EN中文

Journal of Geographical Sciences >

2017 , Vol. 27 >Issue 12: 1450 - 1462

DOI: https://doi.org/10.1007/s11442-017-1446-9

Research Articles

Spatial econometric analysis on influencing factors of water consumption efficiency in urbanizing China

  • BAO Chao , 1, 2, 3 ,
  • CHEN Xiaojie 1, 2, 3, 4
Expand
  • 1. Institute of Geographic Sciences and Natural Resource Research, CAS, Beijing 100101, China
  • 2. Key Laboratory of Regional Sustainable Development Modeling, CAS, Beijing 100101, China
  • 3. College of Resources and Environment, University of Chinese Academy of Sciences, Beijing 100049, China
  • 4. China Urban Construction Design & Research Institute CO.LTD, Beijing 100120, China;

Author: Bao Chao (1978-), PhD and Associate Professor, specialized in urbanization and urban sustainable development. E-mail:

Received date: 2017-03-04

  Accepted date: 2017-06-13

  Online published: 2017-12-10

Supported by

Major Projects of the National Natural Science Foundation of China, No.41590844

National Natural Science Foundation of China, No.41571156

Service Project on the Cultivation and Construction for the Characteristic Research Institute of the Chinese Academy of Sciences, No.TSYJS02

Copyright

Journal of Geographical Sciences, All Rights Reserved
Fold

Abstract

Due to the limitation of total amount of water resources, it is necessary to enhance water consumption efficiency to meet the increasing water demand of urbanizing China. Based on the panel data of 31 provinces in China in 1997-2013, we analyze the influencing factors of water consumption efficiency by spatial econometric models. Results show that, 1) Due to the notable spatial autocorrelation characteristics of water consumption efficiency among different provinces in China, general panel data regression model which previous studies often used may be improper to reveal its influencing factors. However, spatial Durbin model may best estimate their relationship. 2) Water consumption efficiency of a certain province may be influenced not only by its socio-economic and eco-environmental indicators, but also by water consumption efficiency in its neighboring provinces. Moreover, it may be influenced by the neighboring provinces’ socio-economic and eco-environmental indicators. 3) For the macro average case of the 31 provinces in China, if water consumption efficiency in neighboring provinces increased 1%, water consumption efficiency of the local province would increase 0.34%. 4) Among the ten specific indicators we selected, per capita GDP and urbanization level of itself and its neighboring provinces have the most prominent positive effects on water consumption efficiency, and the indirect effects of neighboring provinces are much larger. Therefore, the spatial spillover effects of the economic development level and urbanization level are the primary influencing factors for improving China’s water consumption efficiency. 5) Policy implications indicate that, to improve water consumption efficiency, each province should properly consider potential influences caused by its neighboring provinces, especially needs to enhance the economic cooperation and urbanization interaction with neighboring provinces.

Key words： water consumption efficiency; water resources management; urbanization; spatial spillover effects; spatial Durbin model

Cite this article

BAO Chao , CHEN Xiaojie . Spatial econometric analysis on influencing factors of water consumption efficiency in urbanizing China[J]. Journal of Geographical Sciences, 2017 , 27(12) : 1450 -1462 . DOI: 10.1007/s11442-017-1446-9

1 Introduction

During rapid urbanization, China has confronted steady increase of water consumption in the past few decades (Bao and He, 2015). However, China’s available water resources are limited and unevenly distributed (Jiang, 2009). Water scarcity and related eco-environmental degeneration have significantly hindered China’s socio-economic development (Bao and Fang, 2007; Fang et al., 2007; Li and Ma, 2014). Constrained by total water resources and increasingly difficult cross-regional water diversion, it is necessary to improve water consumption efficiency for the decoupling of economic growth and water resources utilization in China, especially in water scarce regions and urbanizing areas (Cai, 2008; Bao and Fang, 2012; Bao and Chen, 2015).
Chinese government has set up “The Most Stringent Water Management System” in January 2012 and established “red lines” of water consumption efficiency for each province. The growth rate of water consumption efficiency in China and each province should be larger than the standard which the Ministry of Water Resources of China has planned (Huang, 2015; Zang et al., 2016). Many scholars have studied China’s water consumption efficiency and relevant influencing factors by various methods, including stochastic frontier analysis (SFA), data envelopment analysis (DEA), slacks-based measure (SBM) model, undesirable and meta-frontier model, directional distance function, etc. (Kaneko et al., 2004; Li and Ma, 2014; Sun et al., 2014; Wang et al., 2015; Ma et al., 2016; Deng et al., 2016). They have estimated the absolute or relative value of water consumption efficiency for China’s provinces, and revealed the main influencing factors, including water resources endowment, water consumption structure, industrial structure, technical progress, economic development level, per capita education expense, etc.
However, the variation of China’s water consumption efficiency is complex and the dominant influencing factors vary in different regions and development stages. Current studies seldom take the spatial interactions of water consumption efficiency and relevant influencing factors into consideration (Li and Ma, 2015). Most of them ignored the spatial spillover effects (Sun et al., 2014). For example, we used “water consumption efficiency” (or “water use efficiency” or “water resources utilization efficiency”) and “spatial spillover effect” (or “spatial effect” or “spatial interaction” or “spatial econometric”) as topics (including Title, Abstracts and Keywords) to search literatures based on Web of Science - Science Citation Index Expanded (1900 to present) and Web of Science - Social Sciences Citation Index (1996 to present). There are only 20 results and two of them are related to China. In fact, water consumption efficiency of a certain province might be influenced not only by its socio-economic and eco-environmental indicators, but also by water consumption efficiency in its neighboring provinces. Moreover, it might be influenced by the neighboring provinces’ socio-economic and eco-environmental indicators (Sun et al., 2014; Li and Ma, 2015).
To address the above gaps and verify the above viewpoint, this paper explained spatial econometric models in Section 2, and applied them to investigate the spatial interactions between water consumption efficiency and its relevant influencing factors. Our empirical analysis used the panel data of 31 provinces in China during the period of 1997-2013 to examine the best appropriate model. Results and outcomes were presented in Section 3. Conclusions and implications were put forward in Section 4. It might help to thoroughly understand the influencing factors of China’s water consumption efficiency from a geographical perspective, and promote a more efficient use of water resources in urbanizing China.

2 Methodology and data sources

2.1 Spatial econometric methodology

Spatial econometric model emerged when econometrics turned to focus on spatial characters (Elhorst, 2010; Halleck and Elhorst, 2015). It originates general panel data regression model. However, the general panel data regression model can only reveal the relationship between the explained variable and the explanatory variable without considering the spatial spillover effects of the explained variable. When the explained variable has significant spatial spillover effects, if we use the general panel data regression model, the parameter estimated by ordinary least squares will be biased and inconsistent (Lesage and Pace, 2009). That’s to say, the regression equation cannot pass the validity test. Under this circumstance, we may use spatial econometric model to resolve the problem.
Before constructing a spatial econometric model, we may judge the spatial dependence of the explained variable by spatial autocorrelation analysis preliminarily. If spatial dependence exists, we can use Lagrange Multiplier (LM) to test whether a model without any spatial interaction effect is better to describe the data than the spatial lag model or spatial error model. If the model without any spatial interaction effect is rejected, we can use Wald test or Likelihood Ratio (LR) test to judge whether spatial Durbin model can be simplified to spatial lag model or spatial error model. If it can’t be simplified, it indicates that spatial Durbin model best describes the data, and we can use Hausman test or Likelihood Ratio (LR) test to judge whether a spatial Durbin model with spatial and/or time fixed or random effects is more appropriate (Lesage and Pace, 2009). The specific methods go as follows:
2.1.1 General panel data regression model
As for panel data, if they are stationary and cointegrated, a general panel data regression model can be used to reveal the relationship between the explained and explanatory variable, regardless of the spatial dependence. The general form is as follows:
yit = φ + xit β + ci + αt + εit (1)
where yit is the explained variable in region i and period t; xit is the explanatory variable in region i and period t; β is the regression coefficient of the explanatory variable, representing the influence of xit on yit; φ is the constant term; ci and αt are the intercept term in region i and period t respectively; εit is the error term; εit~N(μ, σ2). In spatial (or time) fixed effects model, ci (or αt) and εit are dependent. However, they are independent in spatial (or time) random effects model.
2.1.2 Spatial autocorrelation analysis
Spatial autocorrelation analysis is a useful method to describe the spatial dependence and spatial heterogeneity by panel data. If the variable has spatial autocorrelation characteristics, it indicates that a spatial econometric model is better to describe the data than a general panel data regression model. In 1950, Moran proposed Global Moran’s I indicator based on spatial stochastic distribution phenomenon (Moran, 1950), which is widely used as an index to measure spatial autocorrelation (Lin and Wang, 2016). The specific calculation formula is as follows:
\[{{I}_{t}}=\frac{n}{\sum\limits_{i=1}^{n}{\sum\limits_{j=1}^{n}{{{w}_{ijt}}}}}\times \frac{\sum\limits_{i=1}^{n}{\sum\limits_{j=1}^{n}{{{w}_{ijt}}({{x}_{it}}-{{{\bar{x}}}_{t}})}}({{x}_{jt}}-{{{\bar{x}}}_{t}})}{\sum\limits_{i=1}^{n}{{{({{x}_{it}}-{{{\bar{x}}}_{t}})}^{2}}}} \ \ (2)\]
where It is the Global Moran’s I index of the whole region in period t; xit is the research variable in region i and period t; \({{\bar{x}}_{t}}\) is the mean value of the research variable in period t; n is the number of spatial units; wijt is the spatial weights matrix of region i and region j in period t, and if the ith and jth spatial units are neighboring, wijt = 1; if not, wijt = 0. It varies from -1 to 1. If it is above zero, it indicates a positive spatial autocorrelation. If it is less than zero, it indicates a negative spatial autocorrelation. If it is close to zero, it indicates that the research variable is randomly distributed. Furthermore, the bigger the absolute value of It is, the greater the spatial autocorrelation is.
2.1.3 Spatial econometric model
Spatial econometric models comprise spatial lag model (SLM), spatial error model (SEM) and spatial Durbin model (SDM) (Lesage and Pace, 2009; Lee and Yu, 2010). Specifically, spatial lag model contains the spatial lag term of the explained variable in general panel data regression model, which indicates the interaction effect of the explained variable in neighboring units on the explained variable in a specific unit. Spatial error model contains a spatial error term, which indicates the interaction effect of the explanatory variables in neighboring units on the explained variable in a specific unit. However, spatial Durbin model is a comprehensive form of them, and contains the spatial lag term of the explained variable and the spatial error term of the explanatory variables. The three kinds of spatial econometric models are expressed as follows (Elhorst, 2003, 2014):
\[{{y}_{it}}=\lambda \sum\limits_{j=1}^{N}{{{W}_{ij}}{{y}_{jt}}}+\phi +{{x}_{it}}\beta +{{c}_{i}}+{{\alpha }_{t}}+{{\varepsilon }_{it}} \ \ (3)\]
\[{{y}_{it}}=\phi +{{x}_{it}}\beta +{{c}_{i}}+{{\alpha }_{t}}+{{u}_{it}}{{u}_{it}}=\rho \sum\limits_{j=1}^{N}{{{W}_{ij}}{{u}_{jt}}}+{{\varepsilon }_{it}} \ \ (4)\]
\[{{y}_{it}}=\lambda \sum\limits_{j=1}^{N}{{{W}_{ij}}{{y}_{jt}}}+\phi +{{x}_{it}}\beta +\sum\limits_{j=1}^{N}{{{W}_{ij}}{{x}_{ijt}}\theta }+{{c}_{i}}+{{\alpha }_{t}}+{{\varepsilon }_{it}} \ \ (5)\]
where λ is the spatial regression coefficient, which indicates the influence of the explained variable in neighboring spatial units (marked as yn) on the explained variable in the local spatial unit (marked as ys). If λ is significantly positive, it indicates that the improvement of the explained variable in a spatial unit may correspond with the improvement of this variable in other spatial units in the study area. uit is the spatial autocorrelation error term; ρ is the spatial error coefficient, which indicates the influence of the spatial error term in neighboring spatial units (marked as un) on the spatial error term in the local spatial unit (marked as us); θ is the spatial lag term coefficient of the explanatory variable, which indicates the influence of the explanatory variables in neighboring spatial units (marked as xn) on ys; N is number of spatial units; Wij is spatial weights matrix, and if the i, jth spatial unit is neighboring, Wij = 1; if not, Wij = 0; the rest items are of the same as those noted in the general panel data regression model.
To judge whether the spatial lag model or spatial error model is better to describe the data, we may use LM test to examine the significance of the spatial lag effect and the spatial error effect. If one of them is significant while the other is not, we can choose the significant model. However, if both of them are significant or insignificant, we need apply the spatial Durbin model. Then we can use Wald test or LR test to test the hypotheses H0:θ = 0 and H0:θ + λβ = 0. From the first hypothesis, we can decide whether it can be simplified to the spatial lag model. From the second hypothesis, we can decide whether it can be simplified to the spatial error model. If both hypotheses are rejected, it cannot be simplified and we should use spatial Durbin model itself (Burridge, 1981; Elhorst, 2014).
2.1.4 Estimation of direct and indirect effect
From the formula of spatial Durbin model, we can see that the explained variable in the local spatial unit (marked as ys) is mainly influenced by three factors. One is the explained variable in neighboring spatial units (marked as yn). The second is the explanatory variables in the local spatial unit (marked as xs). The third is the explanatory variables in neighboring spatial units (marked as xn). Obviously, xs can have direct influence on ys(xs→ys). It also can influence yn and then influence ys through spatial autocorrelation of yn(xs→yn→ys). These two kinds of combined effects which xs has on ys are named direct effect of the explanatory variables. Similarly, xs can have direct influence on yn (xs→yn). It also can influence ys and then influence yn through spatial autocorrelation of ys(xs→ys→yn). These two kinds of combined effects which xs has on yn are named indirect effect of the explanatory variables, or spatial spillover effect of the explanatory variables.
In fact, direct and indirect effects are combined influences which the explanatory variables have on the explained variable through complex interaction. They are different from the regression coefficient of the explanatory variable β(xs→ys) and the spatial lag term coefficient θ(xn→ys) in formula (5). Furthermore, spatial Durbin model can be adapted into the following formula (Lesage and Pace, 2009; Elhorst, 2014):
\[{{Y}_{t}}={{(I-\lambda W)}^{-1}}\varphi {{\iota }_{N}}+{{(I-\lambda W)}^{-1}}({{X}_{t}}\beta +W{{X}_{t}}\theta )+{{(I-\lambda W)}^{-1}}\nu _{t}^{*} \ \ (6) \]
where φtN is the spatial error term; is the time error term with time and spatial effects. Thus, the partial derivative of the explained variable with respect to the kth explanatory variable at a particular point in time is as follows:
\[\left[ \frac{\partial Y}{\partial {{x}_{1k}}}\cdot \frac{\partial Y}{\partial {{x}_{Nk}}} \right]={{(I-\lambda W)}^{-1}}\left[ \begin{matrix} {{\beta }_{k}} & {{w}_{12}}{{\theta }_{k}} & \cdot & {{w}_{1N}}{{\theta }_{k}} \\ {{w}_{21}}{{\theta }_{k}} & {{\beta }_{k}} & \cdot & {{w}_{2N}}{{\theta }_{k}} \\ \cdot & \cdot & \cdot & \cdot \\ {{w}_{N1}}{{\theta }_{k}} & {{w}_{N2}}{{\theta }_{k}} & \cdot & {{\beta }_{k}} \\ \end{matrix} \right] \ \ (7)\]
In the above equation, the average of the diagonal elements of the matrix on the right-hand side is defined as the direct effect, and the average of either t.he row sums or the column sums of the off-diagonal elements of that matrix is defined as the indirect effect (Lesage and Pace, 2009; Elhorst, 2014).

2.2 Data sources and descriptive statistics

According to literature review and data availability, we use frequently used statistical indicators to represent water consumption efficiency and its influencing factors (Table 1). Specifically, water consumption efficiency is the explained variable. It is usually indicated by the economic output of per cubic meter water consumption, or the water consumption of per economic output. Normally, Chinese used to use water consumption per ten thousand yuan GDP to measure it. According to China Water Resources Bulletin, total water consumption is divided into four categories, such as water consumption for agriculture, water consumption for industry, water consumption for residents, and water consumption for eco-environment provided by artificial measures. Among them, water consumption for agriculture includes water consumption for farmland irrigation, water consumption for forestry, fishery and animal husbandry. Water consumption for residents includes domestic water for rural and urban residents. The influencing factors are the explanatory variables, including water resources endowment, water resources development level, economic development level, urbanization level, industrial structure optimization level, investment level, social consumption level, urban and rural income level, and agricultural development level, which are indicated by ten specific indicators (Table 1). It should be pointed out that water consumption per ten thousand yuan GDP is a negative indicator while others are positive indicators.
Table 1 The description or calculation of variables used in the present study
Variable Description or calculation
Water consumption efficiency Total water consumption / GDP
Per capita water resources Total water resources / Total population
Utilization ratio of water resources Total water consumption / Total water resources
Per capita GDP GDP / Total population
Urbanization level Urban population / Total population
Ratio of value added of the tertiary industry in GDP Value added of the tertiary industry / GDP
Per capita total investment in fixed assets Total investment in fixed assets / Total population
Per capita social retail sales of consumer goods Social retail sales of consumer goods / Total population
Urban per capita disposable income National sample survey data
Rural per capita net income National sample survey data
Per capita grain yield Total grain yield / Total population
Data of the above variables in 1997-2013 can be directly obtained or calculated based on China Statistic Yearbook and China Water Resources Bulletin. However, China’s population data have poor comparability among different periods and regions (Zhou and Yu, 2002). For higher comparability, we use the population data in China and its 31 provinces which were modified and forecasted by Shen (2006) and Fang et al. (2009). In addition, we use economic data in China and its 31 provinces at constant price in 1997.

3 Empirical results and outcomes

3.1 Confusing information of general panel data regression model

Before constructing the general panel data regression model, we use unit root test to check the stationarity of the panel data series (Dickey and Fuller, 1979). Results show that such indicators as water consumption per ten thousand yuan GDP, per capita water resources, utilization ratio of water resources, ratio of value added of the tertiary industry in GDP are stationary series. Other indicators, including per capita GDP, urbanization level, per capita total investment in fixed assets, per capita social retail sales of consumer goods, urban and rural per capita disposable income, and per capita grain yield are non-stationary series. However, they are all stationary after being expressed in natural logs. Therefore, these indicators are all expressed in natural logs. Subsequently, we use cointegration test to check the cointegration relationship (Johansen and Juselius, 1990). Results show that there are cointegration relationships among these variables.
For a general panel data regression model, the selection of time or spatial form may have a strong effect on the simulation results. In the perspective of mathematic, unobserved effects in fixed effects model (ci or αt) are related to the error term (εit) because there are still unknown factors that influence the explained variable besides the observed ones. Consequently, the fixed effects model can only be used to describe the observed objects. The results cannot be expanded in spatial or time series. In contrast, unobserved effects in random effects model are independent to the error term because there are no other factors that influence the explained variable besides the observed ones. Consequently, the random effects model can be expanded in spatial or time series. Due to the particularity of each geographic unit, spatial fixed effects model tends to work better in spatial sequence. However, due to the continuity of socio-economic development in time series, time random effects model tends to work better. Therefore, we use the spatial fixed and time random form in the general panel data regression model.
The results of the general panel data regression model are listed in Table 2. It shows that, per capita water resources, utilization ratio of water resources and per capita social retail sales of consumer goods have significant positive effects on water consumption per ten thousand yuan GDP. However, urbanization level, per capita total investment in fixed assets and urban per capita disposable income have significant negative effects. Other indicators, including per capita GDP, ratio of value added of the tertiary industry in GDP, rural per capita net income and per capita grain production have no significant effects on water consumption efficiency.
Table 2 Results of general panel data regression model for water consumption efficiency in China
Variable Regression coefficient t
Intercept term 6934.39 5.85(0.0000)***
Per capita water resources 0.02 4.69(0.0000)***
Utilization ratio of water resources 0.86 5.01(0.0000)***
Per capita GDP 257.05 1.37(0.1698)
Urbanization level -605.77 -2.52(0.0119)**
Ratio of value added of the tertiary industry in GDP -2.68 -0.63(0.5316)
Per capita total investment in fixed assets -183.01 -2.62(0.0090)***
Per capita social retail sales of consumer goods 199.55 2.21(0.0273)**
Urban per capita disposable income -712.97 -3.79(0.0002)***
Rural per capita net income 46.32 0.29(0.7709)
Per capita grain yield -108.18 -1.41(0.1606)

Note: p is listed in the bracket; ***, **, * denote level of significance at 1%, 5% and 10% respectively.

The above results are inconsistent with our qualitative cognition. In general, per capita GDP, which is positively correlated to urbanization level, per capita total investment in fixed assets and per capita social retail sales of consumer goods, should have a significant positive effect on water consumption efficiency. However, when we used spatio-temporal mixed data, the results may be confusing and misleading because per capita GDP might be positively correlated to water consumption efficiency in time series while negatively correlated to water consumption efficiency in spatial series.

3.2 Spatial autocorrelation, spatial lag and spatial error effects test

We use Global Moran’s I indictor to test spatial autocorrelation of water consumption efficiency for provincial China. Results show that Global Moran's I of China’s water consumption efficiency fluctuated around 0.5. It indicates that water consumption efficiency has significant positive spatial autocorrelation among different provinces in 1997-2013, and water consumption efficiency in neighboring provinces may have a positive effect on itself. Therefore, it’s necessary to build spatial econometric model to improve the accuracy of the general panel data regression model.
We use LM test and Robustness LM test to examine the spatial lag effects and the spatial error effects in four different forms, including time and spatial random effects, time random and spatial fixed effects, spatial random and time fixed effects, time and spatial fixed effects. The results are listed in Table 3. It shows that only the spatial error effects in the form of time and spatial random effects, spatial random and time fixed effects fail to pass the Robustness LM test. The spatial error and spatial lag effects in other forms all pass the LM test and Robustness LM test. Besides, the spatial lag effects are more significant than spatial error effects. Therefore, it’s necessary to further propose a spatial Dubin model.
Table 3 Results of LM test with different spatial-specific and time-specific effects in China
Item Time and spatial
random effects		 Time random and
spatial fixed effects		 Spatial random and
time fixed effects		 Time and spatial fixed effects
Spatial lag effect
LM test		 41.11
(0.000)***		 71.73
(0.000)***		 30.43
(0.000)***		 16.33
(0.000)***
Spatial lag effect
Robustness LM test		 13.98
(0.000)***		 33.8
(0.000)***		 10.29
(0.001)***		 27.74
(0.000)***
Spatial error effect
LM test		 27.38
(0.000)***		 50.19
(0.000)***		 20.48
(0.000)***		 6.23
(0.013)**
Spatial error effect
Robustness LM test		 0.24
(0.621)		 12.26
(0.000)***		 0.34
(0.560)		 17.65
(0.000)***

Note: p is listed in the bracket; ***, **, * denote level of significance at 1%, 5% and 10% respectively.

3.3 Best estimation of spatial Dubin model

We use Hausman test to examine the spatial Dubin model in four different forms, including time and spatial random effects, time random and spatial fixed effects, spatial random and time fixed effects, and time and spatial fixed effects. The results show that, spatial Dubin model with time and spatial random effects is rejected while spatial Dubin model with time and spatial fixed effects is acceptable. Spatial Dubin model with spatial random and time fixed effects is rejected while spatial Dubin model with time random and spatial fixed effects is acceptable. According to the results of Hausman test and empirical experience, we finally choose spatial Dubin model with time random and spatial fixed effects to estimate the relationship between water consumption efficiency and relevant influencing factors. Subsequently, we use Wald test to examine the spatial Dubin model with time random and spatial fixed effects. Results show that it can’t be simplified to spatial lag or spatial error model. Therefore, the spatial Dubin model with time random and spatial fixed effects best describes the data. The results are listed in Table 4.
Table 4 Results of spatial Dubin model with time random and spatial fixed effects in China
Variable Regression coefficient Lag term coefficient
β t θ t
Spatial regression coefficient λ 0.34 6.20 (0.0000)*** - -
Per capita water resources 0.01 3.35 (0.0008) *** 0.01 0.81 (0.4152)
Utilization ratio of water resources 1.00 6.10 (0.0000) *** -1.65 -4.26 (0.0000) ***
Per capita GDP -456.17 -2.48 (0.0132) ** -2044.96 -5.75 (0.0000) ***
Urbanization level -186.48 -0.75 (0.0045) *** -976.51 -1.94 (0.0520) *
Ratio of value added of tertiary industry in GDP -4.23 -1.01 (0.3114) -24.1 -2.49 (0.0130) **
Per capita total investment in fixed assets -115.88 -1.77 (0.0772)* 19.9 0.15 (0.8793)
Per capita social retail sales of consumer goods -255.41 -3.05 (0.0023) *** 79.76 0.46 (0.6488)
Urban per capita disposable income -45.23 -0.17 (0.8640) 467.9 1.20 (0.2311)
Rural per capita net income -117.76 -0.50 (0.6171) 1479.13 4.31 (0.0000) ***
Per capita grain yield -106.05 -1.25 (0.2104) -431.69 -2.61 (0.0090) ***

Note: p is listed in the bracket; ***, **, * denote level of significance at 1%, 5% and 10% respectively.

Firstly, the spatial regression coefficient λ is significantly positive. It indicates that water consumption efficiency in one province is obviously influenced by its neighboring provinces. At the same time, it also has obvious positive effects on water consumption efficiency in neighboring provinces. In other words, water consumption efficiency has obvious spatial spillover effects in provincial China. For the macro average case of the 31 provinces in China, if water consumption efficiency in neighboring provinces increased 1%, water consumption efficiency of the local province would increase 0.34%.
Secondly, the significant test of the regression coefficient β indicates that water consumption efficiency of each province in China is influenced by six indicators of itself, including per capita water resources, utilization ratio of water resources, per capita GDP, urbanization level, per capita total investment in fixed assets and per capita social retail sales of consumer goods. However, the other four indicators have no significant effect on water consumption efficiency in itself. From the values of the regression coefficient β of those six indicators which have significant effects, two indicators such as per capita water resources and utilization ratio of water resources have positive effects on water consumption per ten thousand yuan GDP, or have negative effects on water consumption efficiency. The other four indicators all have negative effects on water consumption per ten thousand yuan GDP, or have positive effects on water consumption efficiency.
Finally, the significant test of the lag term coefficient θ indicates that water consumption efficiency of each province in China is influenced by six indicators of its neighboring provinces, including utilization ratio of water resources, per capita GDP, urbanization level, ratio of value added of the tertiary industry in GDP, rural per capita net income and per capita grain yield. However, the other four indicators of its neighboring provinces have no significant effect. From the values of the lag term coefficient θ of those six indicators which have significant effect, only rural per capita net income of its neighboring provinces has positive effects on water consumption per ten thousand yuan GDP, or has negative effects on water consumption efficiency. The other five indicators of its neighboring provinces all have negative effects on water consumption per ten thousand yuan GDP, or have positive effects on water consumption efficiency.

3.4 Direct and indirect effects of different influencing factors

To be mentioned, the regression coefficient β and the lag term coefficient θ only indicate the interaction effects which the observed indicators of the local province and its neighboring provinces have on water consumption efficiency of itself. The secondary interaction effects through spatial spillovers of water consumption efficiency are not considered. Therefore, based on the results of spatial Dubin model, we further calculate the direct and indirect effects of different influencing factors on China’s water consumption efficiency. We listed the results in Table 5. From it, we can determine whether the direct and indirect effects are positive or negative according to the significance of their coefficients. We can also determine which kind of effects is larger according to their absolute values.
Table 5 Direct and indirect effects of different influencing factors on China’s water consumption efficiency
Variable Direct effects Indirect effects
Coefficient t Coefficient t
Per capita water resources 0.02 3.41 (0.0018)*** 0.02 1.34 (0.1897)
Utilization ratio of water resources 1.23 5.41 (0.0000)*** -1.87 -3.32 (0.0023)***
Per capita GDP -692.34 -1.56 (0.0012)*** -2701.77 -5.01 (0.0000)***
Urbanization level -192.87 -1.08 (0.0028)*** -1513.65 -2.17 (0.0374)**
Ratio of value added of the tertiary industry in GDP -12.12 -0.58 (0.5657) -32.03 -2.19 (0.0362)**
Per capita total investment in fixed assets -24.49 -1.67 (0.1042) -25.68 -0.13 (0.8938)
Per capita social retail sales of consumer goods -432.92 -3.1 (0.0041)*** -242.73 -0.98 (0.3358)
Urban per capita disposable income -488.69 -0.09 (0.9266) -656.44 1.28 (0.2092)
Rural per capita net income 448.58 0.02 (0.9863) 2052.82 4.81 (0.0000)***
Per capita grain yield 6.82 1.83 (0.0768)* -679.06 -3.03 (0.0050)***

Note: p is listed in the bracket; ***, **, * denote level of significance at 1%, 5% and 10% respectively.

Specifically, we can see that only the direct and indirect effects of two indicators such as per capita GDP and urbanization level are significantly negative. It indicates that these two indicators have significantly positive effect on water consumption efficiency in both the local province and its neighboring provinces. Moreover, the absolute values of the indirect effects are much larger than that of the direct effects. On average, if the natural logs of per capita GDP in each province increased 1%, water consumption per ten thousand yuan GDP in the local province and its neighboring provinces would decrease 692.34% and 2701.77%, and that of the whole country would decrease 3394.11%. If the natural logs of urbanization level in each province increased 1%, water consumption per ten thousand yuan GDP in the local province and its neighboring provinces would decrease 192.87% and 1513.65%, and that of the whole country would decrease 1706.52%. It indicates that the spatial spillover effects of these two indicators are primary influencing factors for the improvement of China’s water consumption efficiency.
However, the direct and indirect effects of the other eight indicators are much different and complicated. The direct effect of per capita water resources is significantly positive while the indirect effect is insignificant. The direct effects of utilization ratio of water resources and per capita grain yield are significantly positive while the indirect effects are significantly negative. The direct effect of ratio of value added of the tertiary industry in GDP is insignificant while the indirect effect is significantly negative. The direct and indirect effects of per capita total investment in fixed assets and urban per capita disposable income are all insignificant. The direct effect of per capita social retail sales of consumer goods is significantly negative while the indirect effect is insignificant. The direct effect of rural per capita net income is insignificant while the indirect effect is significantly positive. Among these direct and indirect effects which are significant, the absolute values are all small. We can call these indicators as the secondary influencing factors.

4 Conclusions and implications

This study analyzes the influencing factors of water consumption efficiency in urbanizing China by spatial econometric models from the aspects of water resources endowment, water resources development level, economic development level, urbanization level, industrial structure optimization level, investment level, social consumption level, urban and rural income level and agricultural development level. We may get the following conclusions and policy implications:
(1) The spatial autocorrelation characteristics of water consumption efficiency among different provinces in China are significant. Therefore, the general panel data regression model which previous studies often used may be improper to reveal its influencing factors because it did not consider the spatial spillover effects of water consumption efficiency, and the results may be confusing and misleading. However, spatial econometric model may extract the significant factors more accurately because it considers the complicated spatial interaction effects. According to the results of LM test, Hausman test and Wald test, the spatial Dubin model with time random and spatial fixed effects can provide the optimal estimation on influencing factors of water consumption efficiency in urbanizing China.
(2) China’s water consumption efficiency has obvious spatial spillover effect. In other words, water consumption efficiency in each province is significantly influenced by water consumption efficiency of its neighboring provinces. On the other hand, water consumption efficiency in each province may significantly influence the water consumption efficiency of its neighboring provinces. For the macro average case of the 31 provinces in China, if water consumption efficiency in neighboring provinces increased 1%, water consumption efficiency of the local province would increase 0.34%. Therefore, to control the “red line” of water consumption efficiency, each province should properly consider potential influence caused by its neighboring provinces, and a unified water resources management system among different provinces should be further strengthened.
(3) China’s water consumption efficiency in local province is significantly influenced by some specific socio-economic and eco-environmental indicators of itself and its neighboring provinces. Among the ten specific indicators we selected, per capita GDP and urbanization level of itself and its neighboring provinces have the most prominent positive effects on water consumption efficiency, and the indirect effects of neighboring provinces are much larger. Therefore, the spatial spillover effects of the economic development level and urbanization level are the primary influencing factors for improving China’s water consumption efficiency. Moreover, though some of the direct or indirect effects of the other eight indicators are significant, the absolute values are all small. These indicators are the secondary influencing factors. On the whole, beyond water itself, a coordinated socio-economic development strategy among different provinces should also be strengthened. Especially, enhancing the economic cooperation and urbanization interaction with neighboring provinces may help to improve water consumption efficiency in each province.
However, though we have detected the positive/negative and direct/indirect effects of the influencing factors on water consumption efficiency in urbanizing China, we cannot reveal the essential reason behind it at present. The specific and practical interaction mechanism still remains to be studied in the future. Meanwhile, we just showed an application of spatial econometric model, and did not consider some technical matters such as elimination of the multicollinearity and endogeneity. Besides, due to the availability of data, some important indicators may be missing, such as agricultural modernization level, informatization level and education level. If data is available in the future, the influencing factors could be identified more accurately.

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Bao C, Chen X J, 2015. The driving effects of urbanization on economic growth and water use change in China: A provincial-level analysis in 1997-2011.Journal of Geographical Sciences, 25(5): 530-544.As one of the key issues in China’s sustainable development, rapid urbanization and continuous economic growth are accompanied by a steady increase of water consumption and a severe urban water crisis. A better understanding of the relationship among urbanization, economic growth and water use change is necessary for Chinese decision makers at various levels to address the positive and negative effects of urbanization. Thus, we established a complete decomposition model to quantify the driving effects of urbanization on economic growth and water use change for China and its 31 provincial administrative regions during the period of 1997–2011. The results show that, (1) China’s urbanization only contributed about 30% of the economic growth. Therefore, such idea as urbanization is the major driving force of economic growth may be weakened. (2) China’s urbanization increased 2352×10 8 m 3 of water use by increasing the economic aggregate. However, it decreased 4530×10 8 m 3 of water use by optimizing the industrial structure and improving the water use efficiency. Therefore, such idea as urbanization is the major driving force of water demand growth may be reacquainted. (3) Urbanization usually made greater contribution to economic and water use growth in the provincial administrative regions in east and central China, which had larger population and economic aggregate and stepped into the accelerating period of urbanization. However, it also made greater contribution to industrial structure optimization and water use efficiency improvement, and then largely decreased total water use. In total, urbanization had negative effects on water use growth in most provincial administrative regions in China, and the spatiotemporal differences among them were lessened on the whole. (4) Though urbanization helps to decrease water use for China and most provincial administrative regions, it may cause water crisis in urban built-up areas or urban agglomerations. Therefore, China should construct the water transfer and compensation mechanisms between urban and rural areas, or low and high density urban areas as soon as possible.

DOI

[2]
Bao C, Fang C L, 2007. Water resources constraint force on urbanization in water deficient regions: A case study of the Hexi Corridor, arid area of NW China.Ecological Economics, 62(3/4): 508-517.No abstract is available for this item.

DOI

[3]
Bao C, Fang C L, 2012. Water resources flows related to urbanization in China: Challenges and perspectives for water management and urban development.Water Resources Management, 26(2): 531-552.China has been experiencing rapid urbanization since the reform and open policy launched in 1978, leading to the growth of urban water demands and aggravating water scarcity especially in the new millennium. Accordingly, water resources previously used for agriculture and environmental systems tend to be transferred to urban systems. Limited by the total quantity and frail environments, the patterns of water resources flows among different sectors and regions change obviously. Water related problems induced by rapid urbanization have become one of the key concerns for scientists and governments. The purpose of this paper is to analyze the new features of water resources flows related to urbanization in China, mainly with regard to bidirectional water resources flows between rural and urban areas, environmental and socio-economic systems, real and virtual water flows between the south and north. This paper also considers the socio-economic and environmental challenges which are resulted from water resources flows in such a case, and provides some countermeasures on how to promote water resources to flow healthily and swimmingly, so as to improve the urban development constrained by scarce water resources.

DOI

[4]
Bao C, He D, 2015. The causal relationship between urbanization, economic growth and water use change in provincial China.Sustainability, 7(12): 16076-16085.The relationship between urbanization, economic growth, and water use change is one of the key issues for China’s sustainable development, as rapid urbanization and continuous economic growth are accompanied by a steady water stress. Thus, we applied a cointegration test and a VECM (vector error correction model) Granger causality test to investigate the causal relationship between the urbanization level, the economic development level, and the total water use in China and its 31 provincial administrative regions during 1997–2013. Results show that the three indicators have a long-run equilibrium relationship in most provincial administrative regions in China. However, the short-run effects and Granger causal relationship are insignificant for China and most provincial administrative regions. Therefore, that an idea such as urbanization as the engine or major driving force of economic growth, and that China’s urbanization and economic growth will bring a water crisis and will be strongly constrained by water resources, might be properly weakened. Targeted and relatively separate policies should be emphasized more for the coordinated development of China’s urbanization, economy, and water resources.

DOI

[5]
Burridge P, 1981. Testing for a common factor in a spatial autoregression model.Environment and Planning A, 13(7): 795-800.Three asymptotically equivalent tests for the presence of a common factor in a spatial autoregression model are presented. When such a common factor exists, the model reduces to the simple regression with spatially correlated disturbances. The tests can thus be used to examine the appropriateness of the latter specification. The procedure is illustrated by application to a model of Irish agricultural consumption discussed by O'Sullivan (1968), and Cliff and Ord (1973).

DOI

[6]
Cai X M, 2008. Water stress, water transfer and social equity in northern China: Implications for policy reforms.Journal of Environmental Management, 87(1): 14-25.Water stress in Northern China is characterized with major, inefficient irrigation water use and rapidly growing non-agricultural water demands, as well as limited water quantity and declining water quality. Water use in the region is undergoing transfer from agricultural to municipal and industrial sectors. Currently, part of the economic loss and environmental damage due to water stress can be considered as a consequence of water transfer failures, including the current transfers, which hurt farmers' livelihood and income, and the needed transfers, which industry and cities have been waiting for but have not received. This paper starts with a discussion of the causes of water stress in Northern China, which is fundamental to understand the necessity and complexity of agricultural water transfers. Following that, it reviews water transfers in Northern China as a cause for concern over the social stability, economy and environment of the region. Based on an integrated analysis of economic, environmental, fiscal and social implications, this paper begins by identifying critical barriers to smooth water redistribution; and ends with implications for policy reforms, ensuring that farmers can and will save water. It is concluded that the decisions of water reallocation under water stress should be shared by communities at all levels, from the local to the national, to ensure equal access of water, especially the availability of the basic water need for all groups.

DOI PMID

[7]
Deng G, Li L, Song Yet al., 2016. Provincial water use efficiency measurement and factor analysis in China: Based on SBM-DEA model.Ecological Indicators, 69: 12-18.This paper estimates water use efficiency of 31 provinces in China during 2004-2013 using slack based measure-data envelopment analysis (SBM-DEA) model which takes consideration of sewage. By using panel data model, factors that influence water use efficiency are explored. The results show that: (1) water use efficiency is higher in economically developed provinces such as Beijing, Shanghai and Tianjin. (2) In general, inefficiencies of labor input and water input are larger than that of capital input. (3) After taking into account sewage as unexpected output, inefficiency of gross domestic product (GDP) for each province is quite low, but most provinces have inefficient sewage emission. (4) It is found in the analysis of influencing factors that the ratio of added value of agricultural sector, water usage per capita and sewage per unit of output have negative impact on water resources use efficiency, while the import dependence and export dependence have positive impact. The results are robust.

DOI

[8]
Dickey D A, Fuller W A, 1979. Distribution of the estimators for autoregressive time series with a unit root.Journal of the American Statistical Association, 74(366a): 427-431.Let n observations Y1, Y2, ..., Yn be generated by the model Yt = ρ Yt - 1 + et, where Y0 is a fixed constant and {et}t = 1n is a sequence of independent normal random variables with mean 0 and variance σ2. Properties of the regression estimator of ρ are obtained under the assumption that ρ = ± 1. Representations for the limit distributions of the estimator of ρ and of the regression t test are derived. The estimator of ρ and the regression t test furnish methods of testing the hypothesis that ρ = 1.

DOI

[9]
Elhorst J P, 2003. Specification and estimation of spatial panel data models.International Regional Science Review, 26(3): 244-268.

DOI

[10]
Elhorst J P, 2010. Spatial Panel Data Models. Berlin: Springer.

[11]
Elhorst J P, 2014. Matlab software for spatial panels.International Regional Science Review, 37(3): 389-405.

DOI

[12]
Fang C Let al., 2009. Report on China’s Urbanization and the Guarantee of Resources and Eco-environment. Beijing: Science Press. (in Chinese)

[13]
Fang C L, Bao C, Huang J C, 2007. Management implications to water resources constraint force on socio-economic system in rapid urbanization: A case study of Hexi Corridor, NW China.Water Resources Management, 21(9): 1613-1633.As water has become the shortest resources in arid, semi-arid and rapid urbanization areas when the water resources utilization has approached or exceeded its threshold, water resources system slows down the socio-economic growth rate and destroys the projected targets to eradicate poverty and realize sustainable development. We put forward the concept of Water Resources Constraint Force (WRCF) and constructed a conceptual framework on it. Conceptual models on the interactions and feedbacks between water resources and socio-economic systems in water scarce regions or river basins indicate that, if the socio-economic system always aims at sustainable development, WRCF will vary with a normal distribution curve. Rational water resources management plays an important role on this optimistic variation law. Specifically, Water Demand Management (WDM) and Integrated Water Resources Management (IWRM) are considered as an important perspective and approach to alleviate WRCF. A case study in the Hexi Corridor of NW China indicates that, water resources management has great impact on WRCF both in Zhangye and Wuwei Region, and also the river basins where they are located. The drastic transformation of water resources management pattern and the experimental project - Building Water-saving Society in Zhangye Region alleviated the WRCF to some extent. However, from a water resources management view, WRCF in Zhangye Region still belongs to the severe constraint type. It will soon step into the very severe constraint type. In order to shorten the periods from the very severe constraint type finally to the slight constraint type, WDM and IWRM in the Hei River Basin should be improved as soon as possible. However, in the Shiyang River Basin, WRCF belongs to the very severe constraint type at present due to poor water resources management in the past. Though the socio-economic system adapted itself and alleviated the WRCF to some extent, the Shiyang River Basin had to transform the water supply management pattern to WDM, and seek IWRM in recent years. It is concluded that WDM and IWRM is a natural selection to alleviate the WRCF on the socio-economic system and realize sustainable development.

DOI

[14]
Halleck V S, Elhorst J P, 2015. The SLX model.Journal of Regional Science, 55(3): 339-363.

DOI

[15]
Huang G W, 2015. From water-constrained to water-driven sustainable development: A case of water policy impact evaluation.Sustainability, 7(7): 8950-8964.A water allocation policy that aimed to balance water demand with water availability to ensure sustainability was implemented in an arid region of China over ten years ago. This policy-success was assessed across three dimensions: society, the environment, and the economy. While the assessment was not intended to be comprehensive, it highlighted the best outcomes of the policy intervention while revealing some hidden issues. It was found that although the policy was successful in placing a ceiling on water use in the middle reaches of the Heihe River, the Water User Association, one of the main actors in water policy implementation, was under-recognized, even though it functioned well. Moreover, the economic structural adjustment at the macro level had not led to any significant reduction in water use, the reasons for which were explored.

DOI

[16]
Jiang Y, 2009. China’s water scarcity.Journal of Environmental Management, 90(11): 3185-3196.

[17]
Johansen S, Juselius K, 1990. Maximum likelihood estimation and inference on cointegration: With application to the demand for money.Oxford Bulletin of Economics and Statistics, 52(2): 169-210.

[18]
Kaneko S, Tanaka K, Toyota Tet al., 2004. Water efficiency of agricultural production in China: Regional comparison from 1999 to 2002. International Journal of Agricultural Resources, Governance and Ecology, (3/4): 231-251.The mode of action of R-91650, an arylpiperazinyl fluoroquinolone, on feline immunodeficiency virus (FIV) replication inhibitory activity was investigated. R-91650 inhibited replication of FIV at non-cytotoxic concentration levels in both acutely infected peripheral blood mononuclear cells and chronically infected P-CrFK cells. The compound reduced the intracellular p24 concentration levels in P-CrFK cells in a dose-dependent manner. Northern blot analysis revealed that R-91650 selectively prevented the accumulation of FIV mRNA in P-CrFK cells. However, the compound did not inhibit FIV-long terminal repeat (LTR) promoter activity in the reporter gene expression analysis. These data suggest that R-91650 is a novel inhibitor of FIV replication that inhibits a certain step or steps following transcription initiation of the FIV-LTR promoter.

DOI PMID

[19]
Lee L F, Yu J, 2010. Some recent developments in spatial panel data models.Regional Science & Urban Economics, 40(5): 255-271.Spatial econometrics has been an ongoing research field. Recently, it has been extended to panel data settings. Spatial panel data models can allow cross sectional dependence as well as state dependence, and can also enable researchers to control for unknown heterogeneity. This paper reports some recent developments in econometric specification and estimation of spatial panel data models. We develop a general framework and specialize it to investigate different spatial and time dynamics. Monte Carlo studies are provided to investigate finite sample properties of estimates and possible consequences of misspecifications. Two applications illustrate the relevance of spatial panel data models for empirical studies.

DOI

[20]
Lesage J P, Pace R K, 2009. Introduction to Spatial Econometrics. Boca Raton, FL: CRC Press Taylor & Francis Group.

[21]
Li J, Ma X, 2015. Econometric analysis of industrial water use efficiency in China.Environment, Development and Sustainability, 17(5): 1209-1226.Low industrial water use efficiency has become a resource bottleneck to industrial development in China. The SBM-undesirable and meta-frontier models were used in combination with empirical data in 30 provinces in mainland China (Tibet excluded due to data missing from 1999 to 2013), to compare industrial water use efficiency in mainland China under meta-frontier and group-frontier, and explore the influencing factors. The empirical results of the study reveal that: (a) there is a large difference in the industrial water use efficiency between meta-frontier and group-frontier in mainland China, due to the heterogeneity in the levels of industrial water use technology; (b) given the low recycle rate of polluted industrial water, there is room for improvement in the industrial water use efficiency in the 30 provinces in mainland China. Further, the study finds that the current price of industrial water is distorted to some extent, failing to coordinate with the use of water resources. Policy implications indicate that industrial water use efficiency is not only related to technological heterogeneity in different regions, but also the control and treatment of industrial water pollution. Therefore, the current price of industrial water should be gradually raised. A scalar water pricing system as residential water could also be applied to industrial water.

DOI

[22]
Li S S, Ma Y, 2014. Urbanization, economic development and environmental change.Sustainability, 6(8): 5143-5161.This paper applies the pressure-state-response (PSR) model to establish environmental quality indices for 30 administrative regions in China from 2003 to 2011 and employs panel data analysis to study the relationships among the urbanization rate, economic development and environmental change. The results reveal a remarkable inverted-U-shaped relationship between the urbanization rate and changes in regional environmental quality; the “turning point” generally appears near an urbanization rate of 60%. In addition, the degree and mode of economic development have significant, but anisotropic effects on the regional environment. Generally, at a higher degree of economic development, the environment will tend to improve, but an extensive economic growth program that simply aims to increase GDP has a clear negative impact on the environment. Overall, the results of this paper not only further confirm the “environmental Kuznets curve hypothesis”, but also expand it in a manner. The analysis in this paper implies that the inverted-U-shaped evolving relationship between environmental quality and economic growth (urbanization) is universally applicable.

DOI

[23]
Lin X, Wang D, 2016. Spatiotemporal evolution of urban air quality and socioeconomic driving forces in China.Journal of Geographical Sciences, 26(11): 1533-1549.Air pollution is a serious problem brought by the rapid urbanization and economic development in China, imposing great challenges and threats to population health and the sustainability of the society. Based on the real-time air quality monitoring data obtained for each Chinese city from 2013 to 2014, the spatiotemporal characteristics of air pollution are analyzed using various exploratory spatial data analysis tools. With spatial econometric models, this paper further quantifies the influences of socioeconomic factors on air quality at both the national and regional scales. The results are as follows: (1) From 2013 to 2014, the percentage of days compliance of urban air quality increased but air pollution deteriorated and the worsening situation in regions with poor air quality became more obvious. (2) Changes of air quality show a clear temporal coupling with regional socioeconomic activities, basically “relatively poor at daytime and relatively good at night”. (3) Urban air pollution shows a spatial pattern of “heavy in the east and light in the west, and heavy in the north and light in the south”. (4) The overall extent and distribution of regional urban air pollution have clearly different characteristics. The formation and evolution of regional air pollution can be basically induced as “the pollution of key cities is aggravated—pollution of those cities spreads—regional overall pollution is aggravated—the key cities lead in pollution governance—regional pollution joint prevention and control is implemented—regional overall pollution is reduced”. (5) At the national level, energy consumption, industrialization and technological progress are the major factors in the worsening of urban air quality, economic development is a significant driver for the improvement of that quality. (6) Influenced by resources, environment and the development stage, the socioeconomic factors had strongly variable impacts on air quality, in both direction and intensity in different regions. Based on the conclusion, the regional differentiation and development idea of the relationship between economic development and environmental changes in China are discussed.

DOI

[24]
Ma H, Shi C, Chou N, 2016. China’s water utilization efficiency: An analysis with environmental considerations.Sustainability, 8(6): 516-530.

DOI

[25]
Moran P A P, 1950. Notes on continuous stochastic phenomena.Biometrika, 37(1/2): 17-23.Biometrika. 1950 Jun;37(1-2):17-23.

DOI PMID

[26]
Shen J F, 2006. Estimating urbanization levels in Chinese provinces in 1982-2000.International Statistical Review, 74(1): 89-108.Abstract Summary No consistent and reliable annual data series on the urbanization level for provincial regions of China is available. Making use of urban population data from the 1982 and 2000 population censuses, this paper estimates an annual data series of the urbanization level for provincial regions using an estimation approach developed on the basis of a conceptual model of dual-track urbanization. Based on such estimated new urban data of provincial regions, the major trends of urbanization in Chinese provinces and the relationship between urbanization and economic development are analysed for the period 1982–2000. Résumé On ne dispose pas de séries de données annuelles cohérentes et fiables sur le niveau d'urbanisation des régions de province chinoises. En utilisant les données de population urbaine des recensements de 1982 et 2000, cet article estime une série annuelle du niveau d'urbanisation des régions de province gr09ce à une approche développée sur la base d'un modèle conceptuel d'urbanisation différenciée. Sur la base de nouvelles données urbaines ainsi estimées pour les régions provinciales, on analyse les grandes tendances d'urbanisation dans les provinces chinoises et la relation entre l'urbanisation et le développement économique pour la période 1982–2000.

DOI

[27]
Sun C, Zhao L, Zou Wet al., 2014. Water resource utilization efficiency and spatial spillover effects in China.Journal of Geographical Sciences, 24(5): 771-788.

DOI

[28]
Wang Y, Bian Y, Xu Het al., 2015. Water use efficiency and related pollutants' abatement costs of regional industrial systems in China: A slacks-based measure approach.Journal of Cleaner Production, 101: 301-310.Water saving and wastewater cleaning are two important ways to alleviate water shortage crisis in China. This paper aims to examine water use efficiency and abatement costs of two main pollutants in wastewater, i.e., chemical oxygen demand (COD) and ammonia nitrogen (NH 4 –N), of regional industrial systems in China. For this purpose, a slacks-based measure model based on the assumption of undesirable output weak disposability is developed. Based on the proposed model, the measures of water saving potential, COD reduction potential, NH 4 –N reduction potential, COD abatement cost and NH 4 –N abatement cost are presented. We apply the proposed approaches to investigate water use efficiency, water saving potential, the two pollutants' reduction potentials and their related abatement costs during 2009–2010. Results of the application show that there are great potentials to reduce water consumption and pollutants' discharges in China's regional industrial production process, and there exist evident geographic disparities in water saving potentials, pollutants' reduction potentials and their abatement costs. It is found that water use efficiency increases from 2009 to 2010, while pollutants abatement costs decrease. Some other findings and implications for efficiency improvement of China's regional industrial systems in water management and pollutants' abatement are achieved.

DOI

[29]
Zang Z, Zou X, Xi Xet al.., 2016. Quantitative characterization and comprehensive evaluation of regional water resources using the Three Red Lines method.Journal of Geographical Sciences, 26(4): 397-414.基于新工业化，快速的都市化和农业现代化( IUAM )的合作发展，并且从在水之间的交互关系的观点资源和地区性的人口，eco环境，经济和协会，水资源紧张( WRI )的概念，水环境紧张(魏)，水资源亲戚效率( WRRE )和水环境亲戚效率()与精力紧张，资源效率和 env 的参考被定义根据 benchmarking 理论，三红的量的描述和评估方法排队(水资源分配的上面的限制，水资源的利用效率和污水分泌物的上面的限制的基线) 被建议。根据这些概念和模型，三红的实验分析在在 2003 和 2012 之间的中国大陆上水资源排队被执行。结果证明在中国的东方、中央、西方的部分的那全部的水消费拥有俱乐部集中特征，它意味着这些区域把类似的内部条件显得在发展的集中。在水消费的内部省的差别在东方、中央的区域继续到减少，而是 northsouth 区别特征仍然是相对明显的，当在东方部分的省的差别是时至少并且有的中央区域最大。西南河和 Huaihe 河盆里的所有四个部门的水资源效率(WRE ) 通常高。Songhua 河，长江和珀尔河盆， Songhua 河里的农业 WRRE，黄河和西北的河盆里的工业 WRRE 和 Liaohe 河，长江和珀尔河盆里的国内 WRRE 都是低的。东南的河和长江盆里的 Eco 环境的 WRRE 是低的，但是显示出一个向上的趋势。除了西北的河盆，另外的河盆有高 eco 环境的 WRRE 与一向下趋势。西方的中国，特别西北的部分，有水环境( WERI )和高综合的水环境管理( IWEM )的低相对紧张性能，而是水资源( WRRI )的相对紧张相当高，并且在这些区域的综合的水资源管理( IWRM )的全面表演是低的。在南部的中国，特别东南的部分， IWEM 相当高，但是全面 IWRM 更低。

DOI

[30]
Zhou Y X, Yu H B, 2002. Suggestions on reconstructing the urbanization levels in China based on the fifth population census.Statistical Research, (4): 44-47. (in Chinese)

Options
Outlines

/

〈 〉