Next Article in Journal
Construction-Induced Waterlogging Simulation in Pinglu Canal Using a Coupled SWMM-HEC-RAS Model: Implications for Inland Waterway Engineering
Previous Article in Journal
Evaluation of Groundwater Quality and Health Risk Assessment During the Dry Season in the Xin’an River Basin, China
Previous Article in Special Issue
Water Rights Trading and Agricultural Water Use Efficiency: Evidence from China
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Does Water Rights Trading Improve Agricultural Water Use Efficiency? Evidence from a Quasi-Natural Experiment

1
College of Economics and Management, Huazhong Agricultural University, Wuhan 430070, China
2
School of Business, Jiangsu Ocean University, Lianyungang 222005, China
3
College of Public Administration, Zhongnan University of Economics and Law, Wuhan 430073, China
*
Author to whom correspondence should be addressed.
Water 2025, 17(16), 2414; https://doi.org/10.3390/w17162414
Submission received: 8 July 2025 / Revised: 12 August 2025 / Accepted: 13 August 2025 / Published: 15 August 2025

Abstract

Global water scarcity has emerged as a critical barrier to sustainable socio-economic development, stimulating water rights trading to serve as a policy instrument designed to enhance water use efficiency. This study systematically evaluates the impact of water rights trading (WRT) on agricultural water use efficiency (AWE) using panel data from 30 provinces (2011–2022) and a difference-in-difference (DID) model, while thoroughly investigating the underlying mechanisms and spatial spillover effects. The following are primary conclusions: (1) WRT significantly improves efficiency, reducing water consumption per unit of agricultural output by 4.5% in pilot regions, with robustness checks confirming reliability; (2) the policy’s effects on agricultural water use efficiency vary across regions; (3) mechanism analysis suggests that efficiency improvements are primarily driven by optimized crop planting patterns, adoption of water-saving irrigation technologies, advancements in agricultural mechanization, and strengthened environmental regulations; and (4) the policy exhibits notable spatial spillover effects. These findings contribute to the evaluation of WRT policy and offer practical insights for market-based water allocation reforms, suggesting further expansion of WRT with an emphasis on regional coordination and cross-regional cooperation mechanisms.

1. Introduction

Water is a natural resource that is essential for the survival of humans. With the increasing number of people worldwide and rapid urbanization [1], the scarcity of water has become a global concern [2,3] and has constrained the sustainable development of socioeconomics in most countries [4]. Agriculture is the most water-intensive industry and serves as the foundation of many nations. In China, the agricultural sector consumes more than 60% of national water resources [5]. However, the utilization coefficient of irrigation water in cropland is significantly lower than the global standard of 0.7 to 0.8, with the coefficient of effective utilization standing at 0.572 in 2022 [6]. Drought and water shortages continue to afflict the agricultural sector, characterized by severe mismatches between supply and demand and persistently low water use efficiency [7].
Sustainable water management presents a critical solution to agricultural water scarcity [8,9,10]. To improve water use efficiency, governments worldwide are implementing various policy instruments. Water rights trading, as a market-based mechanism [11,12], encourages water users to trade water resources, generating economic benefits [13,14,15] while optimizing water resource allocation, alleviating water scarcity [16], and improving agricultural water use efficiency. This is an effective path towards sustainable development [17]. The Ministry of Water Resources (MWR) issued the Notice on Piloting of Water Rights in 2014, which stipulates that seven provinces in China will serve as pilot zones. By the end of 2023, the effective utilization coefficient of farmland irrigation water had improved from 0.530 to 0.576.
Existing research has examined the implications of water rights trading from the perspective of property rights theory. They illustrate that water rights trading reorganizes water resource property rights through market-based allocation systems, generating price signals that indicate water shortage [18]. This market mechanism endows water resources with economic good characteristics, raising usage costs and curbing excessive consumption [19]. Moreover, agricultural water rights trading facilitates the exchange of surplus water among farmers [20], creating economic value while reallocating water from low-efficiency to high-efficiency users [21,22]. In terms of input–output, based on the Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA) models, Razzaq et al. (2019) discovered that farmers in Pakistan can improve the water use efficiency by participating in water rights trading [23]. To alleviate slacks in inputs and outputs, Deng et al. (2025) enhanced the DEA model by using SBM model to calculate the water use efficiency [24]. In terms of the underlying mechanism, research has demonstrated that optimizing agricultural crop structures can reduce irrigation water consumption [25]. Machine learning techniques were used by Akbari et al. (2022) to show that groundwater use might be decreased by substituting winter barley for winter wheat and cotton [26]. In a similar vein, Osama et al. (2017) showed using linear programming models that improving crop structures in the face of water constraints increases farmers’ financial gains [27]. Recent advancements integrate crop structure optimization with water rights trading policies, offering a dual approach to sustainable water governance. Dong et al. (2024) found that water rights trading systems raise production costs for water-intensive crops [28]. Consequently, farmers are motivated to reassess the economic yields of their cropping patterns and modify their planting strategies to optimize water resource utilization [29,30]. Wang et al. (2021) indicated that water rights trading rules incentivize farmers to implement efficient irrigation technology [31], including sprinkler and drip irrigation systems. This can control water application based on crop growth stages and soil parameters [32,33,34], hence minimizing inefficiencies [35] compared to conventional methods [36].
Although existing studies have made contributions to the policy implications of water rights trading systems, current research reveals notable limitations. First, most studies focus on the influence of water rights trade on regional general water use efficiency, with few studies focusing on specific industries. Among those studies that investigate the effects of water rights trade on agricultural water-use efficiency, the processes are primarily focused on technology innovation and crop structure changes, resulting in limited analytical framework that requires further expansion. For example, agricultural mechanization can transform production modes and enhance water-use efficiency [37]; regional environmental regulations strengthen ecological conservation, ensure water supply quality, and provide a secure trading environment for water rights [38,39]. Additionally, while DEA and SFA are commonly employed for water use efficiency calculations, these methods have inherent limitations. Specifically, their mathematical models fail to account for crucial environmental variables, thereby limiting the efficiency assessment. Finally, potential spatial spillover effects of water rights trading policies remain largely unexplored.
Therefore, to fill these research gaps, this study is based on a DID model and uses panel data of 30 provinces from 2011 to 2022. The analysis aims at gauging the heterogeneity and spatial spillover of the policy and examining the mechanism of influence of the policy on agricultural water use efficiency and thus provide evidence-based policy suggestions. Marginal contributions of the study are in the following dimensions:
(1)
This paper focuses on agriculture, with particular attention to the impact of water rights trading on agricultural sector’s water use. By doing this, it expands our understanding of agricultural water policy.
(2)
This study uses a new indicator, “water consumption per unit of agricultural output value,” which represents the amount of water required per unit of agricultural output. This gives a direct reflection of economic value of water use and, when paired with a DID framework, allows for a precise assessment of policy impacts. This approach also avoids potential biases associated with efficiency measurement methods like DEA and SFA, which depend on model specifications and weight selection.
(3)
This paper thoroughly explores the spatial spillover effects of water rights trading policies using spatial econometric methods. Currently, there are few studies that have considered the spatial spillover effects of water rights trading policies, so investigating the spatial spillover effects will help fill this research gap.
(4)
This paper not only examines the effects of such trading on agricultural water efficiency, but also offers an exploration of the underlying mechanisms. In doing so, it addresses the gaps in mechanism identification and lays a stronger foundation for evaluating the effectiveness of water rights trading policies.

2. Research Hypotheses

2.1. Water Rights Trading and Agricultural Water Use Efficiency

The clear definition and initial distribution of property rights over water resources are the cornerstone for the efficient operation of the water rights trading system. This system incentivizes farmers to engage in water resource trading through price signals and financial incentives to achieve the most optimal distribution of water resources [40]. In addition, the establishment of dedicated trading platforms also promotes transparency in the water allocation process [41]. By reducing information asymmetries and facilitating efficient exchanges, it strengthens market performance in resource allocation. In this way, the water rights trading system not only offers institutional support but also provides market-driven incentives for the enhancement of agricultural water use efficiency. Based on the above theory, we formulate the following hypothesis:
H1: 
The implementation of water rights trading enhances regional agricultural water use efficiency.

2.2. Mechanisms of Water Rights Trading Affect Agricultural Water Use Efficiency

First, optimizing crop structure is an important way to improve agricultural water use efficiency. Farmers are often unable to maximize their financial gains from traditional water-intensive crops due to their low economic yield per unit of water resources [42]. Therefore, water rights trading will motivate producers to compare the water input requirements and economic outputs of various crops and adjust their planting structures to achieve more efficient use of water resources. Crop structure optimization and adjustment not only contribute to increased land productivity [43], but also promote the economic efficiency of agricultural water usage and encourage the upgrading of the agricultural industrial structure. This paper proposes the following hypothesis:
H2: 
Water rights trading enhances agricultural water use efficiency by promoting crop structure adjustment and reducing the cultivation of water-intensive crops.
Second, intelligent water-saving irrigation should be promoted, which possesses significance for modern agricultural development [44]. More importantly, farmers can obtain economic benefits through the trading market, which strengthens their incentive to conserve water [45,46]. Therefore, farmers are motivated to adopt water-conserving irrigation technology to decrease production expenses and increase profitability, while minimizing irrigation water wastage and enhancing water use efficiency.
H3: 
Water rights trading promotes improvements in agricultural water use efficiency by facilitating the adoption of water-saving irrigation technologies.
Third, the implementation of a water rights trading system supports the advancement of agricultural mechanization. The market-based water allocation mechanism has motivated agricultural producers to improve production efficiency [47] in farm machinery and equipment. Agricultural mechanization has increased irrigation efficiency while also enhancing field preparation quality, soil structure, and water retention ability. Furthermore, it has accelerated technology developments in agricultural production systems [48], laying the groundwork for precision agriculture [33], and causing a fundamental shift in farming techniques. Considering these observations, we propose the following hypothesis:
H4: 
Water rights trading enhances agricultural water use efficiency by fostering higher levels of agricultural mechanization.
Finally, local environmental regulations also have practical significance. The water rights system requires that traded water resources meet ecological sustainability standards [49]. This requirement not only facilitates wider adoption of water rights trading but also establishes an environmentally protective virtuous cycle. On the one hand, water rights trading rules encourage relevant departments to adopt market-based measures to manage water resource development [50], as well as to curb excessive water extraction, unlawful sewage disposal, and other activities that harm the water environment. On the other hand, more stringent environmental regulations have compelled producers to adopt clean production technologies [51,52], and enhance the efficiency of water resource recycling. Environmental regulations have accelerated the development of the environmental protection industry, providing strong institutional support for innovations [53] and sustainable development. Therefore, we propose the following hypothesis:
H5: 
The water rights trading mechanism enhances regional environmental protection, thereby improving agricultural water use efficiency.
To summarize, we contend that water rights trading policies influence agricultural water-use efficiency through multiple pathways. The overall framework of this study is illustrated in Figure 1.

3. Data and Methods

3.1. Empirical Model

3.1.1. DID Model Design

This study used the DID approach to investigate the impact of the water rights trading program on agricultural water use efficiency; the following model can be used:
A W E i , t = α 0 + α 1 W R T i , t + α k X i , t + γ i + δ t + ε i , t
Among these terms, AW E i , t represents the agricultural water use efficiency of region i in year t. WR T i , t is a dummy variable for this paper’s main independent variable, the water rights trading pilot policy. When region i becomes a pilot region in year t, WR T i , t takes the value of 1; otherwise, it is 0. X i , t denotes control variables, γ i represents individual fixed effects, δ t represents time fixed effects, and ε i , t is the random disturbance term.

3.1.2. Impact Mechanism Testing Design

We proposed that the water rights trading policy may enhance agricultural water use efficiency through multiple channels, including optimizing cropping structure, expanding water-saving irrigation areas, promoting agricultural mechanization, and strengthening regional environmental regulation intensity. To investigate these pathways, this study employs a mediation effect model based on Model (1). The model is specified as follows:
M I D i , t = α 0 + α 1 W R T i , t + α k X i , t + γ i + δ t + ε i , t
A W E i , t = α 0 + α 1 W R T i , t + α 2 M I D i , t + α k X i , t + γ i + δ t + ε i , t
M I D i , t represents mediating variable, which includes cropping structure, area of water-saving irrigation, level of agricultural mechanization, and degree of regional environmental protection. Other control variables remain consistent with those used in Model (1).

3.1.3. Spatial Econometric Model Design

To examine potential spatial spillover effects of water rights trading (WRT) policies on agricultural water use efficiency, this study employs spatial econometric modeling. Moran’s I is calculated to test spatial autocorrelation in the variables, with the formula
I = i = 1 n j = 1 n W i j | x i x | | x j x | S 2 i = 1 n i = 1 n W i j
In this formulation, n denotes the total number of regional units in the sample, W i j represents the elements of the spatial weight matrix, X indicates the mean of the attribute values, and S 2 corresponds to the sample variance.
This study employs a spatial Durbin model (SDM) that considers spatial correlation between the dependent variable and explanatory factors, and the model is as follows:
A W E i , t = β 0 + ρ j = 1 n W i , j A W E i , t + β 1 W R T i , t + β 2 X i , t + β 3 j = 1 n W i , j W R T i , t + β 4 j = 1 n W i , j X j , t + γ i + δ t + ε i , t
Among these, β 1 and β 2 represent the impact of water rights trading policies and corresponding control variables in the region on agricultural water use efficiency, while β 3 and β 4 represent the impact of water rights trading policies and series of control variables in neighboring regions through spatial interaction.

3.2. Variable Descriptions

3.2.1. Core Explanatory Variable: Water Rights Trading (WRT)

The Notice of MWR on the Piloting of Water Rights, released in 2014, named seven pilot regions that included Ningxia, Jiangxi, Hubei, Inner Mongolia, Henan, Gansu, and Guangdong. Based on this policy milestone, the year 2014 is referred to as the inception year of the pilot program. Accordingly, the core explanatory variable WRT is specified as a dummy variable: provinces included in the pilot program are assigned a value of 1 (treatment group), while others are assigned 0 (control group).

3.2.2. Dependent Variable: Agricultural Water Use Efficiency (AWE)

In this study, agricultural water usage efficiency is measured by the volume of total water consumed per unit of gross agricultural output. Although contemporary research often employs methodologies such as SFA or DEA to estimate efficiency, these approaches involve complex mathematical modeling and possess certain limitations—for example, the inability to incorporate environmental factors such as soil conditions or humidity. By contrast, gross agricultural output serves as a direct indicator of the economic value generated by agricultural production. Thus, this study adopts the ratio of agricultural water consumption to gross agricultural output as a proxy for water use efficiency This metric is not only more intuitive but also establishes a clearer linkage between water usage and economic returns, thereby offering stronger explanatory utility.

3.2.3. Control Variables

Agricultural water use efficiency is affected not only by farming practices but also by a variety of regional factors, including economic development levels, technological investment intensity, and the quality of public infrastructure. For example, accelerated urbanization and population growth tend to increase aggregate water demand [54], whereas the adoption of water-saving technologies can significantly enhance efficiency [55]. Moreover, regions with a heavier focus on secondary or tertiary industries may devote comparatively less attention to the agricultural sector. Informed by the existing literature, we incorporate a range of urban and regional characteristics as control variables in the models. A detailed description of these variables is presented in Table 1 below.

3.3. Data Resource

This study establishes a panel of data from 30 provinces in China (2011–2022) to explore the mechanisms by which policies of trading water rights affect agricultural water use efficiency. The Provincial Statistical yearbooks, China Environmental Statistics Yearbook, and China Rural Statistical Yearbook are the major data sources. To address incomplete information caused by missing values, observations with missing data were excluded from the analysis. Furthermore, all continuous variables were winsorized at the 1st and 99th percentiles to reduce potential bias from extreme values. Following the data cleaning procedures, we conducted descriptive statistical analysis on the final sample and the statistics presented in Table 2. It exhibits sufficient dispersion of all variables to justify the proposed econometric specifications.

4. Empirical Results and Analysis

4.1. Parallel Trend Test

Before employing the DID approach, it is essential to validate the parallel trend assumption, which posits that before the implementation of the policy, the treatment and control units have exhibited similar paths. This research employs an event study analysis, utilizing a series of time-relative dummy variables that indicate the number of years preceding and after the implementation of the policy, rather than the original water rights trading variable. The basic model is then re-specified as the Equation (6) to be analyzed. This procedure is performed for two main reasons; first, it allows the evaluation of the dynamic effects and persistence of the water rights trading policy through time, and second, it gives a sound test of the parallel trend assumption. The specification is as follows:
I W E i , t = α 0 + k = 1 3 r k P k + j = 0 4 ω j L j + θ k X i , t + γ i + δ t + ε i , t
In Model (6), P k represents the k-th year before implementation of the water rights trading policy in each of the regions and it is the independent variable that is used to test whether the group getting the treatment and the non-selected control group followed the same trend in the years before the implementation. When r k is not found to be statistically significant, it then shows that no statistically significant difference existed between the treatment group and the control group prior to the implementation of the policy. The effects of water rights trading policy can be identified by means of L j , which represents the j-th year since the adoption of water rights trading policy. The meanings of the other symbols are identical to those in Equation (1).
Figure 2 shows that there are no significant differences between the treatment and control groups prior to policy implementation, confirming the parallel trend assumption. After the policy, the water use efficiency in the agricultural system in the treatment group drastically decreases as compared to those in the control group. Furthermore, the effects of the policy grow first and then settle down slowly implying the time lag reaction towards the introduction of the system of water rights trading. Overall, the outcomes of the parallel trend test give strong support to the further use of the DID framework to make a causal inference, but also give the initial evidence that the policy of water rights trading can help improve the efficiency of water use in agriculture.

4.2. Baseline Regression Results

The impact of water rights trading on agricultural water use efficiency will be assessed by baseline regression. Results are reported in Table 3, together with the estimation from four model specifications. In column (1) the only effect consists of province-fixed effects and no control variables. Column (2) goes further to introduce year-fixed effects in the specification. The coefficients in the two models are both estimated as statistically significant at 1% and negative, indicating that introduction of water rights trading leads to a drop in water use per unit of gross agricultural production. To account for possible problems of omitted variable bias, further control variables are added in columns (3) and (4), with the latter allowing both fixed effects of province and year from the observations.
In the findings, the core explanatory variable, water rights trading, is negative and statistically significant with a probability of 1% in all the model specifications. This implies that the benefit of implementing the water rights trading policy is to reduce water consumption per unit of gross agricultural production, and this improves agricultural water use efficiency. Particularly, the coefficient (−0.045) means that the pilot policy enhances agricultural water efficiency by the average of 4.5%.
Concerning the control variables, the coefficient of the industrial structure is significantly positive in columns (3) and (4) and the industrial structure indicator indicates that higher percentage of the tertiary sector may be an impediment to advances in water use efficiency. This can be due to the past shift of focus and investment on other areas except the agricultural sector in response to the growth of services. Secondly, openness to international markets has a positive coefficient in row (4), and it is possible that as it becomes more open to external markets, it will have more exports of water-intensive agricultural products hence causing water use efficiency to contract. On the other hand, the level of transport infrastructure reflects high negative influences on the water use efficiency of agriculture, and it can be assumed that around −0.168 is the value of the coefficient of this impact. This also notes how a better transportation system would help in the dissemination of water saving technologies and the benefits in the balance of agricultural resources. The baseline regression results are a clear demonstration that water rights trading policy can play a substantial role in the improvement in agricultural water use efficiency.

4.3. Robustness Tests

4.3.1. PSM + DID

The issue of non-random policy assignment might be a problem related to the traditional DID estimation method. Thus, it is a PSM-DID based study that may be regarded as effective since it does not only account for systematic variation in observable characteristics between treatment and control groups, but also removes the impact of unobservable factors that are unsystematically distributed over time. Consequently, it would be possible to accurately measure the policy effect of water rights trading on a proper basis.
The Table 4 showed that before matching there was a significant difference between the groups on various variables. The standardized deviations of variables, in the unmatched sample, were high, which included the values of per capita GDP, industrial structure, level of openness, and R&D intensity. But at the stage of PSM application, such deviations became considerably lower, and no significative difference was left between groups. This helps to determine the efficacy of the matching procedure thus bringing a rough foundation to the following DID analysis. The matching procedure eliminates the selection bias as well as makes the claim of causality more plausible. After that, the findings of PSM-DID also confirm the effect of water rights trading on agricultural water use efficiency. Shown in column (1) of Table 5, following the treatment of selection bias with the matching, the coefficient of the water rights trading variable (wrt) is −0.045, and significant at the 1% level. This suggests that the water rights trading pilot program improved the effectiveness of agricultural water use. Furthermore, the result also agrees with the estimates of the baseline regression in magnitude and statistical significance, hence further complementing the soundness of the conclusions.

4.3.2. Excluding the Influence of the Water Resource Tax Reform

Water usage cost have gone up because of tax reforms affecting water resources. Since water rights trading and the reform to place the primary responsibility on taxation of the fees used to manage water resources can occur concurrently, and both processes can have an impact on the agricultural water use efficiency, there is a risk that the true value of the market-based water rights trading scheme will be underestimated if the impact of the tax reform is not identified. This introduces bias with regards to the assessment of its impact. To mitigate this possibly confounding factor, this research incorporates a control variable of the water fee-to-tax reform (wft). Column (2) of Table 5 shows that the coefficient of the water rights trading variable (wrt) is −0.046, which is not only significant at the 1% level, but also indicates that the more the trading of water rights the less the water is used. In the meantime, the coefficient of the water fee-to tax-reform variable is non-significant. This implies that the water resource tax reform does not confound the impact of water rights trading on agricultural water use efficiency, hence supporting the findings.

4.3.3. Excluding Certain Samples

The endowment of resources in the region might affect the efficiency of policy implementation. Consequently, due to the distinct administrative and economic characteristics of municipalities governed by the central government, Model (2) excludes the four cities (Beijing, Shanghai, Tianjin, and Chongqing) as a robustness check. As reported in column (3) of Table 5, the findings indicate that the coefficient of the water rights trading variable is −0.042 and significant at the 5% level, which means that the result does not change after the municipalities are excluded. Moreover, COVID-19 brought in external shocks to agricultural production. To decrease the risk of this possibly confounding factor, the study removes the data from years 2019 and after. Results corresponding to this are given in column (4). The water rights trading variable in this specification also has a coefficient of −0.041, which is significant at the 1% level. In sum, all the four robustness checks yield the coefficient of the water rights trading variable that is remarkably similar in sign and magnitude, falling within −0.036 and −0.046. These results solidly support the invulnerability of the positive effects of the policy on agricultural efficiency of water use.

4.3.4. One-Period Lag

Lastly, with the possible non-immediate impact of implementation of policies, the treatment variable is lagged by a period in column (5) in Table 5. The findings indicate that the lagged coefficient of the water right trading variable is −0.036 and significant at the 1% level, thus indicating that the policy has a lasting impact in the longer-term.

4.3.5. Placebo Test

In the analysis, observable systematic differences across provinces have been controlled for, and the potential confounding influence of other policies has been excluded. However, a question remains: could the observed policy effect still be driven by unobservable factors? To address this concern, a placebo test is performed by creating a series of counterfactual situations to reassess the baseline outcomes. Specifically, fictitious policy treatment variables are randomly generated to simulate scenarios in which the water rights trading policy was assigned arbitrarily. If improvements in agricultural water use efficiency are observed under these false treatments, it would suggest that the effect may be due to unobserved influences rather than the actual implementation of the policy.
The purpose of placebo test is to determine whether external factors have an impact on the policy effect by randomly creating pseudo-treatment variables depending on the treatment variable. This method helps determine whether the estimated treatment effect arises from true causality or is the result of model misspecification or random chance. The Figure 3 presents the distribution of the placebo test results: the horizontal axis represents the coefficients derived from random assignment, the left vertical axis shows the density, and the right vertical axis indicates the corresponding p-values. The results show that the coefficients from the randomized samples are mostly concentrated within the range of −0.010 to 0.010, forming a bell-shaped curve centered around zero. The density plot illustrates that most of these pseudo-treatment effects are close to zero. Meanwhile, the scatter plot of p-values indicates that most values lie above the 0.05 threshold, suggesting that the simulated effects are largely statistically insignificant. These findings suggest that with random treatment assignment, it is improbable to achieve substantial outcomes akin to the observed impact. This provides compelling evidence that the impact of the water rights trading policy on agricultural water use efficiency reflects a genuine causal relationship, rather than being driven by random variation or model specification errors.

4.4. Heterogeneity Analysis

4.4.1. Water Resource Allocation

As a key determinant of agricultural water use efficiency, water resource allocation directly influences the degree of regional water scarcity and the strength of incentives for water-saving behavior. To explore the heterogeneous effects of the policy, this study divides the sample into two groups—water-abundant and water-scarce regions—based on the annual median of per capita water resources. Grouped regressions are then performed to assess how the impact of the water rights trading policy varies across different levels of water resource allocation. This contributes to a deeper understanding of the mechanisms through which the policy operates and the conditions under which it is most effective.
The result in Table 6 demonstrates that the water rights trading policy exerts a more substantial effect in regions with scarce water resources. In these areas, the coefficient of the water rights trading variable is −0.068 and significant at the 1% level. In comparison, the coefficient in water-abundant regions is −0.044, significant at the 10% level. A formal test of heterogeneity confirms that the difference between the two coefficients is statistically significant. These results suggest that the policy’s impact is considerably greater in water-scarce regions, where the effects are more pronounced than in water-abundant areas.
This differentiated impact primarily arises from the contrasting behavioral responses of agricultural producers under varying water resource endowments. In water-scarce regions, the tighter constraints on water availability heighten producers’ sensitivity to water costs. As a result, the price signals embedded in the water rights trading system are more effective in encouraging water conservation behaviors and prompting adjustments in agricultural practices. Conversely, in regions with abundant water resources, it dampens the policy’s incentive effects. Furthermore, areas facing water scarcity often exhibit a stronger demand for institutional innovation and a greater willingness to adopt policy instruments, which enhances the effectiveness of the water rights trading system.

4.4.2. Agricultural Reliance

Agricultural reliance reflects the significance of the agricultural sector within a region’s economic structure and the extent to which the region relies on agricultural water use. It directly influences the sensitivity and intensity of agricultural producers’ responses to the water rights trading policy. Regions with differing levels of agricultural dependence often display marked variations in water resource allocation, agricultural production modes, and policy receptiveness. Therefore, assessing the policy’s effectiveness is essential to understanding the structural characteristics of its impact. To investigate the effects of the water rights trading policy across regions with different agricultural profiles, this study divides the sample into two groups—high and low agricultural dependence—based on the provincial annual median of agricultural water consumption. A heterogeneity analysis is then conducted to capture the differentiated policy impacts.
The analysis in the Table 7 indicates that the water rights trading policy has a more pronounced effect in regions with high agricultural dependence. In these areas, the coefficient of the water rights trading variable is −0.053 and statistically significant at the 1% level. By contrast, in regions with low agricultural dependence, the coefficient is −0.035 and not significant. The analysis reveals that the policy’s influence on agricultural water usage efficiency is predominantly focused on places where agriculture plays a more substantial role, whereas its effect is comparatively diminished in areas with lesser agricultural dependence.
This heterogeneity stems from fundamental differences in how regions with varying levels of agricultural dependence respond to the water rights trading system. In regions heavily reliant on agriculture, this sector plays a crucial role, generally necessitating significant water usage and a strong dependence on water resources. The implementation of a water rights trading mechanism can more directly influence agricultural production decisions, motivating producers to optimize water use. Farmers in these areas are more responsive to variations in water costs and are more likely to modify cropping structures, employ water-efficient technology, and apply other techniques in response to price signals. Conversely, in areas with lower agricultural dependence, where agriculture represents a smaller share of economic activity, the influence of the policy is less direct and the behavioral response from agricultural producers correspondingly weaker.

4.4.3. Grain-Producing Regions

As critical pillars of the national food security strategy, major grain-producing regions exhibit marked differences from non-grain-producing regions. Grain-producing regions are generally characterized by a predominance of staple crop cultivation and adherence to traditional production methods. In contrast, non-grain-producing regions are more oriented toward cash crops and specialty agriculture, often adopting market-responsive production practices. These divergent agricultural roles shape the ways in which producers in different regions respond to water rights trading policy and influence their capacity to adapt to it. Therefore, it is essential to examine the impacts of the policy across regions with varying agricultural functions. To capture this heterogeneity, the sample is divided into two groups—grain-producing and non-grain-producing regions—based on whether a region is officially designated as a national major grain-producing area.
The findings in Table 8 reveal that the water rights trading policy exerts a more pronounced effect in non-grain-producing regions. In these areas, the coefficient for the water rights trading variable is −0.062, significant at the 1% level. By contrast, in grain-producing regions, the coefficient is −0.037 and fails to achieve statistical significance. These results suggest that the policy’s effect on agricultural water use efficiency is primarily concentrated in non-grain-producing regions, whereas its impact in major grain-producing areas remains limited. This differentiated effect highlights the substantial variation in the mechanisms through which the policy takes effect. Non-grain-producing regions have more diverse agricultural structures and higher crop selection flexibility; thus, producers are better positioned to respond to price signals encoded in the water rights trading system. They can increase water efficiency by modifying planting patterns and implementing water-saving farming practices. Furthermore, agricultural production in these areas is typically driven by economic incentives, resulting in a greater flexibility to market-based resource allocation methods. In contrast, major grain-producing regions are responsible for ensuring national food security, which limits cropping structure changes. Agricultural producers in these areas tend to make production decisions that are more influenced by policy directives, resulting in a relatively limited response to the water rights trading system.

4.4.4. Level of Economic Development

To investigate the impact of economic development levels on the efficiency of water rights trading regulations, this study employs an economic development-based heterogeneity analysis. Using GDP per capita as criterion, the sample is classified into two groups: high and low economic development. The differing consequences of water rights trading schemes in different economic development situations are evaluated independently.
The results in Table 9 show that in high-level economic development areas, the coefficient for the water rights trading variable is −0.047, significant at the 1% level. By contrast, in low-level economic development regions, the coefficient is −0.061. Results indicate that the policy has a more pronounced effect in regions with lower economic development levels. The following reasons may be responsible: in regions with low levels of economic development, the agricultural water use technologies and management levels are relatively backward, which leaves more room for improvement in water resource utilization efficiency. The marginal improvement effects brought about by the introduction of water rights trading systems are more pronounced. In regions with high levels of economic development, water resource management systems may be at a relatively high level prior to the implementation of water rights trading policies, resulting in limited incremental effects. Furthermore, regions with low levels of economic development may exhibit a greater intrinsic motivation to enhance resource utilization efficiency and may respond more favorably to water rights trading as a market-based allocation mechanism. In contrast, regions with high levels of economic growth have generally stable resource consumption patterns and may be more resistant to changing their production practices. Water rights trading, as a market-based allocation mechanism, has been shown to increase efficiency and positive responses. Finally, we use a coefficient visualization chart (Figure 4) to present the results obtained in a more intuitive manner.

5. Further Analysis

5.1. Mechanism Analysis

The empirical results have confirmed that the implementation of water rights trading policies promotes agricultural water use efficiency. To further elucidate the underlying mechanisms, this study employs a mediation effect model to examine the impact mechanisms of water rights trading on agricultural water use efficiency from four key pathways: crop structure, water-saving irrigation area, agricultural mechanization, and regional environmental regulation intensity.
Column (1) of Table 10 shows that water rights trade mediates cropping structure with a coefficient of 0.040, significant at 1%. This shows that the water rights trading policy encourages non-grain crop development. Column (2) demonstrates that cropping structure negatively affects agricultural water usage efficiency with a coefficient of −0.100. Water rights trading has a −0.041 coefficient at 5% after adjusting for cropping structure. These findings support that cropping structure modifications mediate the water rights trading policy’s effect on agricultural water usage efficiency.
Column (3) of Table 10 demonstrates that water rights trade affects water-saving irrigation by 0.116, significant at 5%. This shows that the policy has greatly pushed water-saving irrigation systems. Column (4) shows that water-saving irrigation has a negative influence on agricultural water usage efficiency, with a coefficient of −0.025, significant at a 10% level. The direct effect of water rights trade is strong at 5% level with a coefficient of −0.042. These findings confirm that water rights trade enhances agricultural water usage efficiency by expanding water-saving irrigation. The mediation analysis shows that the water rights trading policy improves agricultural water usage efficiency by restructuring cropping patterns and encouraging water-saving irrigation. This indicates that policy actively optimizes agricultural resource allocation.
Table 11 shows the results for agricultural mechanization and regional environmental control intensity. Column (1) demonstrates that the coefficient of the water rights trading variable (wrt) on agricultural mechanization (agrimech) is 0.844, which is significant at 5%. This suggests that the water rights trading strategy boosts agricultural mechanization. Column (2) shows that agricultural mechanization decreases water consumption efficiency by −0.006 at 10% level. Water rights trading has a −0.039 coefficient and is significant at 5%. These findings show that agricultural mechanization mediates the water rights trading policy’s influence on agricultural water usage efficiency.
In column (3) of Table 11, the coefficient of the water rights trading variable on environmental regulation is 0.003, significant at the 1% level, showing that the policy strengthens regional environmental regulation. Column (4) shows that environmental regulation reduces agricultural water usage efficiency by −2.192. The direct effect of water rights trade is strong at 5% level with a coefficient of −0.039. These data support that the water rights trading regime improves agricultural water use efficiency through increased environmental regulation. The water rights trading policy indirectly enhances agricultural water utilization efficiency by promoting automation and environmental regulation, as demonstrated by mediation analysis. The effects of the policy on agricultural technology and external environmental circumstances are illustrated by these findings.

5.2. Spatial Spillover Effects Analysis

The water rights trading policy aims to create a transparent and well-regulated market that incentivizes individuals, enterprises, and regions to participate in water trades according to their specific usage requirements. It promotes water conservation and minimizes resource wastage among consumers. Beyond facilitating transactions among farmers, the policy allows for inter-regional and cross-sectoral water trading. The continuous relaxation of geographic limitations unavoidably strengthens spatial connections between regions and encourages wider involvement in the water rights trading system, thereby producing geographical spillover effects. These strong demonstration and imitation effects are expected to continually extend the spatial influence of the water rights trading mechanism, further amplifying its spillover impact on agricultural water use efficiency.

5.2.1. Spatial Autocorrelation Test

Prior to assessing the efficiency of water usage on farms in comparison to neighboring farms, it is evident that evaluating spatial autocorrelation is an essential preliminary step for spatial econometric analysis. The study will be using the Global Moran’s I statistic in the determination of the spatial correlation of agricultural water use efficiency in the Chinese provinces between the years 2011 and 2022 informing the need to adopt the spatial econometric models to undertake the analysis.
The Global Moran’s I is frequently utilized as an indicator for assessing spatial connection among regional variables. The values range from −1 to 1, with positive values indicating positive geographical correlation, signifying a propensity for similar agricultural water usage efficiency in adjacent regions, hence suggesting spatial clustering tendencies.
The results Table 12 indicate that the Global Moran’s I value for each year from 2011 to 2022 is consistently positive, ranging from 0.229 to 0.314, and significant at the 1% or 5% levels. These findings provide strong evidence of significant spatial autocorrelation in agricultural water usage efficiency over time—a clear spatial clustering pattern is seen in the tendency of nearby locations to have comparable efficiency levels. From a temporal perspective, the Global Moran’s I demonstrates a declining yet fluctuating trend. The values were relatively high in 2011 and 2012, at 0.306 and 0.314, respectively, but declined to 0.232 by 2021. This suggests that while the degree of spatial dependence in agricultural water use efficiency has diminished over time, the spatial clustering effect remains statistically significant. The significance of the Moran’s I statistic underscores the importance of accounting for spatial effects in the analysis. Ignoring spatial dependence would likely result in biased estimates, thereby necessitating the adoption of spatial econometric models to properly address spatial autocorrelation.

5.2.2. Diagnostic Tests for Spatial Models

As it became obvious that the correlation between spatial factors was found, the most suitable spatial econometric model should be found. In this regard, variability of the diagnostic tests such as the Lagrange Multiplier (LM) test, the Likelihood Ratio (LR) test and the Wald test is used to establish the best model specification.
The LM test in Table 13 results indicate that LM statistics for the SEM and the Spatial Lag Model (SLM) are 200.700 and 144.021, respectively. This provides strong evidence for the simultaneous presence of spatial error and spatial lag effects. The robust version of the LM test further reveals that the robust statistic for the SEM is 57.063 and remains significant at the 1% level, whereas the robust statistic for the SLM is not significant. These results offer preliminary support for the SEM being a more suitable specification for the data. The LR tests comparing the SDM with the SLM and SEM yield test statistics of 29.37 and 45.94, respectively, both significant at the 1% level. These results indicate that the SDM cannot be reduced to either the SLM or SEM. This conclusion is further corroborated by the Wald test, which likewise rejects the simplification of the SDM to either alternative. In addition, the Hausman test produces a statistic of 168.00, also significant at the 1% level, suggesting that fixed effect model is more appropriate. Taken together, these diagnostic tests consistently point to the SDM as the most appropriate method in this study. Nevertheless, regression results from the SLM, SEM, and SDM models are presented.

5.2.3. Spatial Regression Analysis

Table 14 shows the findings from the SLM, SDM, and SEM. The coefficients of the water rights trading variable are −0.043, −0.041, and −0.041. All coefficients are significant at 1% and closely match the baseline regression, indicating the robustness of the findings. The SDM spatial interaction term Wx has a coefficient of −0.024 and is significant at the 5% level, implying that the water rights trading strategy in one region enhances local agricultural water use efficiency and has positive spillover effects in nearby districts. The SLM and SDM have spatial autoregressive coefficients (ρ) of 0.336 and 0.210, respectively, whereas the SEM has a spatial error coefficient (λ) of 0.308. All coefficients are significant at 1%, indicating considerable geographical dependency in agricultural water usage efficiency. In conclusion, the water rights trading strategy improves water usage efficiency in pilot districts and provides positive spillover effects.

6. Conclusions and Policy Implication

6.1. Conclusions

The water rights trading policy creates a market-based mechanism for water resource distribution, encouraging users to conserve water and trade surplus resource, thereby improving total water usage efficiency. As a fundamental water-saving strategy, accurately evaluating its policy effects is essential for refining water resource governance instruments and strengthening governance capacity.
This study uses panel data from 30 Chinese provinces spanning 2011 to 2022. It examines how the water rights trading system affects agricultural water usage efficiency. This study additionally explores the policy’s regional consequences by testing spillover effects. The primary conclusions are as follows: First, the water rights trading scheme improves agricultural water usage efficiency. Pilot provinces had 4.5% greater agricultural water usage efficiency than non-pilot provinces, a conclusion that holds across model settings. Second, regional resource endowment, agricultural dependence and economic development affect the policy’s consequences. Third, the policy enhances agricultural water use efficiency by optimizing crop structures, promoting water-saving irrigation technologies, improving agricultural mechanization, and strengthening regional environmental regulation. Finally, the water rights trading policy has positive spillover effects, implying that advances in one location may assist surrounding places.

6.2. Policy Implication

The above findings present significant policy implications for enhancing the water rights trading framework, developing an effective trading market, and further improving agricultural water use efficiency.
First, the water rights trading policy demonstrates a significantly positive effect on agricultural water efficiency. As agriculture is a cornerstone of China’s economy and water is an essential input in agricultural production, the government should continue promoting the policy while encouraging greater participation from water users. This entails enhancing public education and awareness, addressing practical challenges encountered during policy execution, and incorporating best practices in a way that is consistent with China’s specific conditions. Through continuous exploration and adaptive learning, a water rights trading system and market aligned with local realities should be developed. Concurrently, regulatory supervision must be strengthened to guarantee transparent information disclosure and furnish water users with a dependable trading platform.
Moreover, given that water rights trading influences agricultural water use efficiency through pathways such as crop structure optimization, adoption of water-saving irrigation technologies and stronger environmental regulations, governments should guide farmers in adopting scientifically based cropping practices. In addition, related subsidy policies should be improved, the diffusion of smart water-saving technologies should be promoted, and the level of agricultural mechanization should be enhanced. Efforts must also be directed toward improving environmental governance and pollution control to maximize the effectiveness of the water rights trading policy. Particularly, compared to capital-intensive approaches, crop structure optimization is a more economical adaptation strategy with significant potential for widespread deployment. Therefore, governments ought to prioritize the alteration of crop structures within existing agricultural policy frameworks. Water-intensive crop production may be limited in water-scarce areas, while low-water-use, drought-tolerant crop cultivation may be encouraged. Government agricultural agencies ought to execute focused extension initiatives to advocate for sustainable farming methods among cultivators. These activities should promote the use of crop rotation and intercropping systems, alongside a balanced cultivation of cash and food crops, to alleviate the excessive water usage linked to monoculture practices.
Finally, considering the spatial spillover impacts of the water rights trading policy, governments should systematically document both experiences and challenges in policy implementation. Cross-regional communication and cooperation should be strengthened to establish a normalized feedback mechanism and a long-term collaborative governance structure. Such efforts would facilitate the gradual scaling-up of policy impacts from localized pilot areas to broader regions.
Despite the comprehensive analysis presented in this study, several limitations remain. First, due to constraints in data availability, the analysis relies solely on provincial-level panel data and lacks detailed transaction-level information, such as trading frequency and unit prices. As more granular data—such as county-level statistics—become accessible in the future, further research could be conducted to validate and enhance the robustness of the findings. Second, because some key indicators are only available up to the year 2022, the long-term effects of the water rights trading policy have not yet been fully captured. Future studies should consider using extended time-series data to reassess and further substantiate the conclusions drawn in this research.

Author Contributions

Conceptualization, H.L., B.H. and W.C.; methodology, H.L.; software, H.L.; validation, H.L., B.H. and W.C.; formal analysis, H.L. and B.H.; investigation, H.L.; resources, H.L., B.H. and W.C.; data curation, H.L.; writing—original draft preparation, H.L.; writing—review and editing, B.H. and W.C.; visualization, B.H.; supervision, B.H. and W.C.; project administration, B.H. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the National Social Science Fund of China (22BJY036).

Data Availability Statement

The dataset is available on request from the authors.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Xu, H.; Yang, R. The impact of water rights reform on economic development: Evidence from city-level panel data in China. J. Environ. Manag. 2025, 374, 124082. [Google Scholar] [CrossRef] [PubMed]
  2. McDonald, R.I.; Weber, K.; Padowski, J.; Flörke, M.; Schneider, C.; Green, P.A.; Gleeson, T.; Eckman, S.; Lehner, B.; Balk, D. Water on an urban planet: Urbanization and the reach of urban water infrastructure. Glob. Environ. Change 2014, 27, 96–105. [Google Scholar] [CrossRef]
  3. Shah, W.U.H.; Hao, G.; Yasmeen, R.; Yan, H.; Shen, J.; Lu, Y. Role of China’s agricultural water policy reforms and production technology heterogeneity on agriculture water usage efficiency and total factor productivity change. Agric. Water Manag. 2023, 287, 108429. [Google Scholar] [CrossRef]
  4. Zhang, H.; Zhou, Q.; Zhang, C. Evaluation of agricultural water-saving effects in the context of water rights trading: An empirical study from China’s water rights pilots. J. Clean. Prod. 2021, 313, 127725. [Google Scholar] [CrossRef]
  5. Wang, T.; Lu, C.; Shi, B.; Liang, S.; Lu, Q. Measurement and Spatial and Temporal Evolution Analysis of Agricultural Water Stress in China. China Rural Water Hydrop. 2023, 12, 42–48. [Google Scholar] [CrossRef]
  6. Fengjun, J.; Zhicong, Y.; Zhuo, C.; Li, M. National water network construction for serving the chinese path to modernization: Development direction and strategic approach. Econ. Geogr. 2024, 44, 148–156. [Google Scholar] [CrossRef]
  7. Bao, X.; Chen, H. Water-saving potential in China’s agriculture and its realization pathways. Rural Econ. 2025, 03, 87–94. [Google Scholar]
  8. Wang, M.; Bodirsky, B.L.; Rijneveld, R.; Beier, F.; Bak, M.P.; Batool, M.; Droppers, B.; Popp, A.; van Vliet, M.T.H.; Strokal, M. A triple increase in global river basins with water scarcity due to future pollution. Nat. Commun. 2024, 15, 880. [Google Scholar] [CrossRef]
  9. Brown, C.M.; Lund, J.R.; Cai, X.; Reed, P.M.; Zagona, E.A.; Ostfeld, A.; Hall, J.; Characklis, G.W.; Yu, W.; Brekke, L. The future of water resources systems analysis: Toward a scientific framework for sustainable water management. Water Resour. Res. 2015, 51, 6110–6124. [Google Scholar] [CrossRef]
  10. Bajaj, A.; Singh, S.; Nayak, D. Impact of water markets on equity and efficiency in irrigation water use: A systematic review and meta-analysis. Agric. Water Manag. 2022, 259, 107182. [Google Scholar] [CrossRef]
  11. Rosegrant, M.W.; Binswanger, H.P. Markets in tradable water rights: Potential for efficiency gains in developing country water resource allocation. World Dev. 1994, 22, 1613–1625. [Google Scholar] [CrossRef]
  12. Zheng, H.; Liu, Y.; Zhao, J. Understanding water rights and water trading systems in China: A systematic framework. Water Secur. 2021, 13, 100094. [Google Scholar] [CrossRef]
  13. Du, Y.; Fang, K.; Zhao, D.; Liu, Q.; Xu, Z.; Peng, J. How far are we from possible ideal virtual water transfer? Evidence from assessing vulnerability of global virtual water trade. Sci. Total Environ. 2022, 828, 154493. [Google Scholar] [CrossRef]
  14. Gohar, A.A.; Ward, F.A. Gains from expanded irrigation water trading in Egypt: An integrated basin approach. Ecol. Econ. 2010, 69, 2535–2548. [Google Scholar] [CrossRef]
  15. Green, G.; Sunding, D.; Zilberman, D.; Parker, D. Explaining irrigation technology choices: A microparameter approach. Am. J. Agric. Econ. 1996, 78, 1064–1072. [Google Scholar] [CrossRef]
  16. Yang, G.; Liu, Z.; Liang, W.; Tian, G. Water resource compensation mechanism for water transfer projects from the perspective of synergistic development. Resour. Sci. 1996, 78, 469–484. [Google Scholar] [CrossRef]
  17. Razzaq, A.; Xiao, M.; Zhou, Y.; Liu, H.; Abbas, A.; Liang, W.; Naseer, M.A.U.R. Impact of participation in groundwater market on farmland, income, and water access: Evidence from Pakistan. Water 2022, 14, 1832. [Google Scholar] [CrossRef]
  18. Shen, H.M. Water rights trading and government innovation: A case study of the dongyang-yiwu water agreement. Manag. World 2005, 6, 45–56. [Google Scholar] [CrossRef]
  19. Liu, J.; Shi, K.; Wang, Z.; Jin, C. Policy effects of water rights trading (WRT). Environ. Sustain. Indic. 2024, 24, 100537. [Google Scholar] [CrossRef]
  20. Ma, J.J.; Cui, Y.; Dong, C. Creditavailability, water rights confirmation and water-saving technology investment: Evidence based on quasi-natural experiment in water rights confirmation pilots. China Rural Econ. 2022, 08, 70–92. [Google Scholar]
  21. Fu, Q.; Zhao, K.; Liu, D.; Jiang, Q.; Li, T.; Zhu, C. The application of a water rights trading model based on two-stage interval-parameter stochastic programming. Water Resour. Manag. 2016, 30, 2227–2243. [Google Scholar] [CrossRef]
  22. Yan, R.; Zhao, N.; Wang, Y.; Liu, X. The impact of water rights trading on water resource use efficiency: Evidence from China’s water rights trading pilots. Water Resour. Econ. 2024, 46, 100241. [Google Scholar] [CrossRef]
  23. Razzaq, A.; Qing, P.; Abid, M.; Anwar, M.; Javed, I. Can the informal groundwater markets improve water use efficiency and equity? Evidence from a semi-arid region of Pakistan. Sci. Total Environ. 2019, 666, 849–857. [Google Scholar] [CrossRef] [PubMed]
  24. Deng, Y.; Zhang, L. Water rights trading and agricultural water use efficiency: Evidence from China. Water 2025, 17, 2047. [Google Scholar] [CrossRef]
  25. Hao, L.; Zhang, S.; Zhang, F.; Ren, Y.; Zhang, X.; Yan, J. Balancing water saving, market attractiveness, and pollution control in crop spatial planting structure planning of arid regions. Agric. Water Manag. 2025, 309, 109348. [Google Scholar] [CrossRef]
  26. Akbari, F.; Shourian, M.; Moridi, A. Assessment of the climate change impacts on the watershed-scale optimal crop pattern using a surface-groundwater interaction hydro-agronomic model. Agric. Water Manag. 2022, 265, 107508. [Google Scholar] [CrossRef]
  27. Osama, S.; Elkholy, M.; Kansoh, R.M. Optimization of the cropping pattern in Egypt. Alex. Eng. J. 2017, 56, 557–566. [Google Scholar] [CrossRef]
  28. Dong, X.; Zhong, F.; Ji, Y. The impact of water prices on farmers’ irrgation water input: Based on the perspective of factor substution. Chin. J. Agric. Resour. Reg. Plan. 2024, 45, 76–86. [Google Scholar]
  29. Du, M.; Huang, C.; Chen, Z. Evaluating the water-saving and wastewater-reducing effects of water rights trading pilots: Evidence from a quasi-natural experiment. J. Environ. Manag. 2022, 319, 115706. [Google Scholar] [CrossRef]
  30. Liu, J.; Umer, M.; Zhao, J.; Li, Q.; Qu, M. Does water rights trading cause economic losses from agricultural water scarcity? Evidence from 264 chinese cities. Agric. Water Manag. 2025, 307, 109233. [Google Scholar] [CrossRef]
  31. Wang, S.; Fu, G.; Ma, X.; Xu, L.; Yang, F. Exploring the optimal crop planting structure to balance water saving, food security and incomes under the spatiotemporal heterogeneity of the agricultural climate. J. Environ. Manag. 2021, 295, 113130. [Google Scholar] [CrossRef]
  32. Bhusal, N.; Adhikari, A.; Lee, M.; Han, A.; Han, A.R.; Kim, H.S. Evaluation of growth responses of six gymnosperm species under long-term excessive irrigation and traits determining species resistance to waterlogging. Agric. For. Meteorol. 2022, 323, 109071. [Google Scholar] [CrossRef]
  33. Gupta, S.; Chowdhury, S.; Govindaraj, R.; Amesho, K.T.T.; Shangdiar, S.; Kadhila, T.; Iikela, S. Smart agriculture using IoT for automated irrigation, water and energy efficiency. Smart Agric. Technol. 2025, 12, 101081. [Google Scholar] [CrossRef]
  34. Du, E.; Cai, X.; Brozović, N.; Minsker, B. Evaluating the impacts of farmers’ behaviors on a hypothetical agricultural water market based on double auction. Water Resour. Res. 2017, 53, 4053–4072. [Google Scholar] [CrossRef]
  35. Zhang, C.; Zhao, Y.; Shi, C.; Chiu, Y.-h. Can china achieve its water use peaking in 2030? A scenario analysis based on LMDI and monte carlo method. J. Clean. Prod. 2021, 278, 123214. [Google Scholar] [CrossRef]
  36. Hu, S.; Ding, Y.; Cui, S.; Li, Y.; Zhao, J. Assessing economic and hydrological effects of water-saving irrigation using a coupled SWAT–MODFLOW–AquaCrop model. Agric. Water Manag. 2025, 314, 109516. [Google Scholar] [CrossRef]
  37. Zanjanian, H.; Abdolabadi, H.; Niksokhan, M.H.; Sarang, A. Influential third party on water right conflict: A game theory approach to achieve the desired equilibrium (case study: Ilam dam, iran). J. Environ. Manag. 2018, 214, 283–294. [Google Scholar] [CrossRef]
  38. Qin, C.; Jiang, S.; Zhao, Y.; Zhu, Y.; Wang, Q.; Wang, L.; Qu, J.; Wang, M. Research on water rights trading and pricing model between agriculture and energy development in ningxia, china. Sustainability 2022, 14, 15748. [Google Scholar] [CrossRef]
  39. Mehmood, F.; Wang, G.; Abubakar, S.A.; Zain, M.; Rahman, S.U.; Gao, Y.; Duan, A. Optimizing irrigation management sustained grain yield, crop water productivity, and mitigated greenhouse gas emissions from the winter wheat field in north China plain. Agric. Water Manag. 2023, 290, 108599. [Google Scholar] [CrossRef]
  40. Chen, S.; Xiao, Y.; Zhang, Z. Resource regulation and green innovation: Evidence from China’s water rights trading policy. Environ. Res. 2024, 258, 119443. [Google Scholar] [CrossRef]
  41. Quentin Grafton, R.; Horne, J.; Wheeler, S.A. On the marketisation of water: Evidence from the murray-darling basin, australia. Water Resour. Manag. 2016, 30, 913–926. [Google Scholar] [CrossRef]
  42. Antonelli, C.; Coromaldi, M.; Pallante, G. Crop and income diversification for rural adaptation: Insights from ugandan panel data. Ecol. Econ. 2022, 195, 107390. [Google Scholar] [CrossRef]
  43. Berrian, D.; Chhapia, G.; Linder, J. Performance of land productivity with single-axis trackers and shade-intolerant crops in agrivoltaic systems. Appl. Energy 2025, 384, 125471. [Google Scholar] [CrossRef]
  44. Zhang, H.; Li, M. Towards an intelligent and automatic irrigation system based on internet of things with authentication feature in VANET. J. Inf. Secur. Appl. 2025, 88, 103927. [Google Scholar] [CrossRef]
  45. Goetz, R.U.; Martínez, Y.; Xabadia, À. Efficiency and acceptance of new water allocation rules-the case of an agricultural water users association. Sci. Total Environ. 2017, 601, 614–625. [Google Scholar] [CrossRef]
  46. Mu, L.; Liu, Y.; Chen, S. Alleviating water scarcity and poverty through water rights trading pilot policy: A quasi-natural experiment based approach. Sci. Total Environ. 2022, 823, 153318. [Google Scholar] [CrossRef]
  47. Garbe, J.; Beevers, L. Modelling the impacts of a water trading scheme on freshwater habitats. Ecol. Eng. 2017, 105, 284–295. [Google Scholar] [CrossRef]
  48. Nettle, R.; Ingram, J. Smart farming technologies and changes to farm work: New insights into on-farm experiences. Technol. Forecast. Soc. Change 2025, 218, 124227. [Google Scholar] [CrossRef]
  49. Ismail, I.J. The influence of green absorptive capacity on agripreneurs’ adoption of precision agriculture technologies: Mediating effects of green innovations. Innov. Green. Dev. 2025, 4, 100248. [Google Scholar] [CrossRef]
  50. Salehi, M. Global water shortage and potable water safety; today’s concern and tomorrow’s crisis. Environ. Int. 2022, 158, 106936. [Google Scholar] [CrossRef]
  51. Yang, J.; Zuo, Z.; Li, Y.; Guo, H. Manufacturing enterprises move towards sustainable development: ESG performance, market-based environmental regulation, and green technological innovation. J. Environ. Manag. 2024, 372, 123244. [Google Scholar] [CrossRef]
  52. Hasan, M.M.; Du, F. Nexus between green financial development, green technological innovation and environmental regulation in China. Renew. Energy 2023, 204, 218–228. [Google Scholar] [CrossRef]
  53. Wang, Z.; He, B.; Fujita, T. Does environmental regulation affect inward foreign direct investment? Evidence from China. Appl. Econ. Lett. 2024, 5, 1–7. [Google Scholar] [CrossRef]
  54. Yan, T.; Wang, J.; Huang, J. Urbanization, agricultural water use, and regional and national crop production in China. Ecol. Modell. 2015, 318, 226–235. [Google Scholar] [CrossRef]
  55. Chang, Y.J.; Zhu, D. Water utilization and treatment efficiency of China’s provinces and decoupling analysis based on policy implementation. Resour. Conserv. Recycl. 2021, 168, 105270. [Google Scholar] [CrossRef]
Figure 1. Framework diagram.
Figure 1. Framework diagram.
Water 17 02414 g001
Figure 2. Parallel trend test.
Figure 2. Parallel trend test.
Water 17 02414 g002
Figure 3. Placebo Test.
Figure 3. Placebo Test.
Water 17 02414 g003
Figure 4. Heterogeneity analysis.
Figure 4. Heterogeneity analysis.
Water 17 02414 g004
Table 1. Variable definitions and descriptions.
Table 1. Variable definitions and descriptions.
VariablesVariable NameVariable DescriptionsData Resource
Dependent VariableAgricultural Water Use EfficiencyAgricultural Water Use Efficiency: Agricultural water consumption (100 million cubic meters)/Total agricultural output value (100 million RMB)China Rural Statistical Yearbook
Independent VariableWater Rights TradingWater Rights Trading Pilot: Assigned a value of 1 if the province is a pilot area for water rights trading each year; otherwise, 0Ministry of Water Resources of the People’s Republic of China
Mediating VariablesCropping StructureShare of Non-Grain Crop Area: Proportion of sown area devoted to non-grain cropsChina Rural Statistical Yearbook
Water-Saving Irrigation AreaWater-Saving Irrigation Rate: (Area under water-saving irrigation/Irrigated arable land area) × 100%
Agricultural MechanizationMechanization Level (Log): Logarithm of total agricultural machinery power per capita (kW/person)
Environmental RegulationIndustrial Pollution Control Investment Intensity: Investment in industrial pollution control/Industrial value addedChina Environmental Statistics Yearbook
Control
Variables
Level of Economic DevelopmentPer Capita GDP (Log): Logarithm of
regional GDP per capita
Industrial StructureTertiary-to-Secondary Industry Ratio: Value added of the tertiary sector/Value added
of the secondary sector
Degree of OpennessTrade Openness: Total import and
export volume/GDP
Government Fiscal ExpenditureFiscal Expenditure Ratio: Government
expenditure/GDP
R&D IntensityScience and Technology Spending Share:
Science and technology expenditure/
Government expenditure
The Provincial Statistical yearbooks
Urbanization LevelUrbanization Rate: Urban population/
Total population
Transportation InfrastructureHighway Mileage (Log): Logarithm of
total highway mileage
Table 2. Statistical description.
Table 2. Statistical description.
Variable NameVariable SymbolMeanSDMinMax
Agricultural Water Use Efficiencyawe0.0860.0750.0060.477
Water Rights Tradingart0.1690.37601
Cropping Structurecropstru0.3540.1410.0340.629
Water-Saving Irrigation Areairriagte0.4370.25701
Agricultural MechanizationagriMech1.6271.1010.176.187
Environmental Regulationer0.0090.00300.016
Level of Economic Developmentpgdp10.8960.4529.88912.065
Industrial Structurestru1.2650.680.5724.525
Degree of Opennessopen0.2590.280.0111.366
Government Fiscal Expendituregov0.2770.1890.1181.216
R&D Intensityrd0.0170.0120.0020.063
Urbanization Levelurban0.5840.1220.2620.893
Transportation Infrastructureroad11.6910.8399.46612.728
Table 3. Baseline regression results.
Table 3. Baseline regression results.
(1)(2)(3)(4)
aweaweaweawe
wrt−0.072 ***−0.051 **−0.038 **−0.045 ***
(0.017)(0.019)(0.015)(0.015)
pgdp −0.0070.046
(0.023)(0.032)
stru 0.035 **0.053 ***
(0.014)(0.017)
open 0.0480.069 *
(0.039)(0.039)
gov −0.194 ***−0.124
(0.066)(0.079)
rd 1.6443.912 **
(1.696)(1.826)
urban −0.138−0.054
(0.093)(0.067)
road −0.183 ***−0.168 ***
(0.062)(0.056)
_cons0.098 ***0.119 ***2.365 ***1.520 **
(0.003)(0.007)(0.635)(0.601)
Year FEYesYesYesYes
Province FENoYesNoYes
N372372372372
Note: Standard errors in parentheses, * p < 0.10, ** p < 0.05, *** p < 0.01.
Table 4. Balance test results of propensity score matching (PSM).
Table 4. Balance test results of propensity score matching (PSM).
UnmatchedMean %Reductt-Test
VariableMatchedTreatedControl%Bias|Bias|tp > |t|
pgdpU10.82410.917−21.8 −1.680.094
M10.82410.854−7.167.3−0.50.616
struU1.0361.332−52.6 −3.560.000
M1.0361.0153.892.80.520.603
openU0.2000.276−27.6 −2.20.028
M0.2000.206−2.291.9−0.180.858
govU0.2540.283−17.9 −1.240.217
M0.2540.2455.768.20.610.544
rdU0.0150.018−27.8 −1.950.052
M0.0150.0150.3990.030.979
urbanU0.5750.587−10.6 −0.790.431
M0.5750.584−7.727.5−0.60.551
Table 5. Robustness test results.
Table 5. Robustness test results.
(1)(2)(3)(4)(5)
aweaweaweaweawe
wrt−0.045 ***−0.046 ***−0.042 ** −0.041 ***
(0.014)(0.015)(0.016) (0.015)
L.wrt −0.036 ***
(0.012)
wft 0.012
(0.011)
pgdp0.126 ***0.0540.0290.050 *−0.018
(0.040)(0.032)(0.036)(0.030)(0.034)
stru0.069 **0.054 ***0.043 **0.054 ***0.037 **
(0.026)(0.017)(0.019)(0.017)(0.017)
open0.254 ***0.068 *0.152 ***0.068 *0.013
(0.059)(0.036)(0.047)(0.040)(0.038)
gov−0.160−0.104−0.148 *−0.090−0.377 ***
(0.124)(0.078)(0.081)(0.078)(0.124)
rd1.9304.131 **4.226 *3.356 *3.907 **
(1.917)(1.795)(2.084)(1.773)(1.688)
urban−1.177 ***−0.071−0.110−0.042−0.023
(0.271)(0.067)(0.152)(0.066)(0.082)
road−0.210 ***−0.157 **−0.176 ***−0.161 **−0.150 **
(0.073)(0.059)(0.061)(0.060)(0.057)
_cons1.781 **1.302 *1.869 ***1.367 **2.052 ***
(0.761)(0.640)(0.658)(0.663)(0.692)
Year FEYesYesYesYesYes
Province FEYesYesYesYesYes
N372372324341279
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 6. Heterogeneity by water resource allocation.
Table 6. Heterogeneity by water resource allocation.
Water-Abundant RegionsWater-Scarce Regions
aweawe
wrt−0.044 *−0.068 ***
(0.022)(0.020)
pgdp−0.0150.108 **
(0.032)(0.044)
stru0.045 ***0.050 **
(0.015)(0.022)
open0.106 *0.077 *
(0.053)(0.042)
gov−0.220 **0.169
(0.084)(0.186)
rd7.639 ***0.661
(1.928)(1.361)
urban0.020−0.075
(0.109)(0.068)
road−0.089−0.137 *
(0.055)(0.073)
_cons1.272 *0.421
(0.615)(0.743)
Province/YearYesYes
N180192
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 7. Heterogeneity by agricultural dependence.
Table 7. Heterogeneity by agricultural dependence.
High Agricultural DependenceLow Agricultural Dependence
aweawe
wrt−0.053 ***−0.035
(0.015)(0.021)
pgdp0.0330.051
(0.035)(0.030)
stru0.068 ***0.010
(0.021)(0.010)
open0.107 **−0.048 **
(0.045)(0.018)
gov−0.157 *0.030
(0.084)(0.099)
rd1.7814.829 **
(2.633)(1.871)
urban−0.104−0.042
(0.120)(0.053)
road−0.234 ***−0.047
(0.066)(0.035)
_cons2.490 ***0.048
(0.794)(0.483)
Year FEYesYes
Province FEYesYes
N180192
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 8. Heterogeneity by grain-producing regions.
Table 8. Heterogeneity by grain-producing regions.
Major Grain-Producing RegionsNon-Major Grain-Producing Regions
aweawe
wrt−0.037−0.062 ***
(0.021)(0.011)
pgdp−0.0370.122 ***
(0.054)(0.033)
stru−0.0060.082 ***
(0.015)(0.013)
open−0.0140.085 **
(0.063)(0.030)
gov−0.216−0.085
(0.152)(0.060)
rd0.7734.378 **
(1.980)(1.875)
urban0.353−0.026
(0.520)(0.038)
road−0.050−0.212 ***
(0.052)(0.059)
_cons0.9341.093 *
(0.799)(0.565)
Year FEYesYes
Province FEYesYes
N144228
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 9. Heterogeneity by level of economic development.
Table 9. Heterogeneity by level of economic development.
High LevelLow Level
aweawe
wrt−0.047 ***−0.061 ***
(0.010)(0.019)
pgdp0.103 ***0.080
(0.025)(0.050)
stru0.069 ***0.059 ***
(0.016)(0.018)
open0.057 **0.247 ***
(0.025)(0.063)
gov0.164−0.054
(0.096)(0.073)
rd−0.7374.653 **
(0.918)(2.152)
urban−0.0420.336
(0.024)(0.285)
road−0.037−0.166 **
(0.044)(0.059)
_cons−0.6941.031
(0.451)(0.722)
Year FEYesYes
Province FEYesYes
N180192
Note: Standard errors in parentheses, ** p < 0.05, *** p < 0.01.
Table 10. Cropping structure and water-saving irrigation area.
Table 10. Cropping structure and water-saving irrigation area.
(1)(2)(3)(4)
cropstruaweirriagteawe
wrt0.040 ***−0.041 **0.116 **−0.042 **
(0.014)(0.015)(0.046)(0.016)
cropstru −0.100 **
(0.047)
irriagte −0.025 *
(0.014)
pgdp0.091 *0.055−0.0870.043
(0.050)(0.032)(0.107)(0.032)
stru0.0230.055 ***−0.0020.053 ***
(0.017)(0.017)(0.058)(0.017)
open−0.166 ***0.0530.2060.074 *
(0.043)(0.036)(0.133)(0.039)
gov−0.017−0.1260.083−0.122
(0.104)(0.078)(0.356)(0.075)
rd−1.1473.797 **1.8833.958 **
(1.685)(1.753)(4.742)(1.795)
urban0.219−0.032−0.186−0.059
(0.129)(0.060)(0.593)(0.064)
road0.076−0.161 ***0.127−0.165 ***
(0.045)(0.056)(0.103)(0.057)
_cons−1.548 *1.365 **−0.1261.516 **
(0.848)(0.614)(1.557)(0.600)
Year FEYesYesYesYes
Province FEYesYesYesYes
N372372372372
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 11. Agricultural mechanization and environmental regulation.
Table 11. Agricultural mechanization and environmental regulation.
(1)(2)(3)(4)
agrimechaweerawe
wrt0.844 **−0.039 **0.003 ***−0.039 **
(0.323)(0.016)(0.001)(0.016)
agrimech −0.006 *
(0.004)
er −2.192 **
(0.895)
pgdp0.2420.047−0.0020.041
(0.875)(0.032)(0.002)(0.031)
stru0.0770.053 ***−0.0010.050 ***
(0.281)(0.017)(0.001)(0.017)
open0.5830.073 *−0.0010.068 *
(0.798)(0.036)(0.002)(0.038)
gov7.258 *−0.0780.002−0.120
(4.144)(0.087)(0.006)(0.076)
rd−21.7863.772 **−0.189 *3.498 *
(44.012)(1.729)(0.099)(1.783)
urban0.707−0.050−0.003−0.060
(1.727)(0.064)(0.005)(0.065)
road0.118−0.167 ***0.003−0.162 ***
(1.010)(0.057)(0.002)(0.057)
_cons−5.0941.487 **0.0031.525 **
(14.130)(0.646)(0.029)(0.606)
Year FEYesYesYesYes
Province FEYesYesYesYes
N372372372372
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Table 12. Spatial autocorrelation test.
Table 12. Spatial autocorrelation test.
YearIzp-Value
20110.3063.3490.001
20120.3143.4610.001
20130.2843.1940.001
20140.2673.1220.002
20150.3003.3810.001
20160.2293.2310.001
20170.2713.0000.003
20180.2702.9190.004
20190.2422.5860.010
20200.2502.6850.007
20210.2322.4700.013
20220.2572.7480.006
Table 13. Results of specification tests for spatial econometric model selection.
Table 13. Results of specification tests for spatial econometric model selection.
TestCategoryTest Statisticp-Value
LM TestSpatial Error Model (SEM)Lagrange Multiplier200.7000.000
Robust Lagrange Multiplier57.0630.000
Spatial Autoregressive Model (SAR)Lagrange Multiplier144.0210.000
Robust Lagrange Multiplier0.3840.536
LR TestSpatial Autoregressive Model (SAR) vs. Spatial Durbin Model (SDM)29.370.000
Spatial Error Model (SEM) vs. Spatial Durbin Model (SDM)45.940.000
Wald TestSARSDM can be simplified to SAR29.850.000
SEMSDM can be simplified to SEM42.470.000
HausmanRandom Effects168.000.000
Table 14. Regression results of spatial econometric models.
Table 14. Regression results of spatial econometric models.
(1)(2)(3)
SARSDMWxSEM
Main
wrt−0.043 ***−0.041 ***−0.024 **−0.041 ***
(−7.607)(−7.160)(−2.111)(−6.966)
pgdp0.048 ***0.034 **−0.061 **0.055 ***
(3.279)(2.016)(−2.052)(3.437)
stru0.050 ***0.040 ***−0.041 ***0.051 ***
(7.738)(5.935)(−2.687)(7.995)
open0.087 ***0.099 ***0.0150.082 ***
(5.255)(5.531)(0.470)(4.803)
gov−0.114 **−0.160 ***−0.186 *−0.108 **
(−2.518)(−3.530)(−1.945)(−2.357)
rd3.004 ***2.312 ***3.654 **3.281 ***
(4.275)(3.104)(2.521)(4.320)
urban−0.136 ***−0.117 **−0.124−0.140 ***
(−2.663)(−2.292)(−1.124)(−2.719)
road−0.149 ***−0.154 ***−0.073−0.150 ***
(−8.246)(−8.404)(−1.536)(−8.057)
ρ 0.336 ***0.210 ***
(6.146)(3.065)
λ 0.308 ***
(4.419)
N372372372372
Note: Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Liu, H.; He, B.; Chen, W. Does Water Rights Trading Improve Agricultural Water Use Efficiency? Evidence from a Quasi-Natural Experiment. Water 2025, 17, 2414. https://doi.org/10.3390/w17162414

AMA Style

Liu H, He B, Chen W. Does Water Rights Trading Improve Agricultural Water Use Efficiency? Evidence from a Quasi-Natural Experiment. Water. 2025; 17(16):2414. https://doi.org/10.3390/w17162414

Chicago/Turabian Style

Liu, Hengyi, Bing He, and Wei Chen. 2025. "Does Water Rights Trading Improve Agricultural Water Use Efficiency? Evidence from a Quasi-Natural Experiment" Water 17, no. 16: 2414. https://doi.org/10.3390/w17162414

APA Style

Liu, H., He, B., & Chen, W. (2025). Does Water Rights Trading Improve Agricultural Water Use Efficiency? Evidence from a Quasi-Natural Experiment. Water, 17(16), 2414. https://doi.org/10.3390/w17162414

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop