2017).

Theoretically, the water resources tax policy is meant to contribute to water conservation in the following ways. First, since a water resources tax will increase the tax burden on each entity's water consumption, it can stimulate entities to adopt new water-conserving technologies, which has long been seen as a powerful approach to conserve water and change the water utilization pattern in high-consumption fields (Benyezza et al. 2021). Second, taxation's mandatory nature improves its efficiency, increases tax consciousness, and causes each unity to recognize the scarcity of water resources via the publicity of the water situation and tax knowledge by relevant departments, ultimately increasing consumer awareness of sustainable water consumption (Gómez-Llanos et al. 2020). Lastly, the levied tax's revenue can be invested in water resource protection and water body restoration via the creation of special accounts and using the special tax fund supply to further protect water resources and conserve water. In contrast, several studies argue that the water resources tax increases the taxpayer burden and is not conducive to water conservation (Kilimani et al. 2015). Literature shows that the existing research results remain controversial. However, knowledge gaps exist about the extent to which the water resources policy can alleviate water poverty, and the theoretical mechanism remains unclear and requires quantification.

Difference in differences (DID) is a commonly used quasi-experimental approach to evaluate the causal effects of specific public policies. Since the implementation of a public policy is usually not influenced by the subject, the policy can be regarded as an exogenous ‘intervention’ to the subject; therefore, the implementation of the policy can also be recognized as a quasi-experiment (Salinas & Solé-Ollé 2018). The fundamental purpose of the DID approach is to comprehensively investigate the differences amongst the differences both with and without and before and after the enforcement of certain policies (Ashenfelter 1978). The DID model estimates the net effect of policy implementation by comparing the differences in dependent variables between the experimental group and the control group before and after the implementation of such policies (Hu et al. 2022). In recent years, increasing numbers of scholars have applied the DID model to conduct causal analysis of public policy effects, such as Moser & Voena (2012), Chagas et al. (2016), and Kosfeld et al. (2021). China's implementation of the water resources ‘fee to tax’ policy provides a fitting case for the scientific evaluation of water poverty's net effect. Therefore, on the basis of the multistage DID model, this study intends to explore three key issues: (1) Does nationwide water poverty exhibit any spatial and temporal characteristics? To what degree does the water resource ‘fee to tax’ policy affect water poverty? (2) What are the mechanisms by which the water resources tax policy affects water poverty? (3) What are the regional differences in and results of heterogeneity analysis of the water resources tax policy?

Water scarcity and uneven spatial distribution of water resources have become two of the major obstacles to sustainable economic development in China. Additionally, excessive extraction of groundwater resources for irrigation to meet the needs of a growing population and ensure food security has largely exacerbated water poverty (Westmore 2018). Therefore, in China's distinctive political system, the central government has adopted a ‘two-handed’ approach to alleviating water poverty that includes a combination of strong central regulation and infrastructure development and the adoption of market principles (Svensson et al. 2021). Economic leverage is the most efficient market tool. Accordingly, following the promulgation of the Water Law of the People's Republic of China in 1988, the water resource fee began to be collected from all citizens. Due to low overall water prices, the ambiguous responsibilities of relevant sectors, and the nonstandard utilization of water resource fees, the current policy of collecting and paying water resource fees can no longer meet national development needs. China is currently reforming the 32-year-old water resources fee policy and plans to replace a more lenient water fee with a stricter water tax; thus, the work of the ‘water resources fee to tax reform’ was implemented.

In July 2016, the Ministry of Water Resources (MWR) of China issued the ‘water resources tax policy’ in Hebei Province to increase water use efficiency and alleviate water poverty. Since the water resource tax reform was implemented, unreasonable water demand has been effectively reduced in Hebei Province, which proves that the water resource tax policy has the basis and conditions to expand. Therefore, in November 2017, the scope of the tax reform was expanded, and the ‘1 + 9’ pilot pattern was formed. These nine provinces are in the eastern (Beijing, Tianjin, and Shandong), central (Henan and Shanxi) and western regions (Inner Mongolia, Ningxia, Sichuan, and Shaanxi) (Figure 1) of China. These representative regions encompass economically developed areas and underdeveloped areas, and the regions show large differences in water resource endowments, major grain-producing areas and nonmajor grain-producing areas. The Chinese government regards the water resources tax policy as a crucial economic incentive to decrease water consumption. However, has the ‘fee to tax’ policy truly alleviated water poverty? If so, to what extent? What are the mechanisms and heterogeneous differences in the effects of water poverty? This study will address these issues.

Price mechanisms are an important tool and means of market regulation (Ergunova et al. 2018). By charging money for resources, we can effectively solve the problem of resource externality, fully protect resources, and promote rational resource utilization (Hotelling 1931; Eisenack et al. 2012; Ing 2020). Levying the water resource tax increases the cost of water utilization and, thus, affects the water intake and use behavior of enterprises and the public (Tian et al. 2021). Therefore, the implementation of the policy of water resources tax will help enterprises and the public develop water-conserving awareness, reduce water consumption, promote the sustainable utilization of water resources, and lastly, reduce water poverty. Based on the above analysis, the first theoretical hypothesis is proposed:

Hypothesis 1: The water resources tax policy can effectively reduce water poverty.

Water resource tax policy may reduce water poverty via several channels. First, groundwater extraction should be controlled. Tax policy can promote the management of groundwater resources (Lorphensri et al. 2011; Wang et al. 2020). The water resources tax policy brings groundwater into the scope of taxation. More importantly, the tax standard for groundwater water resources is higher than that of surface water when the same amount of water is used. Therefore, the reform of the water resource tax increases the tax burden between groundwater and surface water and overexploited and underexploited areas, which is conducive to protecting water resources and alleviating water poverty.

Second, the water use structure should be optimized. Using the difference of tax burden, the water resources tax enables enterprises to adjust water use structures, reduce the proportion of groundwater use and promote the sustainable utilization of water resources. In addition, the proportion of agricultural and industrial water in China is excessive. The water quota specified in several provinces is converted according to the effective utilization coefficient of farmland irrigation water, which is intended to encourage farmers to conserve water. For industrial water, the pilot policy imposes a higher standard on general industrial water than domestic water and agricultural water and doubles the water resource tax for water use exceeding the quota. The objective is to optimize the structure of water use and improve the efficiency of industrial water use, resulting in the improvement in water poverty conditions (Wei et al. 2018). Based on the above analysis, the second theoretical hypothesis is proposed:

Hypothesis 2: The water resources tax policy can reduce water poverty by controlling groundwater extraction and optimizing the structure of water use.

The policy may have clear regional differences in the alleviation level of water poverty. The water poverty situation in the western region of China is more severe than that in the central and eastern regions, where economic and social development is being confronted with a bottleneck caused by water shortages. In addition, the western region is located in an arid zone that has low annual precipitation, agricultural water use efficiency, and high levels of waste. Moreover, the taxes in the western region are generally lower than that in the central and eastern regions, which reflects the government's emphasis on using tax leverage to encourage industries to increase water utilization efficiency (Gao & Yin 2016). Therefore, compared with the western region, the implementation of the water resources tax policy may more effectively reduce water poverty in the central and eastern regions. Based on the above analysis, the third theoretical hypothesis of this study is proposed:

Hypothesis 3: The water resources tax policy will more strongly mitigate water poverty in the eastern and central regions than in the western region.

The WPI method is used to describe the water poverty situation in 31 provinces (municipalities) in China. The WPI encompasses five key dimensions, namely, resources, access, capacity, use and environment. Each component corresponds to different indicators. Table 1 shows the indicators under each component.

Resource: Resource is broadly defined as the total amount of available water in a particular country or region. Here, we use the number of total water resources (R1), the water resources per capita (R2), and precipitation in each province (municipalities) (R3) as three indicators to capture the resource component.

Use: The use dimension typically encompasses the quantity of water used by different water use sectors. In China's water use statistics, water use is usually divided into four categories: agriculture, industry, life, and ecology. Therefore, we select the following four indicators to represent the use components: total agricultural water use (U1), total industrial water use (U2), gross domestic water consumption (U3) and total ecological water use (U4).

Access: Access to sufficient water is essential for production and life. In this study, crop irrigation water, safe domestic water and maximum water supply are considered, and the cultivated land irrigation area (A1), urban water penetration rate (A2) and comprehensive production capacity of water supply (A3) are selected for measurement.

Capacity: Capacity represents the purchasing and management power of water. The financial self-sufficiency rate (C1), local scientific and technological financial support (C2), average number of college students per 100,000 people (C3) and Engel coefficient (C4) are selected as the indicators of capacity. C1 refers to the ratio of local fiscal revenues to local fiscal expenditures, C2 denotes the proportion of local science and technology expenditure in total financial expenditure, C3 represents the education level of the population, and C4 reflects the living standard of the citizens of the province (city), which is used to measure citizens’ ability to buy safe water.

Environment: The environmental component attempts to evaluate the degree of environmental integrity by analyzing the ecological or human factors that may affect the quantity or quality of water in the living environment. Daily urban sewage treatment capacity (E1), park green area per capita (E2), amount of agricultural fertilizer (E3) and area for waterlogging prevention and soil erosion control (E4) are selected as indicators of the environment.

To avoid the arbitrariness of a subjective weighting method, the entropy value method is implemented (Song et al. 2017; Wu et al. 2021). The following steps are applied to decide the weights of each indicator.

Step 1: Decision Matrix

Step 2: Standardization

Step 3: Probability and Entropy

When p_ij = 0, p_ij⋅lnp_ij = 0. In addition, E_i lies in the [0, 1] domain.

Step 4: Entropy Weights

The WPI is recognized as the dependent variable, which is calculated by the entropy weight method. The core independent variable is the water resources tax policy, which is presented by a dummy variable. That is, if province (municipality) i is a pilot area in year t, the value is 1; otherwise, the value is 0. In the model, we also need to control other variables that affect water poverty, including the economic development level (PGDP), water consumption per capita (PWC), population size (POP), urbanization rate (URBAN), per-acre water consumption of arable land (AGRI), water consumption per 10,000 yuan of industrial added value (INDU), and fixed effects of time and provinces (municipalities).

The final sample of this article includes 341 observations (31 provinces from 2009 to 2019) based on data from the ‘China Statistical Yearbook’, ‘China Rural Statistical Yearbook’, ‘China Environmental Statistics Yearbook’, ‘China Water Conservancy Statistical Yearbook’, ‘China Water Resources Bulletin’ and National Bureau of Statistics of China. Table 2 displays the definitions and descriptive statistics for all variables. All the data were analyzed using Stata 15.

Tables 3 and 4 indicates the average value of water poverty in the pilot and non-pilot areas from 2009 to 2019 and shows the changing trend of water poverty by comparing the conditions before and after policy implementation. Before the implementation of the water rights trading pilot policy, the average water poverty value in most pilot areas was lower than the average in the non-pilot areas. After the policy was implemented, the water poverty value in the pilot areas increased significantly, and the water poverty value in non-pilot areas decreased. Moreover, the gap between the pilot provinces and the national average gradually narrowed. This finding demonstrates that compared with non-pilot areas, the implementation of the water resources tax pilot policy has significantly alleviated water poverty in the pilot areas. However, further conclusions must be obtained using an econometric model.

The results in Table 5 imply that the coefficient sign of the variable ‘Treat’ (estimated quantity of the water resource tax policy) is positive and significant at the 1% level, which indicates that the water resources tax policy can effectively mitigate water poverty. The results for the coefficient of Treat show that after policy implementation, the water poverty value of the pilot provinces (municipalities) increases by an average of 0.9 percentage points compared with the non-pilot provinces (municipalities), which validates Hypothesis 1.

The results of the control variables show that the economic development level, population size, per capita water consumption and urbanization rate positively impact water poverty, with significant differences at the levels of 1, 5, and 10%, respectively. Specifically, regions with higher levels of economic development are more capable of improving water poverty via investment in capital and technology. The higher the per capita water consumption in a certain area is, the better the water resource endowment in the area and the more conducive the policies are to eliminate water poverty. A higher urbanization rate results in improved protection of resources and contributes to the reduction of water poverty. In addition, the per-acre water consumption of arable land negatively impacts water poverty, with significant differences at the 1% level. The higher the per-acre water consumption of arable land in a certain area, the lower the efficiency of that area's agricultural water use, and ultimately, the higher the water poverty. Lastly, although the impact of the water consumption per 10,000 yuan of industrial added value on water poverty is negative, the value is insignificant, indicating that it cannot notably affect water poverty.

In the DID model, there should be no significant differences in the severity of water poverty between the experimental and control groups before the implementation of this policy. Therefore, conducting the following parallel trend test is necessary.

Model (10) is used to test the impact of the water resources tax policy on water poverty. If the coefficient α₁ is significant, Model (11) can be used to test the influence of the independent variable (Treat) on the intermediate variable (D). If the coefficient β₁ is significant, Model (12) is used to include both the independent variable (Treat) and intermediate variable (D) for analysis. A significant coefficient γ2 and insignificant γ1 indicate a complete intermediary effect. However, a significant coefficient γ₂ and the coefficient γ₁ indicate a partial mediating effect. An insignificant coefficient γ₂ denotes an untenable mediating effect.

Table 6 reveals how the policy affects water poverty under the action of the three mechanism variables. The mechanism of the groundwater supply in Column 1 explores the independent variable (Treat) on the dependent variable, and the estimated coefficient is 0.009, which is significant at the 1% level. Column 2 in Table 6 displays the impact of the independent variable (Treat) on the intermediate variable (Supply). The coefficient is significantly negative at the 1% level, indicating that the implementation of the policy reduces the exploitation of groundwater. In Column 3, the estimated coefficients of the intermediate variable (Supply) and independent variable (Treat) are significant at the levels of 1 and 5%, respectively. The decrease in the significance level of the independent variable (Treat) proves that groundwater exploitation plays a partial intermediary role between the water resources tax policy and water poverty. Similarly, the mechanism of the water use structure is shown in Columns (4)–(9), where (4)–(6) represent the water use structure of different water sources and (7)-(9) represent the water use structure of different water use departments. Both significance levels of the independent variable (Treat) decreased, which indicates that the water use structure plays a partial intermediary role and that the mitigation effect of the water resources tax policy on water poverty increases as the water structure is optimized. Based on the above regression results, Hypothesis 2 is verified.

The regression results in Table 7 reveal that the influence of water resource tax policy on water poverty in the western region is insignificant compared with those in the eastern and central regions. The impacts on water poverty in the eastern region and central region are positive and significant at the levels of 1 and 5%, respectively. Furthermore, the regression coefficients in Columns (1) and (2) of Table 7 demonstrate that the policy increased the water poverty levels in the experimental group consisting of the eastern and central regions by an average of 1.2 and 1.6%, respectively. The policy effect was clearer. Based on the above empirical test, Hypothesis 3 has been verified.

Previous studies have shown that, although the rapid development of China's economy has improved the availability of water resources and the regional distribution of water resources, the continuous increase in water demand and pollution emissions has led to the expansion and local accumulation of China's regional water poverty (Zhou et al. 2020). Therefore, a method of alleviating China's water poverty and realising the sustainable development of water resources is urgently required. The practical experience of foreign water resources tax shows that implementing a water resources tax policy can increase the water-conserving awareness of enterprises and individuals. The policy effectively promotes the rational utilization of water resources, improves water efficiency and protects water resources (Xu 2020). Based on foreign water resource management policies, the Chinese government has formulated and adapted a water resources tax for China.

This research uses a multistage dynamic DID benchmark regression model to test Hypothesis 1. The findings show that the water resource tax policy can effectively reduce water poverty and that the water poverty value of the pilot provinces (municipalities) increases by an average of 0.9 percentage points compared with other non-pilot provinces (municipalities). This conclusion is consistent with the results of similar studies and further proves the thesis that taxing water resources can effectively reduce water poverty (Guo et al. 2019; Yang et al. 2020). The enforcement of a water resources tax can arouse the awareness of water users to conserve water (Agarwal et al. 2000). Additionally, the leverage effect of the water resources tax relatively increases the price of water resources, which leads enterprises and individuals to adjust their habits of using water and to restore the water environment, thus decreasing the occurrence of water poverty (Li et al. 2018).

Hypothesis 2 proposes that the policy of a water resource tax can reduce water poverty by controlling the exploitation of groundwater. Overexploitation of groundwater and pumping cause excessive disturbance of salt stratification in natural groundwater. As a result, groundwater salinization aggravates water poverty (Foster et al. 2013). The research shows that water resource policies reduce the water intake of groundwater and inhibit groundwater exploitation by increasing the tax difference between surface water and groundwater (Ni et al. 2019). Tax differences can foster the replacement of groundwater with more surface water and reduce the pressure of groundwater resources. In addition, the implementation of a water resource tax increases the cost of agricultural water, which is conducive to prompting farmers to abandon flood irrigation and encourage the use of water-conserving irrigation technology such as micro-irrigation and sprinkler irrigation systems (Parween et al. 2021). The waste of agricultural water resources is one of the important reasons for the contradiction between the supply and demand of water resources (Chang et al. 2016). Using water-conserving irrigation technology is conducive to improving agricultural water use efficiency, reducing groundwater intake, promoting sustainable agricultural development and reducing water poverty (Wang et al. 2019).

Moreover, Hypothesis 2 demonstrates that the policy can achieve the goal of reducing water poverty by optimizing the water-use structure. The tax burden increases the cost of enterprise resource development, which forces the transformation and upgrading of these enterprises and promotes the transformation of society to sustainable economic development by encouraging the adjustment of enterprises’ and society's water use structures (Zhang et al. 2020; Ji et al. 2021). Additionally, China has lower utilization efficiency of agricultural and industrial water resources than that in most developed countries (Giordano 2007; Cao et al. 2021). The water resources tax fully uses the tax lever to adjust the water demand and internalizes its tax burden into the operating cost of water users via the price mechanism, which can guide the water-conserving behaviour and avoid the waste of water resources (Tian et al. 2021). Similar studies prove that the implementation of a water resource tax reduces agricultural water use (Zheng & Chen 2021), improves industrial water use efficiency (Zhao & Zhang 2021), and will help to improve China's water use structure. Adjusting the water use structure is conducive to optimally allocating water resources, eliminating the contradiction of water resource utilization, and reducing water poverty (Wei et al. 2018).

Furthermore, this study validates Hypothesis 3 by testing whether the effects of the water resources policy has spatial heterogeneity. The results of spatial heterogeneity analysis show that the policy more significantly impacts the eastern and central regions than the western region. The water poverty value of the eastern provinces (cities) and the central provinces (cities) have increased by an average of 1.2 and 1.6%, respectively. One possible reason is that the central and eastern regions are richer in water resources and concentrated in the industry, and there is great opportunity to reduce industrial wastewater discharge and improve industrial water efficiency (Geng et al. 2014). Moreover, the eastern and central regions are more economically developed, and the location advantage is clearer than that of the western region; therefore, when the tax policy is implemented, these regions become enabled to adopt clean and green water conservation technology, decrease industrial wastewater discharge and promote water resource recycling (Chen et al. 2021). Therefore, the central and eastern regions show more significant effects than the western region in reducing water poverty when applying price leverage.

Based on panel data from 31 provinces (municipalities) in China from 2009 to 2019, this study constructed a multistage dynamic DID model. We explored the ‘net effect’ of the water resources tax policy on reducing water poverty. The WPI was investigated to evaluate water poverty based on five components: access, resources, use, capacity and environment. The effective governance of water poverty depends on price leverage, and we found that the water resources tax policy can mitigate water poverty. The tax policy alleviates water poverty by controlling groundwater exploitation and encouraging water users to improve water-conserving technology to adjust and optimize their water use structures. In addition, due to differences in economic development, geographical location and resource endowment, the water resources tax policy more significantly impacts the suppression of water poverty in the central and eastern regions than in the western region. The results are expected to provide a reference for policy-makers in water poverty management governance and increase the sustainable development of water conservation businesses in areas with water shortages.

This study provides useful information about the relationship between the water resources tax policy and water poverty, which can lead to the following policy implications. First, in the context of ecological civilisation construction, water resource taxes should be highlighted in the tax structure, and the role of taxes in promoting water resource conservation and improving water poverty should be advanced. A water resource tax should be used to efficient input and output of the water resource-based industry and increase the value of the water industry. Second, a water resources tax policy should be implemented accurately and effectively. Due to differences in water resource endowments and economic development in China's provinces (municipalities), the water poverty problems encountered by various regions will differ. Therefore, the implementation of policies should focus on local conditions, and each region should choose the appropriate path that suits local conditions to fully exploit the role of the water tax transaction policy in reducing water poverty.

This work was financially supported by the National Natural Science Foundation of China (42001221); Shaanxi Provincial Innovation Capability Support Program (2022KRM067); Shaanxi Social Science Fund (2022D049); the Fundamental Research Funds for the Central University (GK202103119); the Fundamental Research Funds for the Central University of Shaanxi Normal University (2022ZYYB25).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.