The equitable allocation of water governance responsibilities is very important yet difficult to achieve, particularly for a basin which involves many stakeholders and policymakers. In this study, the environmental Gini coefficient model was applied to evaluate the inequality of water governance responsibility allocation, and an environmental Gini coefficient optimisation model was built to achieve an optimal adjustment. To illustrate the application of the environmental Gini coefficient, the heavily polluted transboundary Taihu Lake Basin in China, was chosen as a case study. The results show that the original environmental Gini coefficient of the chemical oxygen demand (COD) was greater than 0.2, indicating that the allocation of water governance responsibilities in Taihu Lake Basin was unequal. Of seven decision-making units, three were found to be inequality factors and were adjusted to reduce the water pollutant emissions and to increase the water governance inputs. After the adjustment, the environmental Gini coefficient of the COD was less than 0.2 and the reduction rate was 27.63%. The adjustment process provides clear guidance for policymakers to develop appropriate policies and improve the equality of water governance responsibility allocation.

INTRODUCTION

Water is a public liquid resource, and water governance responsibilities must be allocated among many stakeholders. However, equitably allocating water governance responsibilities is a challenge. Currently, in both developed and developing countries, no appropriate method or standard exists for water governance responsibility allocation.

The Gini coefficient is a commonly used economic measurement for the inequality of income or wealth distribution (Bosi & Seegmuller 2006). In recent years, scholars have attempted to use the Gini coefficient in the governance of liquid resources or pollutants. Heil & Wodon (2000) used the Gini coefficient to analyse future carbon emission inequality, within and between countries and named the method the ‘environmental Gini coefficient’. Millimet & Slottje (2002) used the environmental Gini coefficient to measure inequality in the distribution of per capita emissions across US counties and states. Vass et al. (2013) evaluated the fairness of EU carbon emission policies by calculating the Gini coefficients for six criteria of policy outcomes. Cho & Lee (2014) introduced the environmental resource-based Gini coefficient into the waste load allocation (WLA) model to compute the inequality in waste load discharge with respect to the environmental resources in each region. The suitability of the WLA model was verified by its application to the total maximum daily load of a heavily polluted river in South Korea. In China, the environmental Gini coefficient is widely used for the allocation of regional water pollutant emissions (Sun et al. 2010; Zhang et al. 2010; Wang et al. 2011; Zhang et al. 2012a, b; Chen et al. 2012) and for the inequality analysis of urban water use (Zhang & Shao 2010). To build the environmental Gini coefficient model, these studies generally use the cumulative proportion of various water pollutant emissions as the vertical axis and the cumulative proportion of the GDP or ecological capacity as the horizontal axis to establish the environmental Lorenz curve. Then, the environmental Gini coefficient can be computed. In addition, the inequality factors are based on the efficiency coefficient (Sun et al. 2010, etc.).

As described above, the Gini coefficient (or the environmental Gini coefficient) is increasingly used in the inequality assessment of carbon emissions and water WLA. The environmental Gini coefficient is seldom used in governance responsibility allocation which is very important for environmental governance but is challenging for policymakers.

To improve the equality of water governance, the environmental Gini coefficient was applied in the allocation of water governance responsibility. This research first attempted to apply the environmental Gini coefficient to evaluate the inequality of water governance responsibility allocation. Then, an environmental Gini coefficient optimisation model was established to achieve the optimal adjustment. To verify the practicality of the method, Taihu Lake Basin, a heavily polluted basin in China, was chosen as a case study.

METHODS

Environmental Gini coefficient model

This research applied the environmental Gini coefficient to evaluate the inequality of the environmental governance responsibility allocation of water. The method consists of the following three steps: (1) to establish an environmental Lorenz curve; (2) to calculate the environmental Gini coefficient; and (3) to determine the criteria of the overall inequality evaluation.

Establishment of the environmental Lorenz curve

We assumed that water responsibilities were allocated to n decision-making units (DMUs). The term ‘DMUs’ first proposed by Chames et al. (1978) represented the units whose efficiency were assessed based on the same inputs or outputs, such as different countries, companies, departments, etc (Chames et al. 1978, 1985; He & Lyu 2008). In this paper, DMUs were used to represent different water governance units within a region or basin, such as the n cities within one state or the n pollution control districts within one basin. The environmental Lorenz curve was established by using the n DMUs' water pollution contribution and water governance contribution (see Figure 1). As shown in Figure 1, the horizontal axis represents the cumulative proportion of the water pollution contribution, while the vertical axis represents the cumulative proportion of the water governance contribution.

Figure 1

Environmental Lorenz curve.

Figure 1

Environmental Lorenz curve.

The pollution contribution coefficient was calculated from the weighted sum of m1 pollution contribution indicators. The formulae are as follows: 
formula
1
 
formula
2
where CPi represents the pollution contribution coefficient of the ith DMU, CPij represents the pollution contribution coefficient of the jth pollution contribution indicator in the ith DMU, wj represents the weight of the jth pollution contribution indicator, Pij represents the original value of the jth pollution contribution indicator in the ith DMU, and n represents the number of DMUs.
The governance contribution coefficient was calculated from the weighted sum of m2 governance contribution indicators. The formulae are as follows: 
formula
3
 
formula
4
where CGi represents the governance contribution coefficient of the ith DMU, CGik represents the governance contribution coefficient of the kth governance contribution indicator in the ith DMU, wk represents the weight of the kth governance contribution indicator, Gik represents the original value of the kth governance contribution indicator in the ith DMU, and n represents the number of DMUs.

Calculation of the environmental Gini coefficient

Based on the definition of the Gini coefficient (Sun et al. 2010), the environmental Gini coefficient is expressed as 
formula
5
where SA represents the area between the practical distribution curve and the absolutely equitable distribution curve, and SB represents the area below the practical distribution curve, as shown in Figure 1.
The calculation of the Gini coefficient can be completed using a variety of methods, such as the geometric method, Gini's mean difference method (or relative mean difference method), the covariance method or matrix methods (e.g., Sadras & Bongiovanni 2004; White 2007; and Groves-Kirkby et al. 2009). In this study, we used a trapezoidal approximation algorithm, which is expressed as follows: 
formula
6
where xi represents the cumulative proportion of the water pollution contribution, and yi represents the cumulative proportion of the water governance contribution. When i = 1, (xi−1, yi−1) = (0, 0).

Determination of evaluation criteria

The Gini coefficient has values within the range of 0 (perfectly uniform distribution) to 1 (complete inequality) (Groves-Kirkby et al. 2009). Many scholars have noted that a reasonable range for the environmental Gini coefficient is 0–0.2 (e.g., Xie et al. 2006; Qin et al. 2013). Therefore, we set a value of ≤0.2 as the expected range for the environmental Gini coefficient and as the absolute equality criteria for the governance responsibility allocation.

Environmental Gini coefficient optimisation model

To achieve the optimal adjustment for the inequality assessment results, an environmental Gini coefficient optimisation model was established. The basic concept of this model was to achieve the optimal adjustment target by establishing and following particular adjustment rules. The rules mainly include the adjustment criterion, the adjustment paths, and the upper and lower adjustment limits.

The adjustment criterion: what should be adjusted

To establish the optimisation model, the first step was to set an adjustment criterion to determine what should be adjusted or what the inequality factors are. The ratio of the water governance contribution coefficient to the water pollution contribution coefficient was proposed as the criterion for determining the inequality factors.

The ratio of the water governance contribution coefficient to the water pollution contribution coefficient of each DMU can be expressed as 
formula
7
where CGi and CPi represent the water governance contribution coefficient of the ith DMU and the water pollution contribution coefficient of the ith DMU, respectively.

If CPGi < 1, then the water governance contribution coefficient of the ith DMU is less than the pollution contribution coefficient; thus, the ith DMU is an inequality factor that should be adjusted.

The adjustment paths: how the inequality factors should be adjusted

To establish the optimisation model, the second step was to determine how the inequality factors should be adjusted or how to set the adjustment paths. Following the adjustment guideline of ‘less water pollution and more governance inputs’, the adjustment paths were limited to two types: (1) reducing the amount of the pollution contribution; and (2) increasing the amount of the governance contribution. The path that should initially be selected for the DMU adjustment depends on the ratio of the industrial production contribution coefficient to the water pollution contribution coefficient. If the ratio of the industrial production contribution coefficient to the water pollution contribution coefficient is less than 1, then the pollution contribution should first be adjusted; otherwise, the governance contribution should first be adjusted.

The upper and lower adjustment limits: when should the adjustment stop

To establish the optimisation model, the third step was to determine when the adjustment (including the partial and the overall adjustments) should stop. The upper and lower limits for the water pollution contribution reduction rate and the water governance contribution increase rate were set as the partial constraints. A GE value of 0.2 or less was set as the overall constraint. For example, in the adjustment process, if the water pollution contribution reduction rate of one DMU exceeds the constraint, then the reduction of the water pollution contribution will stop and return to the previous step. If the value of GE is still greater than 0.2, then the adjustment path will change. When the value of GE is no more than 0.2, the overall adjustment stops.

The overall constraint is 
formula
8
The partial constraints are 
formula
9
 
formula
10
where LPi and UPi represent the lower and upper limits of the water pollutant emissions reduction rate, respectively. LGi and UGi represent the lower and upper limits of the water governance inputs increase rate, respectively. CPi0 and CPi represent the water pollution contributions before and after the adjustment, respectively. CGi0 and CGi represent the water governance contributions before and after the adjustment, respectively. TP0 and TP represent the total amount of water pollutant emissions before and after the adjustment, respectively. TG0 and TG represent the total amount of water governance inputs before and after the adjustment, respectively.

The upper and lower adjustment limits were determined by comprehensively considering the actual situation of the DMU and the feasibility of the adjustment programmes. The optimisation model presented the result of each adjustment step. Specifically, the final adjustment target was achieved using several steps within a particular period. Therefore, the regional water resource management plan was developed and achieved according to the results of each adjustment step.

The adjustment process was completed by developing a secondary algorithm using the MATLAB platform. The basic concept was to set an adjustment step length for the pollutant emission reduction rate and governance input increase rate; the loop calculation did not stop under the constraints until the environmental Gini coefficient was no greater than 0.2.

CASE STUDY

The described method was applied for Taihu Lake Basin, China. Since the reform and opening up policy (1978), industrialisation has rapidly advanced in the Taihu Lake Basin. Meanwhile, industrial pollution has become the most serious problem for water governance in the transboundary basin.

Study area

Taihu Lake Basin is located in Southeast China and is the core area of the Yangtze River Delta (see Figure 2). The basin is the most industrialised area in China, with a high population density and high urbanisation. Dense river networks and numerous lakes constitute the basin. Taihu Lake, the water body of the basin, is the third largest freshwater lake in China. Before the 1980s, the water of Taihu Lake was chilly and the entire area was characterised by beautiful scenery. In the 1990s, rural industries were developed in Sunan (southern Jiangsu Province), Shanghai's economy grew rapidly and the private economy developed extremely quickly in Zhejiang (Jiangsu, Shanghai and Zhejiang are the main locations around Taihu Lake, as shown in Figure 2). Taihu Lake Basin became the most desirable area of the Yangtze River Delta. However, because of the rapid industrialisation and urbanisation, the water quality of the Taihu Lake Basin has been severely compromised. In 2007, a large bloom of blue-green algae in Taihu Lake caused the water quality to deteriorate, which seriously threatened the water supply for residents near the lake. The incident attracted substantial attention from the Chinese Government. To improve the water quality of Taihu Lake and to ensure safe drinking water, ‘the integrated water resources governance planning for Taihu Lake Basin’ was proposed, and a substantial amount of money has been invested in the programme. Currently, the water quality in the basin has improved, but it is not yet satisfactory.

Figure 2

Geographical location and administrative boundary of the Taihu Lake Basin.

Figure 2

Geographical location and administrative boundary of the Taihu Lake Basin.

According to ‘the integrated water resources governance planning for Taihu Lake Basin’, the control areas in Taihu Lake Basin include the following: Suzhou, Wuxi, Changzhou, and Zhenjiang (four cities in Jiangsu Province); Huzhou, Jiaxing, and Hangzhou (three cities in Zhejiang Province); and Qingpu, Jinze, and Zhujiajiao (three towns in Shanghai). Considering data availability and the comparability of DMUs, the seven DMUs of Suzhou, Wuxi, Changzhou, Zhenjiang, Hangzhou, Jiaxing, and Huzhou were chosen as the study areas to explore the allocation equality of water resource governance responsibility in the Taihu Lake Basin.

Environmental Gini coefficient model application

Selection of the pollution contribution indicator

The industrial wastewater chemical oxygen demand (COD) discharge of different industries was chosen as the water pollution contribution indicator and as the water pollution control index.

The industrial wastewater COD discharge is not the only water pollution contribution indicator for this method. Many other water pollution contribution indicators can be used, such as the biochemical oxygen demand (BOD), pH value, ammonia nitrogen (NH4 + –N), total nitrogen (TN), total phosphorus (TP), total number of bacteria, and so on. For example, Cho & Lee (2014) used the BOD as the water pollution contribution indicator, while Wang et al. (2011) used the NH4 + –N. A single indicator or a combination of indicators can be used in this method. The indicator which should be used depends on the specific research and governance objectives of the specific circumstances. A variety of factors should be considered. For example, since COD was one of the main pollutants and had significantly exceeded its standard in recent years, Sun et al. (2010) and Zhang et al. (2012b) use it as the pollution contribution indicator when the local governments were very serious about COD discharge. In this study, we selected the industrial wastewater COD discharge for the following reasons: (1) industrial wastewater treatment in Taihu Lake Basin was our main focus; (2) COD is an important index in the analysis of industrial wastewater pollution; (3) the Chinese government attaches great importance to monitoring and controlling wastewater COD discharge; and (4) the representation, importance and correlation of the indexes were considered.

Currently, no uniform or distinct criteria exist for classifying pollution industries. Zhao (2003) used the equally weighted sum average method to classify the pollution industries using the data for various pollutant emissions (i.e., wastewater, waste gas and solid waste) per unit production from 1991 to 1999 in China. Jiang (2010), based on the method of Zhao (2003), adopted the pollutant emissions per unit production as a classification criteria. The current study applied the classification criteria proposed by Jiang (2010). A total of 18 polluting industries' wastewater COD discharges were selected as the pollution contribution indicators, as shown in Table 1.

Table 1

Environmental pollution contributions of different industries

Industry sector COD discharge coefficient COD weight coefficient 
Manufacturing of paper and paper products 48.512 0.380 
Manufacturing of raw chemical materials and chemical products 4.774 0.037 
Smelting and pressing of ferrous metals 1.086 0.009 
Mining and processing of non-metal ores 1.699 0.013 
Manufacturing of non-metallic mineral products 0.715 0.006 
Manufacturing of foods 7.061 0.055 
Manufacturing of liquor, beverages and refined tea 17.149 0.134 
Manufacturing of medicines 5.450 0.043 
Manufacturing of chemical fibres 21.704 0.170 
Smelting and pressing of non-ferrous metals 0.722 0.006 
Manufacturing of metal products 1.092 0.009 
Printing and reproduction of recording media 0.477 0.004 
Processing of petroleum, coking and processing of nuclear fuel 2.036 0.016 
Manufacturing of rubber and plastic products 0.541 0.004 
Manufacturing of leather, fur, feather and related products 5.472 0.043 
Manufacturing of electrical machinery and apparatus 0.139 0.001 
Manufacturing of computers, communication and other electronic equipment 0.472 0.004 
Manufacturing of textiles 8.602 0.067 
Industry sector COD discharge coefficient COD weight coefficient 
Manufacturing of paper and paper products 48.512 0.380 
Manufacturing of raw chemical materials and chemical products 4.774 0.037 
Smelting and pressing of ferrous metals 1.086 0.009 
Mining and processing of non-metal ores 1.699 0.013 
Manufacturing of non-metallic mineral products 0.715 0.006 
Manufacturing of foods 7.061 0.055 
Manufacturing of liquor, beverages and refined tea 17.149 0.134 
Manufacturing of medicines 5.450 0.043 
Manufacturing of chemical fibres 21.704 0.170 
Smelting and pressing of non-ferrous metals 0.722 0.006 
Manufacturing of metal products 1.092 0.009 
Printing and reproduction of recording media 0.477 0.004 
Processing of petroleum, coking and processing of nuclear fuel 2.036 0.016 
Manufacturing of rubber and plastic products 0.541 0.004 
Manufacturing of leather, fur, feather and related products 5.472 0.043 
Manufacturing of electrical machinery and apparatus 0.139 0.001 
Manufacturing of computers, communication and other electronic equipment 0.472 0.004 
Manufacturing of textiles 8.602 0.067 

Note: The units of the COD discharge coefficient are tons/108 yuan.

According to the COD discharges from each industry's wastewater discharged per unit industrial production in China in 2012, the COD discharge coefficient of each industry was calculated. The proportion of each industry's COD discharge coefficient, which accounts for the total COD emission coefficient, was calculated as the weight coefficient. The results are shown in Table 1.

Selection of the governance contribution indicator

Based on the water pollution control contribution, the ecological capacity contribution, and the data availability, six environmental governance contribution indicators of water were selected. The weight coefficients of each indicator were obtained using the Delphi method, as shown in Table 2.

Table 2

Environmental governance contribution indicators of water

Indicator Weight coefficient 
Water pollution control contribution Number of employed persons in the management of water conservancy and the environment 0.260 
Urban drainage network density 0.139 
Sewage treatment rate of sewage treatment plants 0.233 
Comprehensive utilisation rate of industrial solid wastes 0.102 
Ecological capacity contribution Landscape green area 0.130 
Green coverage of completely urban regions 0.136 
Indicator Weight coefficient 
Water pollution control contribution Number of employed persons in the management of water conservancy and the environment 0.260 
Urban drainage network density 0.139 
Sewage treatment rate of sewage treatment plants 0.233 
Comprehensive utilisation rate of industrial solid wastes 0.102 
Ecological capacity contribution Landscape green area 0.130 
Green coverage of completely urban regions 0.136 

Assessment results and discussion

The pollution contribution coefficient and governance contribution coefficient of each DMU were calculated using formulas (1)–(4). The results are shown in Table 3.

Table 3

Water pollution contribution coefficient and governance contribution coefficient

DMU Pollution contribution coefficient Governance contribution coefficient 
Hangzhou 0.316 0.197 
Suzhou 0.275 0.165 
Jiaxing 0.129 0.119 
Wuxi 0.100 0.176 
Zhenjiang 0.086 0.106 
Huzhou 0.048 0.117 
Changzhou 0.046 0.121 
Total 1.000 1.000 
DMU Pollution contribution coefficient Governance contribution coefficient 
Hangzhou 0.316 0.197 
Suzhou 0.275 0.165 
Jiaxing 0.129 0.119 
Wuxi 0.100 0.176 
Zhenjiang 0.086 0.106 
Huzhou 0.048 0.117 
Changzhou 0.046 0.121 
Total 1.000 1.000 

The environmental Gini coefficient of the COD was calculated using formula (6). The result shows that the environmental Gini coefficient of the COD was 0.275. Thus, the environmental Gini coefficient of the COD was not within the range of ‘absolute equality’ and should be adjusted.

The ratio of the governance contribution to the pollution contribution of each DMU was calculated using formula (7). The results showed that the ratios of the governance contribution to the pollution contribution of Hangzhou, Suzhou, Jiaxing, Wuxi, Zhenjiang, Huzhou, and Changzhou were 0.622, 0.600, 0.918, 1.761, 1.232, 2.431, and 2.642, respectively. Therefore, the ratios of the governance contribution to the pollution contribution of Hangzhou, Suzhou, and Jiaxing were less than 1 and thus should be adjusted.

Environmental Gini coefficient optimisation model application

Adjustment path setting

Two main adjustment paths exist: reducing the water pollution contribution by reducing the water pollutant emissions, and increasing the water governance contribution by increasing the water governance inputs. The ratio of the industrial production contribution coefficient to the water pollution contribution coefficient was proposed as the basis for the choice of the path. For example, if the ratio of the industrial production contribution coefficient to the water pollution contribution coefficient of Hangzhou is less than 1, then its pollution contribution should first be adjusted; otherwise, the governance contribution should first be adjusted.

Adjustment constraint setting

Considering feasibility, efficiency and equality, the adjustment targets for each DMU were set as follows: (1) the lower and upper limits of the water pollutant emissions reduction rate were −20% and 0%, respectively; and (2) the lower and upper limits of the water governance input increase rate were 0% and 20%, respectively.

Adjustment results and discussion

According to the adjustment path and targets set above, the water pollutant emissions and the water governance inputs of Hangzhou, Suzhou, and Jiaxing were adjusted by using the environmental Gini coefficient optimisation model. The adjustment results are shown in Table 4.

Table 4

The values and rates of change of the parameters after the adjustment

DMU Water pollutant emissions Water governance inputs Water pollution contribution Water governance contribution Governance contribution/pollution contribution Production contribution/pollution contribution 
Hangzhou 0.254TP0 (−19.84%) 0.222TG0 (12.68%) 0.291 (−7.94%) 0.208 (5.72%) 0.715 (14.84%) 0.486 (8.63%) 
Suzhou 0.234TP0 (−14.85%) 0.197TG0 (19.61%) 0.268 (−2.22%) 0.185 (12.22%) 0.689 (14.77%) 1.417 (2.27%) 
Jiaxing 0.104TP0 (−19.84%) 0.127TG0 (7.21%) 0.119 (−7.94%) 0.120 (0.59%) 1.003 (9.26%) 0.571 (8.63%) 
Wuxi 0.100TP0 (0.00%) 0.176TG0 (0.00%) 0.115 (14.84%) 0.165 (−6.18%) 1.439 (−18.30%) 1.646 (−12.92%) 
Zhenjiang 0.086TP0 (0.00%) 0.106TG0 (0.00%) 0.098 (14.84%) 0.099 (−6.18%) 1.006 (−18.30%) 0.699 (−12.92%) 
Huzhou 0.048TP0 (0.00%) 0.117TG0 (0.00%) 0.055 (14.84%) 0.109 (−6.18%) 1.986 (−18.30%) 0.718 (−12.92%) 
Changzhou 0.046TP0 (0.00%) 0.121TG0 (0.00%) 0.053 (14.84%) 0.114 (−6.18%) 2.158 (−18.30%) 2.142 (−12.92%) 
Total 0.871TP0 (−12.92%) 1.066TG0 (6.59%) 1.000 (0.00%) 1.000 (0.00%) 8.996 (−11.86%) 7.678 (−7.88%) 
DMU Water pollutant emissions Water governance inputs Water pollution contribution Water governance contribution Governance contribution/pollution contribution Production contribution/pollution contribution 
Hangzhou 0.254TP0 (−19.84%) 0.222TG0 (12.68%) 0.291 (−7.94%) 0.208 (5.72%) 0.715 (14.84%) 0.486 (8.63%) 
Suzhou 0.234TP0 (−14.85%) 0.197TG0 (19.61%) 0.268 (−2.22%) 0.185 (12.22%) 0.689 (14.77%) 1.417 (2.27%) 
Jiaxing 0.104TP0 (−19.84%) 0.127TG0 (7.21%) 0.119 (−7.94%) 0.120 (0.59%) 1.003 (9.26%) 0.571 (8.63%) 
Wuxi 0.100TP0 (0.00%) 0.176TG0 (0.00%) 0.115 (14.84%) 0.165 (−6.18%) 1.439 (−18.30%) 1.646 (−12.92%) 
Zhenjiang 0.086TP0 (0.00%) 0.106TG0 (0.00%) 0.098 (14.84%) 0.099 (−6.18%) 1.006 (−18.30%) 0.699 (−12.92%) 
Huzhou 0.048TP0 (0.00%) 0.117TG0 (0.00%) 0.055 (14.84%) 0.109 (−6.18%) 1.986 (−18.30%) 0.718 (−12.92%) 
Changzhou 0.046TP0 (0.00%) 0.121TG0 (0.00%) 0.053 (14.84%) 0.114 (−6.18%) 2.158 (−18.30%) 2.142 (−12.92%) 
Total 0.871TP0 (−12.92%) 1.066TG0 (6.59%) 1.000 (0.00%) 1.000 (0.00%) 8.996 (−11.86%) 7.678 (−7.88%) 

As shown in Table 4, the water pollutant emissions decreased by 19.84% in Hangzhou, 14.85% in Suzhou, and 19.84% in Jiaxing. The water governance inputs increased by 12.68% in Hangzhou, 19.61% in Suzhou, and 7.21% in Jiaxing. Using the adjustment, the total amount of water pollutant emissions (TP) decreased by 12.92%, and the total amount of water governance inputs (TG) increased by 6.59%. These results meet the adjustment guideline of ‘less water pollution and more governance inputs’. After the adjustment, the environmental Gini coefficient of COD was 0.199, and the reduction rate was −27.63%.

CONCLUSIONS

In this study, the environmental Gini coefficient was applied to the allocation of water governance responsibilities to improve the allocation equality. The environmental Gini coefficient optimisation model was proposed to achieve the optimal adjustment for the assessment results of the water governance responsibility allocation inequality.

To verify the suitability of the method, Taihu Lake Basin, a heavily polluted transboundary basin in China, was chosen as a case study. The case study showed that: (1) the environmental Gini coefficient model can scientifically evaluate the inequality of water governance responsibility allocation and identify the inequality factors; and (2) the environmental Gini coefficient optimisation model can significantly improve the equality of water governance responsibility allocation and optimise water pollutant emissions and water governance inputs.

This research gives decision makers the ability to assess the equality of water governance responsibility allocation and provides clear guidance for developing appropriate policies to optimise water governance responsibility allocation.

ACKNOWLEDGEMENT

This research was supported by the National Natural Social Science Foundation of China under Grant No. 12CGL068.

REFERENCES

REFERENCES
Chames
A.
Cooper
W. W.
Rhodes
E.
1978
Measuring the efficiency of decision making units
.
European Journal of Operational Research
2
(
6
),
429
444
.
Chames
A.
Cooper
W. W.
Golany
B.
Seiford
L.
Stutz
J.
1985
Foundations of data envelopment analysis for Pareto-Koopmans efficient empirical production functions
.
Journal of Econometrics
30
(
1/2
),
91
107
.
Chen
B.
Gu
C. S.
Wang
H. Z.
2012
Research on Gini coefficient application in the distribution of water pollution
.
Journal of Food Agriculture & Environment
10
(
2
),
833
838
.
Groves-Kirkby
C. J.
Denman
A. R.
Phillips
P. S.
2009
Lorenz curve and Gini Coefficient: novel tools for analyzing seasonal variation of environmental radon gas
.
Journal of Environmental Management
90
(
8
),
2480
2487
.
He
G. Z.
Lyu
Y. L.
2008
Efficiency evaluation on Taihu Lake water pollution control
.
Acta Ecologica Sinica
28
(
12
),
6348
6354
.
Heil
M. T.
Wodon
Q. T.
2000
Future inequality in CO2 emissions and the impact of abatement proposals
.
Environmental and Resource Economics
17
(
2
),
163
181
.
Jiang
K.
2010
Decomposing and analyzing the environmental pollution of different type of pollution of industry: based on the panel data of industrial from 1998 to 2005
.
Ecological Economy
4
,
96
99
.
Millimet
D. L.
Slottje
D.
2002
Environmental compliance costs and the distribution of emissions in the U.S
.
Journal of Regional Science
42
(
1
),
87
105
.
Qin
D. L.
Wei
A. L.
Lu
S. Y.
Luo
Y. P.
Luo
Y. P.
Liao
Y. H.
Yi
M.
Song
B. B.
2013
Total water pollutant load allocation in Dongting Lake area based on the environmental Gini coefficient method
.
Research of Environmental Sciences
26
(
1
),
8
15
.
Sadras
V.
Bongiovanni
R.
2004
Use of Lorenz curves and Gini coefficients to assess yield inequality within paddocks
.
Field Crops Research
90
(
2–3
),
303
310
.
Wang
M. L.
Luo
B.
Zhou
W. B.
Huang
Y.
Liu
L.
2011
The application of Gini coefficient in the total load allocation in Jinjiang River, China. In: 2011 International Symposium on Water Resource and Environmental Protection
,
Xi'an
,
China
,
20–22 May 2011, IEEE Computer Society
.
Xie
G.
Peng
Y. B.
Li
B. C.
Hang
Y. C.
Cui
Z. J.
2006
A comparative study on water pollution control between TMDL and small river basin pollution comprehensive treatment
.
Journal of Agricultural Mechanization Research
5
,
189
192
.
Zhang
X. H.
Huang
C. Y.
Zhang
H. W.
2010
The Application of Gini Coefficient in Regional Environmental Pollutants Distribution Plan
. In:
2nd Conference on Environmental Science and Information Application Technology
,
Wuhan
,
China
,
17–18 July 2010, IEEE Computer Society
.
Zhang
B.
Liu
H.
Yu
Q. Q.
Bi
J.
2012a
Equity-based optimisation of regional water pollutant discharge amount allocation: a case study in the Tai Lake Basin
,
Journal of Environmental Planning and Management
55
(
7
),
885
900
.
Zhang
Y.
Wang
X. Y.
Zhang
Z. M.
Shen
B. G.
2012b
Multi-level waste load allocation system for Xi'an-Xianyang Section, Weihe River
. In:
18th Biennial ISEM Conference on Ecological Modelling for Global Change and Coupled Human and Natural System
,
Beijing
,
China
,
20–23 September 2011, Procedia Environmental Sciences
.
Zhao
X. K.
2003
Environmental Protection and Industrial Competitiveness: A Theoretical and Empirical Analysis
.
China Social Sciences Press
,
Beijing
,
China
, pp.
1
473
.