Irrigation water pricing based on the ELES-WPTI model in the Yinda-Jihuang project of China

Irrigation is particularly important for agricultural production in arid and semi-arid regions (Flores Arévalo et al., 2021). Agricultural water consumption accounts for about three-quarters of the total water consumption worldwide and even over 90% of the total water consumption in many developing countries (Harun et al., 2015). Water shortages have worsened the food security situation of the world, especially in northwest China where rainfall is extremely limited. Irrigation often accounts for over 60% of the total water consumption (Wu et al., 2015; Cheng et al., 2019), and many regions experience a shortage of water for irrigation because of rapid industrialization and urbanization, as well as environmental challenges such as climate change and water pollution (Wang et al., 2018). China has built many water conservation and transfer projects to supply water to where it is needed (Chang et al., 2019). These projects play important roles in agricultural water supply and food security in arid and semi-arid regions (Ward et al., 2013). Therefore, it is important to improve the water utilization efficiency of these water conservation and transfer projects.

Water pricing is an important means of improving water allocation and reducing water consumption (Castellano et al., 2008). For farmers, an appropriate water price encourages them to adopt more efficient irrigation techniques, which means that higher income is achievable at a lower cost (Schoengold et al., 2006; Chaudhuri & Roy, 2019). Some attempts have been made to explore the relationship between irrigation water use and water price (Berbel et al., 2018; Saccon, 2018). It is found that the elasticity of demand for irrigation water is very low, which implies farmers are willing to reduce irrigation water only when the water price rises to a level that significantly reduces the income from growing crops. For this reason, water pricing can have a great impact on farmers' water-saving behaviors (Mohseni et al., 2022). It should be noted that an appropriate water price also contributes to increase the maintenance budget for water supply projects and ensure their safe operation. Recently, many methods have been proposed for water pricing, such as the residual method, marginal return or marginal cost pricing, water price tolerance index (WPTI), and contingent valuation method (CVM) (Mu et al., 2019; Zamani et al., 2021; Chebil et al., 2022). WPTI is a simple and effective method to study the tolerance of farmers to water prices. However, it should be noted that water pricing is not just an economic issue, fairness and justice should be equally important (Esmaeili & Vazirzadeh, 2009; Fosli et al., 2021). Especially for irrigation water, which is the means of agricultural production and an important factor to ensure food security, so the charge for water is not a priority indicator. New methods are needed to improve the limitations of previous methods, such as using an extended linear expenditure system (ELES) to determine the affordability of different income groups of farmers, especially low-income farmers (Wang et al., 2019). Here, affordability means farmers' psychological and economic acceptability of water pricing. Due to the large difference in the affordability of farmers in different regions to agricultural water prices, it is necessary to analyze the affordability of farmers with different payment capacities by using the WPTI method.

In this work, an ELES-WPTI model is proposed to evaluate the appropriate agricultural water price in order to maximize the total benefit of the Yinda-Jihuang (YJ) Project and local farmers. The consumption patterns of farmers are analyzed from the perspective of demand analysis that involves the total marginal propensity to consume and the income elasticity of demand. Farmers are classified into different levels of ability to pay according to per capita net income, expenditure on food, tobacco and alcohol and total living expenditure. On the basis of current water expenditure, the water price tolerance is analyzed, and a reasonable water price is determined based on dynamic changes in water price for different groups of farmers. Reasonable water prices should balance the interests of the government and water users, while the costs of the project should be considered.

The study area is located on the Qinghai-Tibet Plateau in Qinghai Province, China, which is known as an alpine region with an arid climate and a harsh natural environment. The mainstream Huangshui River is located in the most economically developed area of the eastern Qinghai Province, which is the economic, political, communication and cultural center of the province. For this reason, a large inter-basin water transfer project, the YJ Project, is built to divert water from the Datong River to the Huangshui River basin in order to ensure a more rational allocation of water resources for sustainable economic and social development of the Huangshui River basin. This project consists of five parts, including Shitouxia Water Control Project, Main Water Transfer Canal, Heiquan Reservoir, North Main Canal Project and West Main Canal Project, and the last two projects are mainly used for industrial, urban and agricultural water supply along the route.

The data on per capita disposable income and expenditure of rural residents in DH and HL districts are obtained from Xining Statistical Yearbook 2020, Haidong Statistical Yearbook 2020, and Qinghai Province Statistical Yearbook 2020. According to China Statistical Yearbook, the living expenditure of rural residents is divided into eight categories, including food, tobacco and alcohol, clothes, residence, daily necessities and services, transportation and communication, education, culture and recreation services, medicine and medical services, miscellaneous and service, as shown in Table 1.

In order to better understand the actual situations of the water-receiving areas, a questionnaire was distributed to 150 randomly selected farmers in DH and HL districts, YJ Engineering Bureau officials and stakeholders. The questionnaire mainly consisted of basic information about farmers, current agricultural water consumption and expenditure, planting structure and production cost and benefit. Eighteen questionnaires were deleted due to incomplete or illogical responses (i.e. inconsistent answers), yielding a total of 132 valid questionnaires.

Step1: ELES model is introduced to analyze the basic consumption demand and expenditure structure in order to define the different levels of ability to pay.

Step2: WPTI method is used to evaluate the tolerance of farmers to current water expenditure by the ratio of water price to agricultural production cost and net income.

Step3: ELES-WPTI is used to analyze the dynamic impact of changes in water price for different levels of ability to pay.

The demand for a given commodity mainly depends on the income level of the consumer and the price of the commodity. The demand can be divided into basic needs and non-basic needs, the former of which is independent of income. When the basic need is satisfied, the remaining income can be allocated to non-basic needs based on the marginal propensity to consume.

Then, the living consumption structures of consumers are analyzed based on the marginal propensity to consume, expenditure on basic needs and income elasticity of demand. According to the ratio of per capita net income to food expenditure and total per capita living expenditure, consumers are defined as having ‘no ability to pay’, ‘limited ability to pay’ and ‘high ability to pay’.

Reasonable agricultural water expenditure accounts for 20–30% of the agricultural production cost and 10–20% of the net income in China (Wang et al., 2019). According to the local economic development and agricultural cultivation structure, if it is lower than the above range, the project operation cost will be difficult to maintain; otherwise, it will be beyond the farmers' economic affordability. The agricultural production cost and income of farmers in DH and HL districts are substituted into Equation (6). The water expenditures for the three levels of ability to pay are obtained under the current water price. Then, the proportion of water expenditures under different water prices can be further simulated. Based on the current water expenditure and dynamic adjustment of water price, the appropriate water price is set for different levels of ability to pay.

α_i and β_i are calculated by using the ordinary least squares (OLS) method (Table 2). The results show that the determination coefficient R² of the regression model is nearly 1, indicating that the ELES model fits the data very well. Thus, there is a relationship between the living expenditure and the annual per capita net income of the family. Table 2 shows that the total expenditure on basic needs is 1,328$ in the DH district and 1,119$ in the HL district, respectively. The expenditures on food, tobacco and alcohol are 498$ and 437$ and the proportion to the total expenditure on basic needs (Engel coefficient) is 37 and 38%, respectively, indicating that farmers in these two regions have high living s. The marginal propensity to consume is calculated by Equation (3). As shown in Table 2, the top three items with the highest marginal propensity to consume are food, tobacco and alcohol, residence, transportation and communication. The total marginal propensity to consume is 0.54 in the DH district and 0.51 in the HL district, respectively, which means that 54 and 51% of farmers' new income is used for living expenditure. The income elasticity of demand ranges from 0.5 to 0.8, indicating no elasticity. Thus, it can be concluded that the expenditure is mainly used for daily necessities. Obviously, the demand for daily necessities will increase with increasing income, but the increasing rate is lower than that of income.

According to the survey, the current agricultural water price is 46.7 $/hm² in the DH district and 36.0 $/hm² in the HL district, respectively. The production cost and net income for farmers with no, limited and high ability to pay are summarized in Table 4. The ratios of agricultural water expenditure to production cost and net income for farmers with different levels of ability to pay under the current water price are calculated by Equation (6). Water expenditure is necessary for agricultural production, and the ratio of water expenditure to production cost is a good indicator of farmers' tolerance to water price from the perspective of cost. The net income is defined as the difference between gross output and total cost of agricultural production per unit area, which can be an intuitive indicator of the economic benefit of agricultural production. As shown in Table 4, water expenditure accounts for 2.1–3.1% of the total production cost and 1.8–2.2% of the net profit, which is far below the acceptable range (20–30% of the total production and 10–20% of the net income, respectively). In this regard, it is concluded that the current agricultural water price is affordable for farmers.

Two important factors should be considered in agricultural water pricing. One is farmers' affordability and the other one is farmers' tolerance level. The YJ Project is built to ensure the sustainable development of the local economy by increasing the crop yield and the per capita net income of farmers. Thus, the water price can be determined separately for DH and HL districts based on the principle that water expenditure accounts for 15% of the total production cost and 10% of the net income, and then the weighted average is obtained according to the principle of one water price for one district. As shown in Table 5, the affordable water price is determined to be 237 $/hm² in the DH district and 205 $/hm² in the HL district, respectively.

The current water price in DH and HL districts still follows the water price of 2006. By comparison, the water pricing based on the ELES-WPTI model has increased by 4.1 and 4.7 times, respectively, which is consistent with the economic development. The GDP and per capita disposable income of Qinghai Province increased by 3.7 and 4.2 times, respectively, compared to 2006. Thus, the water price is reasonable and affordable considering farmers' income and expenditure.

In this study, the ELES-WPTI model that combines the ELES model and the WPTI method is developed for irrigation water pricing of the YJ Project. Farmers' ability to pay for agricultural water is evaluated from living expenditure and family income. The ELES model is introduced to define farmers' ability to pay and the expenditure on basic living needs is determined. It is found that farmers' ability to pay is slightly higher in the DH district (<498$ for no ability; 498–1,329$ for limited ability; and >1,329$ for high ability) than that in HL District (437$ for no ability; 437–1,119$ for limited ability; and >1,119$ for high ability). The impact of changes in water prices on farmers with different levels of ability to pay is further analyzed using the WPTI method. As water expenditure accounts for <15% of the total production cost and <10% of the net income, the affordable water price is determined to be 237 $/hm² in the DH district and 205 $/hm² in the HL district, respectively.

This study provides a new means of water pricing based on farmers' affordability. However, some limitations of this study should also be noted. For example, the impact of agricultural policies was not considered in this study. In addition, the operation cost of the project is not considered in water pricing due to the public welfare nature of the project. However, it should be taken into consideration in other water transfer projects, in order to ensure the efficient operation of the project. Another issue worthy of discussion is the potential for participants to change their crops in an economically feasible way to lower their water demand. Perhaps this should be discussed in a further study.

This work was funded by the National Natural Science Foundation of China (Grant No. 52209042), the Key Research and Development Program of Hebei Province (Grant No. 21374201D), the National Key Research and Development Program of China (Grant No. 2021YFC3001000).

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

The authors declare there is no conflict.