China's reform and opening-up strongly impact its local water resources. The most frequently stated problems are water supply scarcity and pollution issues in China's Yangtze River. What is not yet clear is the correlation between comprehensive water efficiency and its sub-stage efficiencies, water utility efficiency, and water governance efficiency. This study aims to investigate water utility and water governance together. This paper provides a fresh perspective on developing an improved DEA method called the SHAN model that evaluates and decomposes the comprehensive water efficiency to water utility efficiency and water governance efficiency for 30 cities in the Yangtze River Basin from 2005 to 2015. The SHAN model is a modified RAM Network DEA model with the extra consideration of Super SBM and undesirable output. A Tobit model will be adopted to analyze the relationship between comprehensive water efficiency and its sub-stages. This paper uses panel quantile regression to investigate the impact of the degree of opening and other efficiency determinants. The results show that the value of comprehensive water efficiencies in most cities along the Yangtze River Basin are lower than 1, as well as the water utility efficiencies and water governance efficiencies. The results indicate that the government should significantly improve water resource utilization and pollution problems. The findings also show that the degree of opening, population density, GDP per capita, urbanization, and industrial structure positively influence comprehensive water efficiency. Population urbanization lag has negative impacts on the comprehensive water efficiency. The Tobit regression results show that the water resource utilization efficiency and water resource management efficiency are 1.435 and 0.227, respectively. This result indicates that the water utility stage has a more significant effect on comprehensive water efficiency.

  • SHAN model is adopted to decompose and measure the comprehensive water efficiency.

  • Tobit model is adopted to analyze the interrelationship of comprehensive water efficiency and its sub-stages. Water utility stage significantly affects the comprehensive water efficiency.

  • The degree of opening, population density, GDP per capita, urbanization and industrial structure positively affect comprehensive water efficiency, and population urbanization lag negatively affects it.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Water plays a crucial role in sustainable human development and economic progress as an irreplaceable resource. The water issue has received considerable attention globally, and there has been growing recognition of the problems associated with water. China is one of the most water-deficient countries in the world, and the distribution of water resources is highly uneven. In China, the total water resources in the Yangtze River Basin are 1.313 × 1012 m3. However, only 79.2% meet Class I to III water quality standards. The remaining 20.8% are polluted (Yuan 2002). Since reform and opening up, China has made significant progress in economic development, but this has also been accompanied by excessive consumption and pollution of water resources (Zhou et al. 2018). The shortage and excessive pollution of water resources has become a stumbling block to sustainable economic development. Therefore, it is of great significance for sustainable development to improve the efficiency of comprehensive water resources, improve the efficiency of water resource utilization and effectively control water pollution.

In order to solve the current dilemma of water resources, improving the efficiency of the water resources system is a practical solution (Deng et al. 2016). The method of improving the whole system should not only consider the consumption of resources but also reduce pollution and protect the environment (Klemeš & Huisingh 2005). Although there is a surge of interest in research on energy efficiency and ecological efficiency, few studies have evaluated the ‘black box’ of comprehensive water efficiency and the impacts of their determinants. This study aims to use the Super SBM modified RAM Network DEA model with undesirable output (SHAN model) to reveal the inside of the black box and explore the role of globalization on water use efficiency. This paper selects the degree of opening as an essential indicator of globalization for a region. Then the impact of the degree of opening (DO) on the comprehensive water efficiency will be evaluated. The definition of DO in this paper refers to the ratio of total imports and exports to GDP.

The present literature shows that the issue of water efficiency has drawn much attention from academia. The total-factor framework was used to study water efficiency (Hu et al. 2006). The water consumption of the beverage industry was demonstrated to find a solution to excessive water use in the production process (Jagtap et al. 2022). The urban water metabolism framework was used to evaluate the performance of water resources management in three regions in South Korea (Jeong & Park 2020). Improving the efficiency of water use is an effective way to solve the shortage of water resources (Allan 1999). Some other scholars studied irrigation water use efficiency and its impact (Lilienfeld & Asmild 2007). The directional distance function method was adopted to study China's water utility efficiency, considering both desirable output and undesired output (Ma et al. 2016).

The DEA model can be proposed to study the efficiency of decision-making units (Charnes et al. 1978). These are classic DEA models such as CCR and BCC models. The traditional models cannot solve the slack variable problem. The non-radial SBM model has been proposed with the advantage of considering the slack variable (Tone 2001). The Super SBM model can further distinguish the efficiency between multiple effective DMUs based on SBM (Tone 2002). The above models are like ‘black box operations’ and cannot reveal the secrets inside the ‘black box’. The Network DEA method has been utilized to solve the ‘black box’ (Färe et al. 2007). The two-stage network model, the composite network model, and the relationship between the two models has been discussed (Wei 2013). The above models can open the ‘black box’. This paper uses the Super SBM modified RAM Network model with the undesirable output method (SHAN Model) to evaluate the comprehensive water efficiency, water utilization efficiency, water governance efficiency, and their relationship.

Water efficiency in the paper refers to producing the same amount of sound output with the consumption of less water (Sun et al. 2014). Andersen & Petersen (1993) proposed a method that can further distinguish the degree of effectiveness of effective DMU. This method was later called the super-efficiency model (Tone 2002). The essence of the super-efficiency model is to remove the evaluated DMU from the reference set. The efficiency of the evaluated DMU is obtained by referring to the frontier composed of other DMUs so that the effective DMU can be further distinguished. The super-efficiency value can be greater than 1 (Chen 2014).

Many published studies describe the method of wastewater treatment. At present, the sewage treatment methods in the world mainly include activated carbon adsorption, physical precipitation, chemical coagulation, flotation, filtration, biological water treatment, and nano-membrane. The electro-flocculation wastewater treatment method significantly affects wastewater in the mining industry (Wang & Amp 2019). Some scholars introduced specific innovative achievements of membranes in water treatment, such as spiral-wound membrane elements (Nicolaisen 2003). The applicability of cell-based bioassays for water treatment efficiency and water quality assessment have been analyzed (Escher et al. 2014). In a study investigating polyphosphate (poly-P), polyphosphate (poly-P) was the main component in the activated sludge of wastewater treatment plants. It has apparent biological phosphorus removal ability (Staal et al. 2019).

Previous researchers have extensively discussed the influencing factors of water resource efficiency. In terms of the degree of opening, increasing the opening level can improve the water utilization efficiency and reduce wastewater emission regions (Ma et al. 2016). Openness can bring in advanced technology and equipment and new plans and strategies. Moreover, due to openness, there is mutual supervision between different countries to force less-developed regions to absorb new environmental protection strategies and technologies. The technical market turnover has a significant positive effect on the technical efficiency of water resources that shows technical exchanges between different countries can help to improve water efficiency (Sun et al. 2014). The openness even includes the trade of water resources themselves between countries. The transfer of virtual water has made countries rely heavily on water resources from other countries (Clark et al. 2015). GDP per capita is positively related to water efficiency, which indicates that the increasing total economic output is one of the keys to improving water utility efficiency (Du et al. 2015). The proportion of industrial water usage to agricultural water usage can positively affect water resource utilization efficiency. Foreign direct investment and water consumption per capita negatively affect water resource utilization efficiency. Zhao et al. 2014 and Bao et al. (2016) claimed that urbanization and population density positively impact the water use efficiency of local-level cities in Henan Province and neighboring prefecture-level cities.

Although water resource efficiency has attracted extensive academic research, little research has been conducted to reveal the interrelationship between comprehensive efficiency of water resources, water utility efficiency and water governance efficiency. This paper aims to solve the ultimate balance problem between economic growth and environmental pollution, in the context of China's opening-up and water shortages.

This study takes 30 cities along the Yangtze River Basin as a case study and investigates the role of the degree of openness in water efficiency. This paper attempts to expand the existing research from the following contributions. First, this study provides new insights into decomposing the comprehensive water efficiency by adopting the SHAN model. The SHAN model is a new Super SBM modified RAM Network DEA model with undesirable output. Secondly, this paper investigates the interrelationship between comprehensive water efficiency and its sub-stages at the prefecture level along the Yangtze River Basin. Thirdly, this study analyzes the determinants that affect the efficiency of water resources, especially the degree of openness.

This study aims to find out how to improve the comprehensive water efficiency more effectively by investigating its interior and revealing its inner relationship. The results of this study show that the water utility stage is the primary internal link that affects the comprehensive water efficiency in the Yangtze River Basin. The result of this study is significant for solving water issues in the Yangtze River and improving the comprehensive utilization of water resources. Decomposing the comprehensive water efficiency and identifying key stages can ensure the economy's effective development and protect environmental water resources. A balance between the economy and the environment can be achieved. This research will help improve the utilization of water resources and reduce the adverse effects of wastewater emissions on the environment. Integrating water utilization efficiency and governance efficiency will help comprehensively study the degree of sustainable development of water resources from an overall angle. Policymakers can focus on investment in essential stages and achieve the optimal result at a lower cost.

This paper begins with an introduction. It will then go on to the method and data. The third chapter of this paper will be the results and discussions. The final section provides the conclusion with some policy recommendations.

SHAN model

This paper uses the Super SBM modified RAM Network DEA model with undesirable output (SHAN model) to study the comprehensive water efficiency, water utility efficiency, and water governance efficiency. The model used in this paper is derived as follows: to solve the mystery of the ‘black box’ and explore the specific intermediate relationship between each stage. The SBM model and network DEA are combined, and an exploration of intermediate output is added (Tone & Tsutsui 2009). This method is called network DEA with slack-based measurement. Based on the research, the Range Adjusted Measure Network DEA model is further introduced and defines divisional and comprehensive efficiencies (Maruyama 2009).

The Range Adjusted Measure Network DEA (RAM Network DEA) is defined as follows:
formula
(1)
with
formula
s.t.
formula
where
formula

Notations for describing the models:

  • n: the amount of DMUs

  • K: the amount of divisions

  • mk: the amount of inputs to division k

  • rk: the amount of outputs from division k

  • D: the set of divisions in the model. The divisions are numbered from I to K

  • S: the set of divisions which have no approaching connections

  • T: the set of divisions which have no outgoing connections

  • (k,h) : the link (intermediate product) from division k to division h

  • t(k,h) : number of items in link (k,h)

  • L: the set of links

  • Pk = {p | (p,k) L} (set of antecessors to division k)

  • Fk = {q | (k,q) L} (set of successors from division)

  • : input resources to DMUj at division k (k = 1,…,K)

  • : output products from DMUj at division k (k = 1…,K)

  • : linking input resources to DMUj at division h from division k ((k, h) L)

  • =linking output products from DMUj at division k to division h ((k, h) L)

where j denotes jth DMU (j = 1,…,n).

Assume that:

no linking inputs to starting divisions and no linking outputs from terminal divisions.

The production possibility set is defined by the following :{(, , , )}
formula
formula
formula
formula
formula
formula
formula
where is the relative weight of division k, and k is set up by the corresponding importance.

The SBM model was proposed and added the slack measurement into the DEA model (Tone 2001). The Super-Slack-based measure method (Super-SBM) was introduced that further distinguished the effectiveness when multiple DMUs were equally effective (Tone 2002).

Finally, the model in this paper considers the situation of undesired output, and wastewater emission was added to the model to form the Super SBM modified RAM Network DEA model with undesirable output (SHAN model). As an intermediate variable, the wastewater from the first stage of production is also used as an input in the second stage of water governance.

Variable selection and data source

Data for SHAN model

Figure 1 is the flowchart that illustrates how to perform the SHAN model. The SHAN model has two stages. The first stage is the water utility stage. In the water utility stage, the input variables are built-up area, labor force, capital, and water consumption. GDP is considered a desirable output. Initial industrial wastewater discharge is the undesirable output. It is also treated as the intermediate variable and used as the input in the second stage. The second stage is the water governance stage. In the second stage, the initial industrial wastewater and wastewater treatment plant ratio are the input. The output in the water governance stage is industrial wastewater emission up to standard and final wastewater emission. The indicators adopted by the SHAN model in this paper are illustrated in Table 1.
Table 1

Summary of the comprehensive water efficiency

StageIndicator typeVariableUnit
Indicators Water utility efficiency Input Built-up area square kilometres 
Labor 10,000 people 
Capital 10,000 yuan 
Water consumption 10,000 tonne 
Output GDP 10,000 yuan 
Intermediate Initial industrial wastewater discharge 10,000 tonne 
Water governance efficiency Input Wastewater treatment plant ratio 
Output Industrial wastewater emission up to standard 10,000 tonne 
Output Final wastewater emission 10,000 tonne 
StageIndicator typeVariableUnit
Indicators Water utility efficiency Input Built-up area square kilometres 
Labor 10,000 people 
Capital 10,000 yuan 
Water consumption 10,000 tonne 
Output GDP 10,000 yuan 
Intermediate Initial industrial wastewater discharge 10,000 tonne 
Water governance efficiency Input Wastewater treatment plant ratio 
Output Industrial wastewater emission up to standard 10,000 tonne 
Output Final wastewater emission 10,000 tonne 
Figure 1

Flowchart for the indicators of the SHAN Model.

Figure 1

Flowchart for the indicators of the SHAN Model.

Close modal

Data for panel quantile regression

For the factors that affect the comprehensive water efficiency, this paper adopts the following variables:

  • (1)

    GDP per capita. This is calculated by using local GDP divided by population.

  • (2)

    Degree of openness. This is used to measure the level of openness in a region. The data is obtained by using total export and import as a proportion of GDP.

  • (3)

    Urbanization. The percentage of urban population as a proportion of total population.

  • (4)

    Urbanization lag.

According to the research done by Tian & Wang (2019), the urbanization lag is defined as follows:
formula
(2)
  • (5)

    Industrial structure. This is the indicator used to measure the weight of secondary industry in total GDP.

  • (6)

    Population density. The data is collected using population divided by land area.

The summary of influencing factors is shown in Table 2:

Table 2

Influencing factors summary

Influencing factorsIndicatorsAbbreviationUnit
GDP per capita Regional GDP per capita GDPPC yuan per person 
Degree of opening Total local volumes of export and import divided by local GDP DO ratio 
Industrial structure The added value of second any industry divided by GDP IS ratio 
Urbanization The amount of urban population as a proportion of the total population Urban ratio 
Population urbanization lag The amount of population urbanization speed as a proportion of land urbanization speed divided by 1.12 Lag ratio 
Population density Population divided by area PD ratio 
Influencing factorsIndicatorsAbbreviationUnit
GDP per capita Regional GDP per capita GDPPC yuan per person 
Degree of opening Total local volumes of export and import divided by local GDP DO ratio 
Industrial structure The added value of second any industry divided by GDP IS ratio 
Urbanization The amount of urban population as a proportion of the total population Urban ratio 
Population urbanization lag The amount of population urbanization speed as a proportion of land urbanization speed divided by 1.12 Lag ratio 
Population density Population divided by area PD ratio 

Data source

Based on the availability of the data, this study uses data from 30 cities in China's Yangtze River Basin from 2005 to 2015 to measure the comprehensive water efficiency and their two sub-efficiencies. The data sources come from China City Statistical Yearbook and China Environmental Statistical Yearbook.

Model construction

Panel quantile regression

Panel quantile regression is used later in this study to investigate the influencing factors of comprehensive efficiency. The driving factors at different quantile levels can be analyzed, and the framework to evaluate the influencing factors is built (Yan et al. 2019):
formula
formula
(3)
formula

Tobit model

The inter-correlation between comprehensive water efficiency, water utility efficiency, and water governance efficiency is investigated using the Tobit model:
formula
(4)
where i represents city index; t indicates time index; ɛ is the disturbance term with mean zero and finite variance; WCE is comprehensive water efficiency; WUE is water utility efficiency; and WGE is water governance efficiency.

Robustness test

The robustness test, also known as sensitivity analysis, relaxes some paper assumptions to see whether the results are unchanged or robust. The usual practice is to examine the stability of measurement results by changing the sample interval, function form, or measurement method. This study used the ordinary least squares method (OLS) as a robustness test for the econometric model.

As shown in Figure 2, the results show that comprehensive water efficiency lies in the range of 0.622–1. This indicates that the comprehensive water efficiency for 30 cities in China's Yangtze River Basin is relatively high. In particular, E'Zhou, Wuhu, Pan Zhihua, Chi Zhou, and Jiu Jiang have had lower water utility efficiency for those years. This is because most of them are located in the middle and upper reaches of the Yangtze River, with an underdeveloped economy and relatively backward technical level. Moreover, for most cities, the water utility efficiencies are between 0.623 and 1, but the water governance efficiencies are only between 0 and 0.573. This shows that water governance efficiencies are relatively inefficient for most cities compared with water utility efficiency. The result attests that water governance is a shortcoming. This is due to the lack of effort in environmental governance and the public's weak environmental awareness. The water utility efficiencies remain steady in some cities like Wuxi, Tai Zhou, Su Zhou, Pan Zhihua, and Nanjing. The water governance efficiencies in most cities are relatively low, ranging from 0 to 0.573. Among them, Suzhou ranks top, based on water governance efficiency, with an efficiency value of 1. The main reason is that it is located in the Yangtze River Delta and has an excellent economic and technological foundation. At the same time, it is also located along the coast, which facilitates the introduction of foreign technology and capital.

Interestingly, compared with the water utility efficiency, comprehensive water efficiency has a similar value and shows a similar trend. Furthermore, the top three cities with a higher comprehensive water value are Suzhou, Chongqing, and Yibin, with an efficiency value of 1. This shows that the comprehensive efficiency of water resources in these areas is better. Figure 3 presents the heat map of comprehensive water efficiency, water utility efficiency, and water governance efficiency. Based on the Tobit regression shown in Table 3, the relationship between comprehensive water efficiency and its sub-stages is that water utility efficiency plays a more critical role in comprehensive water efficiency.
Table 3

Summary of statistics of Tobit regression

VariableCoefficientStd errort-StatisticProb
Water utility efficiency 0.993 0.019 51 0.000 
Water governance efficiency 0.003 0.008 0.42 0.677 
C 0.001 0.016 0.068 0.947 
R-squared 0.992  Durbin–Watson stat 1.15 
Adjusted R-squared 0.991  F-statistic 1,720 
VariableCoefficientStd errort-StatisticProb
Water utility efficiency 0.993 0.019 51 0.000 
Water governance efficiency 0.003 0.008 0.42 0.677 
C 0.001 0.016 0.068 0.947 
R-squared 0.992  Durbin–Watson stat 1.15 
Adjusted R-squared 0.991  F-statistic 1,720 
Figure 2

Radar chart for water utility efficiency, water governance efficiency, and comprehensive water efficiency of Yangtze River basin cities from 2005 to 2015. The abbreviations for the cities along the Yangzte River are shown in Table A1 in Appendix A.

Figure 2

Radar chart for water utility efficiency, water governance efficiency, and comprehensive water efficiency of Yangtze River basin cities from 2005 to 2015. The abbreviations for the cities along the Yangzte River are shown in Table A1 in Appendix A.

Close modal

There is a notable trend that comprehensive water efficiency and water utility efficiency have similar trends and their values are also quite similar. The value of comprehensive water efficiency ranges from 0.622 to 1, while the value of water utility efficiency fluctuates between 0.623 and 1. This can be explained by water utility efficiency determining comprehensive water efficiency. This conclusion can also be confirmed in another way. The Tobit regression coefficient of water utility (0.993) is much larger than that of water governance (0.003). This indicates that water utility has a significant impact on comprehensive water efficiency. The reason is that most cities are still focused on utility and economic growth. More funds and policies are made to this stage while governance lacks attention. Policymakers and the public are not fully aware of the importance of water governance.

Previous studies only decomposed comprehensive efficiency into scale efficiency and technical efficiency. However, by adopting the SHAN model to decompose the comprehensive efficiency into utilization efficiency and governance efficiency, this paper can analyze resource utilization and resource governance together. The water utility efficiency in this paper can be further decomposed into scale efficiency and technical efficiency. This makes the current study capable of revealing the composition of the comprehensive water efficiency and the interrelationship of its sub-stages.

Degree of opening

The impact of the degree of opening on water efficiency is positive. According to Table 4, the statistics of the 50th–75th quantile and above the 75th quantile have passed 1% significance. Furthermore, the coefficient corresponding to the degree of the opening increases with the increase of the quantile. This shows that areas with high water efficiency values have a more significant impact on water efficiency values. This is similar to the results obtained by other scholars (Sun et al. 2014). For a region with a higher degree of opening, it is more likely to bring more advanced water treatment technology and equipment. This will help to improve the water utilization efficiency and reduce wastewater emissions (Clark et al. 2015).

Table 4

Panel quantile regression results

VariablesOLSQuantiles
0.10.20.30.40.50.60.70.80.9
DO 0.15*** (0.04) 0.05 (0.05) 0.07* (0.04) 0.07* (0.04) 0.010** (0.04) 0.11*** (0.04) 0.12*** (0.04) 0.14*** (0.05) 0.23*** (0.05) 0.32*** (0.04) 
GDPPC 0.20** (0.09) 0.09 (0.25) 0.36* (0.20) 0.34** (0.16) 0.39*** (0.15) 0.25* (0.13) 0.29** (0.13) 0.27** (0.13) 0.24 (0.19) 0.19 (0.18) 
Urban 0.14 (0.09) −0.07 (0.10) 0.02 (0.09) 0.07 (0.09) 0.08 (0.09) 0.17* (0.09) 0.16 (0.1) 0.24** (0.11) 0.26* (0.14) 0.09 (0.12) 
Lag −0.07 (0.12) −0.21* (0.11) −0.16* (0.09) −0.13 (0.09) −0.16* (0.09) −0.16* (0.09) −0.06 (0.11) −0.16 (0.11) −0.08 (0.14) 0.04 (0.13) 
IS 0.3 (0.21) 0.82* (0.42) 0.38 (0.33) 0.28 (0.27) 0.17 (0.2) 0.18 (0.2) 0.19 (0.22) 0.39 (0.28) 0.48 (0.4) 0.2 (0.29) 
PD 0.47*** (0.1) 0.32*** (0.1) 0.46*** (0.08) 0.44*** (0.08) 0.48*** (0.08) 0.53*** (0.08) 0.58*** (0.08) 0.57*** (0.09) 0.69*** (0.11) 0.68*** (0.13) 
Intercept −4.61*** (1.0) −6.47** (1.09) −6.93*** (0.92) −6.45*** (0.86) −5.95*** (0.8) −5.04*** (0.78) −4.92*** (0.81) −5.41*** (0.89) −4.29*** (1.56) −1.8 (1.64) 
VariablesOLSQuantiles
0.10.20.30.40.50.60.70.80.9
DO 0.15*** (0.04) 0.05 (0.05) 0.07* (0.04) 0.07* (0.04) 0.010** (0.04) 0.11*** (0.04) 0.12*** (0.04) 0.14*** (0.05) 0.23*** (0.05) 0.32*** (0.04) 
GDPPC 0.20** (0.09) 0.09 (0.25) 0.36* (0.20) 0.34** (0.16) 0.39*** (0.15) 0.25* (0.13) 0.29** (0.13) 0.27** (0.13) 0.24 (0.19) 0.19 (0.18) 
Urban 0.14 (0.09) −0.07 (0.10) 0.02 (0.09) 0.07 (0.09) 0.08 (0.09) 0.17* (0.09) 0.16 (0.1) 0.24** (0.11) 0.26* (0.14) 0.09 (0.12) 
Lag −0.07 (0.12) −0.21* (0.11) −0.16* (0.09) −0.13 (0.09) −0.16* (0.09) −0.16* (0.09) −0.06 (0.11) −0.16 (0.11) −0.08 (0.14) 0.04 (0.13) 
IS 0.3 (0.21) 0.82* (0.42) 0.38 (0.33) 0.28 (0.27) 0.17 (0.2) 0.18 (0.2) 0.19 (0.22) 0.39 (0.28) 0.48 (0.4) 0.2 (0.29) 
PD 0.47*** (0.1) 0.32*** (0.1) 0.46*** (0.08) 0.44*** (0.08) 0.48*** (0.08) 0.53*** (0.08) 0.58*** (0.08) 0.57*** (0.09) 0.69*** (0.11) 0.68*** (0.13) 
Intercept −4.61*** (1.0) −6.47** (1.09) −6.93*** (0.92) −6.45*** (0.86) −5.95*** (0.8) −5.04*** (0.78) −4.92*** (0.81) −5.41*** (0.89) −4.29*** (1.56) −1.8 (1.64) 

Standard deviation in parentheses *p < 0.1; **p < 0.05; ***p < 0.01.

International mutual supervision among countries has also prompted some regions to accelerate the introduction of advanced international technologies. For example, the Regulations for International Water Resources Management Institutions are an international convention adopted in 1976. Its clauses state that all countries should jointly shoulder the responsibility for water quality control and work together to strengthen the protection of water resources. These international water resources protection conventions will prompt all regions to speed up introducing new technologies to improve the water resources environment. Therefore, these are closely related to the regional degree of opening. The degree of opening is defined as the ratio of total imports and exports to GDP, and the introduction of foreign technology and equipment belongs to the category of imports.

There are three levels at which opening to the outside world can improve the local water's comprehensive efficiency. First of all, from an ideological level, opening up to the outside world will help a region establish a sense of water resources protection and better recognize the global water resources crisis. Secondly, from the institutional level, opening to the outside world will help a region introduce advanced systems and methods from developed regions to improve water utilization efficiency and reduce water pollution. Finally, opening to the outside world will bring advanced equipment, technology, and talents to this area to improve water resource utilization and reduce sewage discharge.

GDPPC

GDP per capita plays a positive role in comprehensive water efficiency, especially in the 25th–50th and 50th–75th quantiles. With the increase of the quantile of comprehensive water efficiency, the coefficient of per capita GDP presents an inverted U-shape. It increases first and then decreases. This shows that the comprehensive water resource efficiencies in developed or underdeveloped regions are lower than those in developing regions (Figure 3).
Figure 3

Heat map for water utility efficiency, water governance efficiency, and comprehensive water efficiency for Yangtze River basin cities.

Figure 3

Heat map for water utility efficiency, water governance efficiency, and comprehensive water efficiency for Yangtze River basin cities.

Close modal

This is similar to the results found by some scholars who used the Super SBM model and panel quantile approach in their study (Zhou et al. 2020). GDP per capita is an important indicator to measure the development level. Developed areas tend to spend more resources on water resources protection and governance.

The comprehensive water efficiency in developing regions is higher than in developed and underdeveloped regions. This can be explained using marginal economic efficiency. The speed of local economic development is greater than the increase in water governance costs. In these areas, the wastewater treatment cost is much lower than the added value from the economic growth. For example, simple treatments such as activated carbon adsorption are much lower than the cost of advanced technology such as nano-membranes and biodegradation. Therefore, the increase in GDP is more significant than the investment in environmental governance.

Urbanization

There is a positive correlation between urbanization and comprehensive water efficiency. Other scholars have also confirmed this view (Ma et al. 2016). Urbanization helps to exert agglomeration and scale effects and facilitates centralized sewage treatment. Highly urbanized areas have a high proportion of non-agricultural industries, and agriculture has always been a high-water-consumption industry. Therefore, the urbanization process has improved the water consumption structure and helped to improve comprehensive water efficiency. Although there has been an emergence of new forms of agriculture, such as drip irrigation, this cannot change the fact that agriculture is still an industry with high water consumption and relatively low GDP output. The non-agricultural proportion has always been an important factor affecting water efficiency. Moreover, urbanization contributes to the renewal and improvement of technology, improving comprehensive water efficiency. Together with the development of industrialization, urbanization has greatly improved economic and operational efficiency.

Population urbanization lag

Population urbanization lag is negatively correlated with comprehensive water efficiency. As the quantile increases, the lag coefficient gradually becomes more prominent. This indicates that as lag hysteresis increases, its influence on comprehensive water efficiency weakens. The lag of urbanization means that the rate of land urbanization is greater than the population urbanization rate. After a simple calculation, it can be concluded that this will reduce the population density of the city and the level of urbanization. It will also bring about problems such as the idleness of urban land and unemployment in rural areas (Zhou et al. 2020). This will also damage the environment and increase the discharge of pollutants such as waste gas and wastewater.

Industrial structure

Industrial structure has a positive impact on comprehensive water efficiency. It is different from previous research on eco-efficiency. This implies that agriculture is an essential factor affecting comprehensive water efficiency. The higher the ratio of the secondary industry, the less the share of the primary industry represented by agriculture. Everything is inseparable from water. Regardless of the type of agriculture, the dependence on water resources is extreme. Agriculture has always been a significant industry of water consumption.

Population density

Population density plays a tremendously positive role in water resource efficiency for cities in all quantiles. The coefficient of population density is the largest, and all quantile statistics have passed the 1% level, indicating that it is very significant. Population growth will increase the consumption of resources and lead to more pollutant emissions (Yan et al. 2020). Water is an indispensable resource in human life and production activities. Population rise will inevitably lead to a substantial increase in water usage. The increase in population density may also produce scale and agglomeration effects (Cheng et al. 2017). The fact is that with more population moving into the city, the city scale continues to expand. Water facilities can be fully utilized, and the sewage is also convenient for concentrated and efficient treatment. These all contribute to the improvement of water resource efficiency. Therefore, the government should encourage more people to enter the city to strengthen this scale and agglomeration effect. It is also convenient to publicize people's water resources protection and raise people's awareness of water conservation and protection.

This paper uses the Super SBM modified RAM Network DEA model with undesirable output to study the comprehensive efficiency of water resources through data analysis of 30 cities in China's Yangtze River Basin from 2005 to 2015. This study has the following accomplishments:

Firstly, a new DEA model named the SHAN model is built to reveal the interrelationship inside the DMUs. By adopting this model, the comprehensive water efficiency can be decomposed into two stages: the water utility stage and the water governance stage. This paper finds that the water utility is a crucial stage affecting comprehensive water efficiency.

Secondly, the degree of openness plays a positive role in the comprehensive efficiency of cities along the Yangtze River Basin. As the quantile of comprehensive water efficiency increases, the effectiveness of the degree of openness on comprehensive water efficiency increases. This finding indicates that China's opening and internationalization policy can impact the local environment.

Thirdly, the application of panel data regression enables this study to analyze some economic indicators that affect comprehensive water efficiency. Specifically, GDP per capita plays a positive role in comprehensive water efficiency. The improvement of comprehensive water efficiency shows an inverted U-shape relationship between GDP per capita and comprehensive water efficiency. In addition, this paper finds that population urbanization lag harms comprehensive water efficiency. Population density has a significantly positive effect on comprehensive water efficiency. Based on the above results, this article makes the following policy recommendations:

  • 1.

    The government should increase the degree of opening to the outside world and accelerate the introduction of advanced water resource management concepts, systems, technical equipment, and talents from developed regions. Moreover, the government should also strengthen supervision, introduce strict sewage discharge standards, and supervise the reduction of total wastewater discharge. Moreover, more investment in water utilization and governance should be made.

  • 2.

    The government should appropriately speed up the process of urbanization and remove the barriers between urban and rural areas. This will increase population density, improve the scale effect and aggregation effect, and improve water resource utilization efficiency and sewage treatment rate. Administrative attention also should be paid to reduce the negative impact of population urbanization lag and avoid idle land in cities.

  • 3.

    Local government should find a proper balance between economic growth and environmental protection. At this stage, the main task of local governments should still be to focus on the improvement of water utilization efficiency and achieving the optimized balance between economic progress and environmental governance.

  • 4.

    Cities should adjust industrial structure and prioritize the development of those industries with better water utilization efficiency, such as tertiary industry. Moreover, water resources should be allocated more wisely.

We are thankful for the suggestions and efforts from the reviewers and editors

The authors have no relevant financial or non-financial interests to disclose.

The authors attest that this paper has not been published elsewhere, the work has not been submitted simultaneously for publication elsewhere and the results presented in this work are true and not manipulated.

Shan, Huang: Conceptualization, Methodology, Software, Data curation, Writing – Original draft preparation, Visualization, Investigation, Validation, Writing – Reviewing and Editing. Ying, Kong: Supervision.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

The authors declare there is no conflict.

Andersen
P.
&
Petersen
N. C.
1993
A procedure for ranking efficient units in Data Envelopment Analysis
.
Management Science
39
(
10
),
1261
1264
.
Bao
C.
,
Chen
X. J.
&
Liang
G. L.
2016
Analysis on the influencing factors of water use efficiency in Henan Province based on spatial econometric models
.
Journal of Natural Resources
31
(
7
),
1138
1148
.
Charnes
A.
,
Cooper
W. W.
&
Rhodes
E.
1978
Measuring the efficiency of decision making units
.
European Journal of Operational Research
2
(
6
),
429
444
.
Chen
G.
2014
Data Envelopment Analysis Method and Max DEA Software
.
Intellectual Property Publishing House
.
Beijing, China (in Chinese)
Du
Y. Q.
,
Su
S. P.
,
Wang
Y. P.
&
Sun
X. X.
2015
Analysis on utilization efficiency and influencing factors of water resource in Fujian province based on DEA–tobit model
.
Resource Development & Market
31
,
46
50
(in Chinese)
.
Escher
B. I.
,
Allinson
M.
,
Altenburger
R.
,
Bain
P. A.
,
Balaguer
P.
,
Busch
W.
,
Crago
J.
,
Denslow
N. D.
,
Dopp
E.
,
Hilscherova
K.
,
Humpage
A. R.
,
Kumar
A.
,
Grimaldi
M.
,
Jayasinghe
B. S.
,
Jarosova
B.
,
Jia
A.
,
Makarov
S.
,
Maruya
K. A.
,
Medvedev
A.
,
Mehinto
A. C.
,
Mendez
J. E.
,
Poulsen
A.
,
Prochazka
E.
,
Richard
J.
,
Schifferli
A.
,
Schlenk
D.
,
Scholz
S.
,
Shiraishi
F.
,
Snyder
S.
,
Su
G.
,
Tang
J. Y. M.
,
van der Burg
B.
,
van der Linden
S. C.
,
Werner
I.
,
Westerheide
S. D.
,
Wong
C. K. C.
,
Yang
M.
,
Yeung
B. H. Y.
,
Zhang
X.
&
Leusch
F. D. L.
2014
Benchmarking organic micropollutants in wastewater, recycled water and drinking water with in vitro bioassays
.
Environmental Science & Technology
48
(
3
),
1940
1956
.
Färe
R.
,
Grosskopf
S.
&
Whittaker
G.
2007
Network DEA. In: Modeling Data Irregularities and Structural Complexities in Data Envelopment Analysis (J. Zhu & W. D. Cook, eds), Springer, Boston, MA, USA, pp. 209–40
.
Hu
J.-L.
,
Wang
S.-C.
&
Yeh
F.-Y.
2006
Total-factor water efficiency of regions in China
.
Resources Policy
31
(
4
),
217
230
.
Jagtap
S.
,
Skouteris
G.
,
Choudhari
V.
&
Rahimifard
S.
2022
Improving water efficiency in the beverage industry with the internet of things. In: Implementing Data Analytics and Architectures for Next Generation Wireless Communications (C. Bhatt, N. Kumar, A. K. Bashir & M. Alazab, eds), IGI Global, Hershey, PA, USA, pp. 18–26
.
Lilienfeld
A.
&
Asmild
M.
2007
Estimation of excess water use in irrigated agriculture: a Data Envelopment Analysis Approach
.
Agricultural Water Management
94
(
1–3
),
73
82
.
Maruyama
Y.
2009
Range adjusted measure network DEA model
.
AIP Conference Proceedings
1168, pp.
949
952
.
Nicolaisen
B.
2003
Developments in membrane technology for water treatment
.
Desalination
153
(
1–3
),
355
360
.
Staal
L. B.
,
Petersen
A. B.
,
Jørgensen
C. A.
,
Nielsen
U. G.
,
Nielsen
P. H.
&
Reitzel
K.
2019
Extraction and quantification of polyphosphates in activated sludge from waste water treatment plants by P NMR spectroscopy
.
Water Research
157
,
346
355
.
Tian
J.
&
Wang
Y.
2019
The impact of China’s population urbanization lag on carbon emissions
.
Journal of Environmental Economics
(in Chinese)
.
Tone
K.
2001
A slacks-based measure of efficiency in data envelopment analysis
.
European Journal of Operational Research
130
(
3
),
498
509
.
Tone
K.
2002
A slacks-based measure of super-efficiency in data envelopment analysis
.
European Journal of Operational Research
143
(
1
),
32
41
.
Tone
K.
&
Tsutsui
M.
2009
Network DEA: a slacks-based measure approach
.
European Journal of Operational Research
197
(
1
),
243
252
.
Wang
S. Y.
&
Amp
C. E.
2019
Research on the waste water treatment and reuse technology of scheelite floatation effluent. Hunan Nonferrous Metals (in Chinese).
Wei
Q. L.
2013
On ‘opening black box evaluation’ network DEA models
.
Mathematics in Practice and Theory
42
(
24
),
184
195
.
(in Chinese).
Yan
D.
,
Kong
Y.
,
Ren
X.
,
Shi
Y.
&
Chiang
S. W.
2019
The determinants of urban sustainability in Chinese resource-based cities: a panel quantile regression approach
.
Science of the Total Environment
686
,
1210
1219
.
Yuan
H.
2002
Challenge and perspective of Yangtze River Basin water resources protection
.
Yangtze River
2002
(
S1
),
50
53
.
Zhao
L.
,
Sun
C.
&
Zheng
D.
2014
Water resource utilization efficiency and its spatial spillover effects measure in China
.
Acta Geographica Sinica
69
(
1
),
121
133
.
Zhou
X.
,
Luo
R.
,
Yao
L.
,
Cao
S.
,
Wang
S.
&
Lev
B.
2018
Assessing integrated water use and wastewater treatment systems in China: a mixed network structure two-stage SBM DEA model
.
Journal of Cleaner Production
185
,
533
546
.
Zhou
Y.
,
Kong
Y.
,
Wang
H.
&
Luo
F.
2020
The impact of population urbanization lag on eco-efficiency: a panel quantile approach
.
Journal of Cleaner Production
244
,
118664
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data