DEA-based Ecological Efficiency Evaluation during the

Process of Industrial Transformation in the Pearl River Delta

Urban Cluster

M R Su

1,*

, D Z Li

, C Ma

, L L Liao

and L R Yu

Research Center for Eco-Environmental Engineering, Dongguan University of

Technology, Dongguan 523808, China.

School of Environment and Civil Engineering, Dongguan University of Technology,

Dongguan 523808, China.

Corresponding author and e-mail: M R Su, sumr@dgut.edu.cn

Abstract. In order to comprehensively evaluate the effect of industrial transformation, the

concept of ecological efficiency was introduced in this paper due to its merit of integrating

social, economic, and environmental factors. The indicator system and model of ecological

efficiency evaluation was subsequently established, while the model based on data

envelopment analysis is hopeful to reduce the subjectivity of evaluation. The case study on the

Pearl River Delta Urban Cluster shows the generally increased ecological efficiency during the

process of industrial transformation. However, the declined returns to scale implies the lack of

satisfactory harmony among different cities in the Pearl River Delta Urban Cluster. Certain

suggestions to improve ecological efficiency was put forward based on the results.

1. Introduction

Industrial transformation is undoubtedly a hopeful option of socio-economic development especially

in the regions where the traditional economy has caused certain eco-environmental problems. In the

past decades, many cities and regions in China has experienced the industrial transformation, the

remained question is that whether the industrial transformation is effective, i.e., realizes the desired

goal of harmonious socio-economic and eco-environmental development.

In order to comprehensively evaluate the effect of industrial transformation, the concept of

ecological efficiency is introduced in this paper regarding its integration and linkage among social,

economic, and environmental factors [1]. Although ecological efficiency has been widely studied on

such scales as nation, region, and enterprise [2–6], there is few attempt from the viewpoint of

industrial transformation. And more effective evaluation method is still necessary to resolve the

existent problems, e.g., reducing the subjectivity of the evaluation.

Focusing on these demands, the evaluation method of ecological efficiency based on data

envelopment analysis is constructed in this paper, attempting to evaluate the overall effect of

industrial transformation. The case study is conducted for the Pearl River Delta Urban Cluster, the

representative of industrial transformation in China with relatively long history and remarkable

Su, M., Li, D., Ma, C., Liao, L. and Yu, L.

DEA-based Ecological Efﬁciency Evaluation During the Process of Industrial Transformation in the Pearl River Delta Urban Cluster.

In Proceedings of the International Workshop on Environmental Management, Science and Engineering (IWEMSE 2018), pages 301-307

ISBN: 978-989-758-344-5

301

outcome. Certain suggestions to improve ecological efficiency is finally proposed based on the

results.

2. Methodology

2.1. Concept of ecological efficiency

Although there are various understandings of ecological efficiency among different organizations and

scholars [2, 4, 6–8], its essential point is always the efficiency, i.e., the ratio of the output to input for

a system. Differing from the general efficiency, ecological efficiency also considers the resource and

environmental factors besides the traditional economic factor, e.g., it investigates such undesired

output as waste gas and wastewater besides economic output along with the resource and

environmental input.

2.2. Evaluation indicators of ecological efficiency

According to the essential point of ecological efficiency, the evaluation indicator system were

confirmed, when referring to indicators established in the traditional German economic account [1],

the data availability and the correlation among different indicators. As indicated in Table 1, it mainly

includes three aspects, i.e., resource, environment, and economy, in which the first two aspects

mainly reflect the input dimension while the last one represents the output dimension. And it should

be pointed out that the undesired output (i.e., environmental discharge) was regarded as the input

indicator rather than output in this paper, when considering the following two facts. The first one is

that it is difficult to collect accurate data of investment for dispose of undesired output, while the

undesired output is an alternative. Another one is that the undesired output can be regarded as a kind

of input per se for the environment since it needs absorption and dispose by the environment.

Table 1. The evaluation indicator system of ecological efficiency.

Dimension

Aspect

Item

Indicator

Input

Resource

Labor

Amount of employment at the end of

the year

Land

Area of cultivated land at the end of

the year

Area of built-up parts

Water resource

Amount of non-residential water

consumption

Energy

Amount of non-residential electricity

consumption

Environment

Water environment

Amount of industrial wastewater

discharge

Atmospheric environment

Amount of industrial waste gas

emission

Output

Economy

Gross domestic product

2.3. Evaluation model of ecological efficiency based on data envelopment analysis

The weights of different indicators are always an open question which may influence the results of

comprehensive assessment. With the ability of conducting assessment without confirming weights of

indicators, the method named as data envelopment analysis was selected to implement evaluation of

ecological efficiency. To better satisfy the demand of dynamic evaluation of ecological efficiency

IWEMSE 2018 - International Workshop on Environmental Management, Science and Engineering

302

during the process of industrial transformation, the BCC (Banker, Charnes, Cooper) model was

applied in which the returns to scale is variable.

In the BCC model, it is assumed that there are n decision making unit and each unit has s input and

t output. When x

represents the input of the jth decision making unit, and y

means the output of the jth

decision making unit, then the efficiency of the jth unit can be converted into a problem of linear

programming solution, when θ means the efficiency.

min θ

(1)

(2)

(3)

Concretely speaking, the decision making unit is the Pearl River Delta Urban Cluster and the nine

typical cities, the input is that from resource and environment, while the output is the economic

outcome.

In order to analyze the change of ecological efficiency during the process of industrial

transformation, the Malmquist index was combined with the merit of time series analysis. The

calculation is given in Eq. (4):

(4)

where 















and 















is the single stage distance function with fixed returns to

scale, while 















and 















is the inter-temporal distance function with fixed

returns to scale.

The result parameter obtained from data envelopment analysis and Malmquist index can be mainly

divided into three types, including the overall efficiency, the input slackness degree, and the change of

efficiency.

(1) The overall efficiency

Pure technological efficiency: the efficiency influenced by management and technology,

Scale efficiency: the efficiency influenced by industrial scale,

Comprehensive efficiency: the integrated efficiency influenced by both technology and industrial

scale.

(2) The input slackness degree

The input is redundant when the value of input slackness degree > 0, and the input is insufficient

when the value < 0, and the input is suitable when the value = 0.

(3) The change of efficiency

Change of technological efficiency: the technological efficiency increases when the value ≥ 1, and

the efficiency decreases when the value < 1,

Change of technological progress: the technological level improves when the value ≥ 1, and the

technological level declines when the value < 1,

Change of pure technological efficiency: the pure technological efficiency increases when the value

≥ 1, and the efficiency decreases when the value < 1,

s.t.

jn

 x



jn

 y



= 1



 0 j  n



t+1



= [

t+1



t+1



CRS



t+1





CRS







t+1



CRS







CRS



]

DEA-based Ecological Efﬁciency Evaluation During the Process of Industrial Transformation in the Pearl River Delta Urban Cluster

303

Change of scale efficiency: the scale efficiency increases when the value ≥ 1, and the efficiency

decreases when the value < 1,

Change of total factor productivity: it means the ratio of total production to total factor input. The

total factor productivity increases when the value ≥ 1, and the efficiency decreases when the value < 1.

2.4. Data sources

Required data during 1997–2015 were mainly collected from various statistical yearbook, e.g.,

Guangdong Statistical Yearbook, China City Statistical Yearbook, Yangtze River Delta & Pearl

River Delta and Hong Kong & Macao SAR & Tai Wan Statistical Yearbook, as well as the statistical

yearbooks of the nine typical cities in the Pearl River Delta. Moreover, some data were acquired from

the online resources, e.g., the website of municipal water authority and Guangdong statistical

information.

3. Results and discussion

3.1. Overall ecological efficiency of the Pearl River Delta Urban Cluster

Based on the BCC model, the overall ecological efficiency of the Pearl River Delta Urban Cluster

and nine cities were obtained, as shown in Table 2. Three findings are investigated from the results.

(1) Except for Dongguan and Zhongshan, the comprehensive efficiency of other seven cities was

higher than that of the Pearl River Delta Urban Cluster during the study period. It means that the

usage of input was efficient in these seven cities when using that of the Pearl River Delta Urban

Cluster as the baseline. (2) Although with relatively low comprehensive efficiency, Dongguan and

Zhongshan had increased returns to scale during the study period, which implies that the ecological

efficiency was gradually improved during the process of industrial transformation. (3) The Pearl

River Delta Urban Cluster showed decreased returns to scale, although none city showed the same

trend during the study period. It reflects that the coordination among different cities in the Pearl

River Delta Urban Cluster was unsatisfactory.

Table 2. The ecological efficiency obtained from BCC model.

Area

Comprehensive

efficiency

Pure

technological

efficiency

Scale

efficiency

Change of scale

efficiency

Guangzhou

1.000

Shenzhen

1.000

Zhuhai

1.000

Huizhou

1.000

Dongguan

0.749

0.861

0.870

irs

Zhongshan

0.848

1.000

0.848

irs

Jiangmen

1.000

Foshan

1.000

Zhaoqing

1.000

The Pearl River Delta

0.897

1.000

0.897

drs

Note: ‘irs’ means the returns to scale increase, ‘drs’ means the returns to scale decrease, and ‘-’ means

the returns to scale remain unchanged.

3.2. The input slackness degree of ecological efficiency for the Pearl River Delta Urban Cluster

The input slackness degree of ecological efficiency for the Pearl River Delta Urban Cluster were also

calculated based on the BCC model. It is denoted in Table 3 that the input of employment and

IWEMSE 2018 - International Workshop on Environmental Management, Science and Engineering

304

non-residential water were non-redundant, whereas that of industrial waste gas, non-residential

electricity, cultivated land, and built-up area showed obvious redundancy during the study period. It

implies that the ecological efficiency during the process of industrial transformation is closely related

with resource and environmental input, and the influence of industrial transformation on resource and

environment should be paid attention to.

Table 3. The input slackness degree based on BCC model.

Aspect

Indicator

Input slackness

degree

Resource

Amount of employment at the end of the year

Area of cultivated land at the end of the year

1.365

Area of built-up parts

0.718

Amount of non-residential water consumption

Amount of non-residential electricity consumption

3.124

Environment

Amount of industrial wastewater discharge

0.053

Amount of industrial waste gas emission

49.376

3.3. The dynamic analysis of ecological efficiency for the Pearl River Delta Urban Cluster

The dynamics of ecological efficiency for the Pearl River Delta Urban Cluster was analyzed based on

the Malmquist index method. As shown in Figure 1, the change value of total factor productivity was

bigger than one in most of years during 1997–2015, representing the ecological efficiency of the

Pearl River Delta Urban Cluster generally improved during the process of industrial transformation.

However, it is also found out that the change value of total factor productivity was smaller than one

during 2012–2015, showing the declined trend of ecological efficiency in recent years which was

mainly caused by the decrease of scale efficiency. Due to the big industrial scale in the Pearl River

Delta Urban Cluster, the harmony among various aspects of different industries and cities has not

been achieved, thus affects the ecological efficiency.

Figure 1. The change of ecological efficiency for the Pearl River Delta Urban Cluster.

0.75

1.25

1.5

Change of technological efficiency

Change of technological progress

Change of pure technological efficiency

Change of scale efficiency

Change of total factor productivity

DEA-based Ecological Efﬁciency Evaluation During the Process of Industrial Transformation in the Pearl River Delta Urban Cluster

305

3.4. Suggestions for improving ecological efficiency in the Pearl River Delta Urban Cluster

Based on the above-mentioned results, the suggestions for improving ecological efficiency in the

Pearl River Delta Urban Cluster was put forward from three aspect. First, some input needs to be

reduced when considering the redundancy, which mainly includes non-residential electricity,

cultivated land, and built-up area. In order to reduce the non-residential electricity consumption, the

industry should transfers into that of low energy consumption, and the energy-saving awareness of

employees should be improved. The productive cultivation technique needs to be studied and

developed to decrease the redundancy of the cultivated land. In terms of the input of built-up area, a

more rational planning are necessary combining with the urban development ability and orientation.

Second, undesired output needs to be reduced, which mainly includes the waste gas and wastewater.

The industry is required transferred from high emission to low emission. Third, the urban network in

the Pearl River Delta Urban Cluster needs to be optimized, which requires a more holistic industrial

planning in the whole urban cluster where different cities develop advantageous complementary

industry.

4. Conclusions

In order to check the effect of industrial transformation, the concept of ecological efficiency was

introduced in this paper regarding its integration of economy, resource, and environment. The

evaluation indicator system of ecological efficiency was established. The evaluation model based on

data envelopment analysis and the Malmquist index was also developed, which is hopeful to reduce

the evaluation subjectivity without confirming the weights of different indicators. The Pearl River

Delta Urban Cluster-the representative of industrial transformation in China-was selected as the case

to demonstrate the application of the established evaluation indicators and mathematical model. It is

indicated that the ecological efficiency of the Pearl River Delta Urban Cluster generally increased

during the process of industrial transformation but with declined returns to scale, which implies the

lack of satisfactory coordination among different cities. Such suggestions to improve ecological

efficiency is finally proposed as reduce certain input and undesired output, and strength the holistic

industrial planning in the Pearl River Delta Urban Cluster.

Acknowledgments

This research was financially supported by the National Key R & D Program of China (No.

2016YFC0502800, 2017YFC0405900), the National Natural Science Foundation of China

(No.71673027), the Natural Science Foundation for Distinguished Young Scholars of Guangdong

Province (No.2017A030306032), GDUPS (2017), and the Scientific Research Foundation for

High-level Talents and Innovation Team in Dongguan University of Technology (No.

KCYKYQD2016001).

References

[1] Hoh H, Scoer K and Seibel S 2001 Eco-efficiency indicators in German

environmental-economic accounting (Federal Statistical Office, Germany)

[2] Zhang Z, Zhu D J, Shi Q H and Cheng M W 2018 Which countries are more ecologically

efficient in improving human well-being? An application of the index of ecological

well-being performance Resou. Conserv. Recy. 129 pp 112–119

[3] Yue S J, Yang Y and Pu Z N 2017 Total-factor ecology efficiency of regions in China Ecol.

Indic. 73 pp 284–292

[4] Silveira J L, Lamas W de Q, Tuna C E, Villela I A de C and Miro L S 2012 Ecological efficiency

and thermoeconomic analysis of a cogeneration system at a hospital Renew.Sust. Energ. Rev.

16 pp 2894–2906

IWEMSE 2018 - International Workshop on Environmental Management, Science and Engineering

306

[5] Sun L Y, Miao C L and Yang L 2017 Ecological-economic efficiency evaluation of green

technology innovation in strategic emerging industries based on entropy weighted TOPSIS

method Ecol. Indic. 73 pp 554–558

[6] Coronado C R, Villela A de C and Silveira J L 2010 Ecological efficiency in CHP: Biodiesel

case Appl. Therm. Eng. 30 pp 458–463

[7] Jollands N 2006 Concepts of efficiency in ecological economics: Sisyphus and the decision

maker Ecol. Econ. 56 pp 359–372

[8] Katsaros G, Stichler P, Subirats J and Guitart J 2016 Estimation and forecasting of ecological

efficiency of virtual machines Future Gener. Comp. Sy. 55 pp 480–494

DEA-based Ecological Efﬁciency Evaluation During the Process of Industrial Transformation in the Pearl River Delta Urban Cluster

307