Energy Efficiency Analysis of China Based on DEA Methods Under

Dual Carbon Target

Jingqi Wang

College of Science, North China University of Technology, Beijing, 100043, China

Keywords: DEA Methods, Energy Efficiency, Energy Consume, Low Carbon, Environment

Abstract: With the increasingly serious environmental problems, the realization of low-carbon energy transformation

in China's energy system has been promoted to the national strategic height. It has certain practical

significance and necessity to investigate its production efficiency scientifically and accurately and analysis

its improvement path. In this paper, the Data envelopment analysis (DEA) method is utilized to measure the

energy of 30 provinces and municipalities in China from 2010 to 2020 using deep2.1 software, from a

temporal and spatial perspective. According to the data analysis, this paper makes suggestions: pay attention

to the important role of technological innovation in improving energy efficiency, strengthen the scientific

research investment of new energy enterprises, cultivate internationally competitive energy talents; strengthen

regional energy cooperation, promote clean energy to replace traditional energy, increase the research on

clean coal utilization technology, improve coal utilization efficiency and reduce carbon emissions, and further

improve environmental protection policies, and increase the punishment of polluting enterprises. At the same

time, encourages enterprises to trade carbon emission rights, increase carbon emission costs, promote

enterprises to improve energy efficiency. In the production process, Chinese companies should try their best

to energy conservation and emission reduction, and reduce energy consumption. In energy consumption,

popularize energy-saving products and improve energy efficiency; strengthen regional energy cooperation

and promote the rational allocation of energy resources. Through regional coordination, the improvement of

energy efficiency can be realized, reduction of national energy consumption level can be promoted.

1 INTRODUCTION

Energy plays a crucial role in supporting human

survival and facilitating social development. With the

increasing energy demand, environmental problems

caused by energy shortage and excessive energy

consumption have gradually attracted people's

attention. How to effectively improve energy

efficiency is crucial for the sustainable development

of China's economy. Therefore, it is of great

theoretical and practical significance to scientifically

evaluate the energy efficiency situation and the major

influencing factors in China to provide reliable

suggestions and empirical data reference for

improving energy efficiency.

Through theoretical induction, the Data

envelopment analysis (DEA) model is widely used in

energy efficiency measurement. Yao et al. proposed

a methodology that decomposes total factor

productivity change into two distinct elements,

technological progress and changed in technical

efficiency (Yao et al, 2023). Guo et al. investigated

how entry and innovation affect total factor

productivity growth (Guo et al, 2023). Yang et al.

observed Significant effects that persist over time

(Yang et al, 2023). Jin et al. developed a framework

to evaluate China's agricultural research investment

trends and its impact on total factor productivity (Jin

et al, 2002). Chen examined heterogeneous total

factor productivity (TFP) (Chen & Moore, 2009).

Asche analyzed total factor productivity change

(Liang & Wang, 2023).

With the gradual improvement of DEA models

and the increasingly prominent problem of the

product of environment, domestic scholars also began

to study green total factor productivity. Li utilized the

Malmquist index and a spatial Durbin model to

analyze the impact of the effect on green total factor

productivity (Asche, 2013, Li & Wu, et al, 2017). Li

used the Super-SBM model to calculate China's

agricultural green total factor productivity according

to carbon emissions (Li & Lin, 2017). The carbon

Wang, J.

Energy Efﬁciency Analysis of China Based on DEA Methods Under Dual Carbon Target.

DOI: 10.5220/0012825500004547

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Data Science and Engineering (ICDSE 2024), pages 91-97

ISBN: 978-989-758-690-3

emissions are usually used to calculate green total

factor production efficiency (Liu et al, 2021).

The main body of this paper is energy efficiency.

Through the literature research, the author

understands the variety of testing methods for energy

efficiency, combined with the knowledge that has

been learned.

2 METHODOLOGY

2.1 Data Source and Description

For the construction of energy efficiency index

system, most scholars choose to consider energy,

labor, technology, capital input, economic output

interaction and alternative relationship of the total

factor energy efficiency index. This paper refers to

the actual research, and chooses capital input, labor

input, technology input three indexes as input

elements, with local fiscal revenue to measure capital

input, with the end of employment to measure labor

input. Research and experimental development

personnel collaborated to measure the impact of

technology, using per capita GDP as the desired

output and the carbon dioxide emissions as the

undesired output. The specific input-output index

system is shown in the following table.

𝐸𝐶 =

∑

𝐸𝐶



∑

𝐸



×𝐶𝐹



×𝐶𝐶



×𝐶𝑂𝐹



×3.6









(1)

Among them, represents the estimated total

carbon dioxide emissions of various energy

consumption; 𝑖 represents energy consumption,

including coal, coke, gasoline, kerosene, fuel oil and

natural gas. 6 is the total energy consumption of each

province: represents the heat value of energy:

represents the carbon content: represents the

oxidation factor of energy in i. It's called carbon

dioxide emission.

2.2 Index Selection

Most researchers prefer to consider the interaction

between energy, labor, science and technology,

capital input, economic output, and the substitution

relationship of the total factor energy efficiency index

when constructing the energy efficiency index

system. Drawing on the actual research, this paper

selects three indicators including capital input, labor

input, and technology input as input elements. Local

fiscal revenue is used to calculate capital input,

employment termination is used to measure labor

input, and research and development personnel are

counted for technology input. Per capita GDP is used

as the desired output, and carbon dioxide emissions

serve as the undesired output. The specific input-

output index system is shown in the following Table

Table 1: Index selection and unit.

Indicator t

e name of index Index unit

Investment index

Local fiscal revenue 100 million

umber of people emplo

ed at the end of the

human bein

Research and trial development personnel human bein

Expect output indicators Per capita GDP firs

Undesired output indicators Carbon dioxide emissions Ten thousand tons

2.3 Research Method

DEA-BCC Model measures the pure technical

efficiency by assuming a variable return of scale, it is

also called the variable return of scale (VRS) model.

BCC The model is shown as follows:

When σ =1, then the pure technical efficiency of

DMU is DEA effective. If the comprehensive

technical efficiency is θ and the pure technical

efficiency is σ, then the scale efficiency of DMU = θ

/ σ (Yang et al, 2023).

Malmquist Model: The Malmquist Index method

is a helpful tool for analyzing efficiency changes

across multiple samples. It can show the relationship

between comprehensive efficiency, technical

efficiency, and total factor production efficiency

index. This method can dynamically track the

changes in efficiency values of sample data over

different time periods. The Malmquist index was first

proposed in 1953 to solve the problem of variations

in the consumption bundle in the consumption

function. CAVES applied the Malmquist index to

analyze energy production efficiency (Asche, 2013).

ICDSE 2024 - International Conference on Data Science and Engineering

𝑀



𝑦



,𝑥



,𝑦



,𝑥





















,















,





∗











,















,





 (2)

Where



𝑥



,𝑦





𝑥



,𝑦





𝐷



𝐷



For input and

output variables in 𝑡+1 and 𝑡; for decision unit

distance function in 𝑡+1 and 𝑡.

3 RESULTS AND DISCUSSION

Energy plays a crucial role as a material foundation

for human survival and social progress. With the

increasing energy demand, environmental problems

caused by energy shortage and excessive energy

consumption have gradually attracted people's

attention.

3.1 Descriptive Statistics

Based on previous data comparing energy

consumption and carbon emissions. In this research,

the DEA-Malmquist method is applied to measure the

energy efficiency of China from 2010-2020 from both

static and dynamic perspectives, and to compare the

efficiency differences of the seven regions (Figure 1).

Figure 1: Comparing energy consumption and carbon emissions plot (Picture credit: Original)

Using the original input and output data and

Deap2.1 software, the total energy factor productivity

and decomposition index were calculated for the 30

provinces in China. and municipalities from 2010 to

2020. Combined with the Malmquist index, the total

factor production efficiency is used to evaluate it

dynamically. As shown in Table 2., the average total

factor production efficiency of 30 provinces and

municipalities in China from 2010 to 2020 is 0.951,

indicating that the energy efficiency of these 30

provinces and municipalities has not yet reached the

effective state. From 2010 to 2014, tfp increased year

by year, with a significant decrease in 2014-2015.

From 2015 to 2016, it recovered to the level of 2013-

2014 and probably decreased from 2016 to 2019.

From 2010 to 2020, the development of energy

efficiency in all regions of China will be divided into

three phases: the first from 2010 to 2014, the second

from 2014 to 2016, and the three from 2016 to 2019.

(Table 2).

5000

10000

15000

20000

25000

30000

35000

40000

0.00

10000.00

20000.00

30000.00

40000.00

50000.00

60000.00

70000.00

80000.00

90000.00

100000.00

Carbon dioxide emissions

Total energy consumption

Carbon dioxide emissions

Total energy consumption

Energy Efﬁciency Analysis of China Based on DEA Methods Under Dual Carbon Target

Table 2: Malmqusit index decomposition of energy efficiency.

ear eff tech

eh sec tf

2010-2011 1.027 0.870 1.015 1.012 0.894

2011-2012 1.021 0.910 1.023 0.998 0.928

2012-2013 1.000 0.961 1.005 0.996 0.962

2013-2014 1.016 0.962 1.016 1.000 0.977

2014-2015 0.994 0.932 0.995 0.999 0.927

2015-2016 1.022 0.958 1.033 0.989 0.979

2016-2017 0.972 0.995 0.960 1.012 0.967

2017-2018 1.010 0.961 1.006 1.003 0.970

2018-2019 1.035 0.929 1.105 0.936 0.961

mean 1.011 0.941 1.017 0.994 0.951

According to the principle of the Malmquist index

method, the total factor production efficiency is the

product of the comprehensive technical efficiency

change index and the technical change index.

In the first stage, the contribution to the growth

of total factor production efficiency is technological

changes. The sudden decline of total factor

production efficiency in the second stage is mainly

due to the decline of technological changes, while the

scale also decreases in efficiency. Combined with the

literature research, environmental protection was

valued in 2014.31 provinces and the Ministry of

Environmental Protection signed the "Target

Responsibility Letter for Air Pollution Prevention and

Control", and the policy had a great impact on the

technical change index, thus affecting the total factor

productivity. In the third stage, the total factor

productivity index continued to decline, but from

2018 to 2019, the technical change index increased.

The scale efficiency index decreased, indicating that

the development of new energy technology in China

is effective. The technology application is in scale,

and it is in the stage of energy technology

transformation and development after the

environmental protection policy.

In general, technical changes play a leading role

in the total factor efficiency of production. In

contrast, the change is relatively flat at present, which

may require further optimization of resource

allocation and scale expansion strategy.

3.2 Regional Trend

Without the influence of environmental effects and

random factors, the total factor growth rate decreased

by 1% annually from 2010 to 2020 (table 3). Surgical

efficiency (TECH) increased by 2.4% annually;

technological progress level. TCH decreased by 3.2%

annually. This shows that the decline in the

productivity of environmental service enterprises is

mainly caused by the relative decline of technology.

From the perspective of segmentation, only the

average annual productivity of environmental

monitoring enterprises has seen a small improvement,

and the main driving force is the obvious

improvement in the technical efficiency of

enterprises. However, the productivity of enterprises

in the other five categories showed a small downward

trend, with a decline rate of 0.56% ~2.25%.

Between, the main reason is the negative impact

of the technological relative regression on

productivity. The results of the first stage show that,

without considering the influence of environmental

factors and random factors, the improvement of

correct management decisions is not enough to offset

the adverse impact of the relative decline of

technology level on the production efficiency of

enterprises.

From the perspective of regional analysis, the

results under environmental constraints energy

efficiency and overall technology progress level are

low (only 0.94) has not reached the equilibrium

degree. The efficiency difference between provinces,

regions, in the study period of Beijing, Shanghai

province energy efficiency mean is 1, every year

reached the effective state, and the average energy

efficiency in Guizhou is only 0.87.

Energy efficiency in North China and Eastern

China is significantly higher than in other regions. In

this regard, the differences in resource endowment of

different regions, as well as the supply and demand of

different resources and the development degree of

utilization, and relevant policies should be formulated

according to local conditions.

From a dynamic point of view, between 2010-

2020 our country's energy total factor productivity

changed overall downward trend, and gradually in the

good direction in recent years, the technological

progress is the main factor of driving energy

efficiency growth, should adhere to resource

ICDSE 2024 - International Conference on Data Science and Engineering

orientation and science and technology, and attach

importance to technology research and development,

actively develop clean coal utilization technology,

unconventional oil and gas exploration and

development technology, but also to promote the

development of energy conservation and emissions

reduction technology, intensify mining environment

monitoring, increase the proportion of clean energy

consumption, under the condition of low carbon

environmental protection improve energy efficiency

(table 4).

Table 3: Static index analysis of energy efficiency.

area firm crste vrste scale

Sichuan 25 0.135 0.135 1.000

Henan 10 0.165 0.165 1.000

Guan

don

5 0.169 0.169 1.000

Shandon

21 0.196 0.196 1.000

Jian

su 15 0.207 0.207 1.000

Hebei 9 0.209 0.209 1.000

Hunan 13 0.212 0.212 1.000

Anhui 1 0.221 0.221 1.000

Hubei 12 0.243 0.243 1.000

Yu nn an 28 0.251 0.251 1.000

Liaonin

17 0.265 0.291 0.913

the Heilon

ian

Rive

11 0.268 0.268 1.000

Shaanxi Province 23 0.271 0.271 1.000

Guan

xi 6 0.287 0.287 1.000

Shanxi 22 0.289 0.289 1.000

Jian

xi 16 0.299 0.299 1.000

Zhe

ian

29 0.314 0.314 1.000

Guizhou 7 0.349 0.349 1.000

Xin

ian

27 0.351 0.351 1.000

Chon

qin

30 0.365 0.365 1.000

Jilin 14 0.387 0.474 0.816

Gansu 4 0.393 0.393 1.000

ei Mon

ol 18 0.430 1.000 0.430

ian 3 0.479 0.479 1.000

Shan

hai 24 0.672 1.000 0.672

Bei

2 1.000 1.000 1.000

Hainan 8 1.000 1.000 1.000

xia 19 1.000 1.000 1.000

Qin

hai 20 1.000 1.000 1.000

Tian

in 26 1.000 1.000 1.000

Table 4: Total factor energy efficiency.

area firm effch techch pech sech tfpch

Anhui 1 1.01 0.89 1.02 7.00 0.90

Bei

2 1.00 1.08 1.00 1.99 1.08

ian 3 1.04 0.95 1.08 0.97 0.99

Gansu 4 1.02 0.88 1.02 2.00 0.90

Guan

don

5 1.01 0.97 1.03 0.98 0.97

Guan

xi 6 1.01 0.89 1.01 2.00 0.89

Guizhou 7 0.98 0.88 0.98 1.10 0.87

Hainan 8 1.00 0.96 1.00 2.00 0.96

Hebei 9 0.99 0.92 0.99 1.09 0.91

Henan 10 1.01 0.90 1.01 2.00 0.91

Energy Efﬁciency Analysis of China Based on DEA Methods Under Dual Carbon Target

Table 4: Total factor energy efficiency (cont.).

area firm effch techch

ech sech tf

The Heilon

ian

Rive

11 1.03 0.92 1.03 2.00 0.94

Hubei 12 1.03 0.95 1.04 9.98 0.98

Hunan 13 1.01 0.93 1.01 7.00 0.94

Jilin 14 1.02 0.96 1.00 1.92 0.98

Jian

su 15 1.05 0.95 1.09 8.96 0.99

Jian

xi 16 0.99 0.89 0.99 1.01 0.88

Liaonin

17 1.01 0.96 1.01 1.10 0.97

ei Mon

ol 18 0.99 1.00 1.00 0.99 0.99

xia 19 1.00 0.98 1.00 1.90 0.98

Qin

hai 20 1.00 0.99 1.00 2.00 0.99

Shandon

21 1.00 0.96 1.01 9.99 0.96

Shanxi 22 1.01 0.91 1.01 1.10 0.91

Shaanxi Province 23 1.03 0.95 1.04 10.00 0.98

Shan

hai 24 1.02 1.04 1.00 1.18 1.06

Sichuan 25 1.03 0.89 1.03 1.99 0.92

Tian

in 26 0.98 0.97 1.00 6.98 0.96

Xin

ian

27 1.01 0.94 1.02 9.99 0.95

Yu nn an 28 1.01 0.88 1.01 1.09 0.89

Zhe

ian

29 1.01 0.95 1.05 9.97 0.96

Chon

qin

30 1.04 0.96 1.06 0.98 0.99

mean 1.01 0.94 1.02 9.99 0.95

3.3 Discussion

According to the above studies, the paper puts

forward the following suggestions:

Chinese energy enterprises should increase their

investment in new energy technology innovation, and

attach importance to the important role of

technological innovation in improving energy

efficiency, to improve China's energy efficiency. The

government and enterprises should increase their

investment in technological innovation, increase the

allocation of research and development funds and

human resources, and cultivate internationally

competitive energy technologies.

China needs to work on optimizing its energy mix

by promoting the use of clean energy in place of

traditional energy sources, and increasing the

proportion of clean energy in the total energy

consumption. It is also important for enterprises to

focus on researching clean coal utilization technology

to improve the efficiency of coal utilization and

reduce carbon emissions. Strengthening

environmental protection policies. The Chinese

government will further improve environmental

protection policies, increase penalties for polluting

enterprises, and guide them to take the path of green

development. At the same time, enterprises are

encouraged to carry out carbon emission rights

trading, increase the cost of carbon emission, and

promote enterprises to improve energy efficiency.

Strengthen the training of energy talents. China

aims to provide comprehensive training for a diverse

pool of energy professionals, equipping them with a

global perspective and enhancing their expertise in

areas such as energy technology research and

development, as well as energy management.

Simultaneously, the Chinese government will

prioritize the enhancement of professional

development programs for individuals already

employed within the energy industry, thereby

elevating the overall quality standards across the

sector.

Strengthen regional coordination. The Chinese

government should strengthen energy cooperation

among regions and promote rational allocation of

energy resources. Through regional coordination,

China can improve energy efficiency and reduce

national energy consumption.

4 CONCLUSION

Based on the measurement of energy data from

various regions in China from 2010 to 2020, this

paper finds that China's energy efficiency has not yet

reached its overall effective state, and there are

significant fluctuations in energy efficiency from

2014 to 2016. Under this premise, this paper puts

forward policy suggestions such as increasing

ICDSE 2024 - International Conference on Data Science and Engineering

investment in technological innovation, optimizing

energy structure and strengthening environmental

protection policies to provide reference for improving

energy efficiency and green and low-carbon

development in China. Through the implementation

of these policies, China's energy enterprises can help

to improve production efficiency, reduce carbon

emissions, and achieve sustainable development. At

the same time, the government, enterprises and

scientific research institutions should make joint

efforts to contribute to improving energy efficiency

and green and low-carbon development in China.

REFERENCES

S. C. Yao, J. Q. Zhi, J. B. Fu, et al. Journal of South China

University of Technology: Natural Science Edition

51(6), 97-108 (2023).

A. Guo, L. Niu, P. J. Liu, et al. Economic Geography 43(9),

172-178 (2023).

H. Yang, Y. H. Pan and G. W. Cai, China Industrial

Economy 7, 46-65 (2023).

S. Q. Jin, J. K. Huang, R. F. Hu and S. Rozelle, International

Political Economy, (2002).

M. X. Chen and M. O. Moore, Journal of International

Economics, (2009).

Y. W. Liang and M. L. Wang, Agricultural Economic

Issues 10, 101-115 (2023).

F. Asche, G. Atle and R. Nielsen, Aquaculture, (2013).

B. Li and S. S. Wu, Journal of Clean Production, (2017).

K. Li and B.Q Lin, Applied Energy, (2017).

D. D. Liu, X. Y. Zhu and Y. F. Wang, Journal of Clean

Production, (2021).

Energy Efﬁciency Analysis of China Based on DEA Methods Under Dual Carbon Target