Energy Policy 97 (2016) 414–420

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Short communication

Study on decoupling analysis between energy consumption and

economic growth in Liaoning Province
Bai Dong a, Ming Zhang b,n, Hailin Mu a, Xuanming Su c
a
Dalian University of Technology, Xuzhou 116024, PR China
b
China University of Mining and Technology, Xuzhou 221116, PR China
c
National Institute for Environmental Studies, Tsukuba-City, 3058506, Japan

H I G H L I G H T S

 The Tapio decoupling method is utilized to study what the decoupling status occurred in Liaoning Province.
 The generalized LMDI method is used to find the reason why the decoupling status appeared.
 The energy intensity decoupling effect played a positive role in the development of decoupling.

art ic l e i nf o a b s t r a c t

Article history: Since 1978, Liaoning province has experienced spectacular economic growth, which has led to more
Received 19 February 2016 energy consumption. The purpose of this paper is to explore the decoupling status between energy
Received in revised form consumption and economic growth in Liaoning Province. Firstly, the generalized LMDI method is used to
12 July 2016
explore the driving forces governing production energy consumption in Liaoning province. Then, the
Accepted 31 July 2016
Available online 8 August 2016
combination of Tapio decoupling indicator and generalized LMDI method is utilized to study what the
decoupling status occurred in Liaoning Province and why the decoupling status appeared. During the
Keywords: study period, only four decoupling status occurred: expansive negative decoupling, expansive coupling,
Generalized LMDI method weak decoupling, and strong decoupling. The energy intensity decoupling effect played a positive role in
Decouple indicator
the appearance of decoupling. However, the economic structure decoupling effect and investment de-
Energy consumption
coupling effect played a negative role in the appearance of decoupling. Over the study period, the energy
structure decoupling effect and labour decoupling effect played a relative small role in the appearance of
decoupling.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction broad categories of decomposition techniques (Ma and Stern, 2008).

Currently, the global warming has become a serious issue in the
world. However, energy-related CO2 emission is the largest con-
tribution to the greenhouse effect. Energy source is the main
driving force of economic development. Along with spectacular
economic growth, more energy has been consumed, which also
results in seriously environmental pollution. Thus, whether an
economy is becoming less dependent on energy consumption has
been paid attention to by many researchers.
The influencing factors governing energy consumption and its
environment emission can be explored by the decomposition
techniques (Zhang et al., 2013). Currently, SDA (Structural Decom-
position Analysis) and IDA (index decomposition analysis) are two
n
Corresponding author.
E-mail address: Zhangmingdlut@163.com (M. Zhang).
The base of SDA approach is the input-output tables. The SDA can
not only differentiate between direct and indirect energy demands,
but also distinguish between a range of technological effects and
structural effects. The SDA was used by Lin and Polenske (1995) to
study China’s energy use between 1981 and 1987. However, the
input-output tables may only be available sporadically (Su and Ang,
2012). The advantage of IDA is the utilization of time-series data.
IDA includes a variety of different indexing methods, such as the
Laspeyres decomposition and LMDI method. By comparing various
IDA methods, Ang (2004) concluded that the LMDI method was the
preferred method. Because there are the logarithmic terms in the
LMDI formulae, complications arise when the data set contains zero
values. Ang and Liu (2007) presented eight strategies to handle zero
values in LMDI method. By combining C-D production function and
LMDI method, Wang et al., (2014) presented a new generalized
LMDI method, which can be utilized to study many factors, such as
fixed asset investment and labour.

http://dx.doi.org/10.1016/j.enpol.2016.07.054
0301-4215/& 2016 Elsevier Ltd. All rights reserved.
 B. Dong et al. / Energy Policy 97 (2016) 414–420 415

The relationship between energy use or related CO2 emission
and economic growth can be explored by many methods, such as
simple regressions, correlation analysis, bivariate causality, panel
cointegration, multivariate cointegration, unit root testing, var-
iance decomposition and vector error correction modeling (Cli-
ment and Pardo, 2007). Among all the existing methods, the de-
coupling method is the best technique to characterize the de-
pendence of economic growth (GDP) on energy consumption. The
notion of decoupling was firstly introduced by Von (1989). How-
ever, Zhang (2000) firstly utilized the decoupling concept to ex-
plore the relationship between China’s energy-related CO2 emis-
sion and economic growth. In 2002, the OECD (2010) developed
that concept into an indicator. Currently, two existing decoupling
indexes have being widely used to figure out the dependence of
economic growth (GDP) on energy consumption.
One is the Taipo decoupling index developed by Tapio (2005).
Based on the elasticity concept, Juknys (2003) defined three kinds
of decoupling. The primary decoupling is defined as the decou-
pling of natural resources consumption from economic growth.
The secondary decoupling is defined as the decoupling of en-
vironmental pollution from natural resources consumption. If the
primary decoupling and secondary decoupling happen at the same
time, double decoupling will occur. Based on the Juknys's decou-
pling idea, Tapio (2005) redefined the Tapio decoupling indicator
and three decoupling statuses (i.e., decoupling, coupling and ne-
gative decoupling), which was also utilized to explore the decou-
pling status in the European transport industry. To distinguish
decoupling state rationally, Tapio divided three decoupling sta-
tuses into eight logical possibilities.
Due to the rational decoupling positions with eight possible
combinations, the Taipo decoupling method has being widely used
by many researchers (Ren and Hu, 2012). For example, Climent
and Pardo (2007) used the Taipo decoupling indicator to in-
vestigate the causal relationship between energy use and Spanish
economic growth. The Taipo decoupling indicator was utilized by
Freitas and Kaneko (2011) to study the occurrence of a decoupling
between Brazil's economic growth and energy-related CO2 emis-
sion over the period 2004–2009. That indicator was also used by
Zhang and Wang (2013) to study the decoupling status between
GDP and energy-related CO2 emission in Jiangsu province during
the period 1995–2009.
The other decoupling indicator is defined based on the de-
composition results received by IDA. Diakoulaki and Mandaraka
(2007) firstly defined a decoupling model based on the result of
the refined Laspeyres decomposition model, which was utilized to
assess the real efforts undertaken in each country and their ef-
fectiveness in dissociating the economic and environmental di-
mensions of development. Using the decomposition results of
LMDI method, Zhang and Wang (2013) developed a decoupling
indicator, which was used to study the decoupling of electricity
consumption from economic growth in China over 1991–2009.
That decoupling method was also used by Zhang and Guo (2013)
to evaluate the progress in decoupling energy consumption from
per capita annual net income of rural households. But this de-
coupling method only defines three kinds of decoupling, i.e, weak
decoupling, strong decoupling, and no decoupling.
As an important old industrial base in China, Liaoning province
has experienced spectacular economic growth with the im-
plementation of the development strategy for reviving northeast
old industrial base, which has led to more energy use. Nowadays,
the energy saving in Liaoning province has some problems, such as
large energy gap, low resource utilization efficiency. Thus, it is
necessary to explore the decoupling status between energy con-
sumption and economic growth in Liaoning Province. The purpose
of this paper is to explore the dependence of economic growth
(GDP) on energy consumption in Liaoning Province over the
period 1995–2012. The Tapio decoupling indicator is used to
complete that purpose. Then the generalized LMDI method pro-
vided by Wang et al. (2014) is used to find more affecting factors
governing decoupling status between energy consumption and
economic growth in Liaoning Province, which is the main con-
tribution to the literature.
The remainder of this paper is organized as follows. Section 2
presents the methodologies of the study and related data. The
main results are presented in Section 3. Finally, we conclude this
study.
2. Methodology and data
2.1. LMDI method
The final energy consumption in year t ( Et ) can be expressed as
the following formula:
Eijt Eit Gt
Et = ∑ × × it × Gt = ∑ ESijt × EIit × Sit × Gt
i, j
Eit Git
G i, j (1)
where t : the time in years; i : industrial sector; j : fuel type; Eit :
energy consumption of the i th industrial sector in year t; Eijt : en-
ergy consumption of the j th fuel type of i th industrial sector in
year t; Gt : the GDP in year t; Git : the GDP of the i th industrial sector
Eijt
in year t; ESijt = : the share of the j th energy form to total energy
Eit
Eit
consumption of the i th industrial sector in year t; EIit = : the
Git
Git
energy intensity of the i th industrial sector in year t; Sit = t: the
G
economic structure of the i th industrial sector in year t.
Nowadays, there exist a number of production functions in
literatures. This paper tries to study the impact of investment and
labour on energy consumption. However, the Cobb–Douglas (C-D)
production function was widely used to represent the technolo-
gical relationship between the amounts of two or more inputs,
particularly physical capital and labour, and the amount of output
that can be produced by those inputs (Wang et al., 2014). Thus, this
paper selects the C-D production function to describe GDP. So the
GDP can be expressed as the following Eq. (2)
Gt = A(K t )α (Lt ) β (2)
Where, A, α , β are unknown constant parameters. Commonly,
α > 0, β > 0. K : the fixed asset investment; L : the amount of labour
input. Substitute Eq. (2) for the GDP on the right of Eq. (1), then we
have the following Eq. (3)
Et = ∑ A × ESijt × EIit × Sit × (K t )α × (Lt ) β
i, j (3)
According to the LMDI method given by Ang (2004), the change
t
of energy consumption ( ΔEtot ) between a base year 0 and a target
year t can be decomposed into the following six factors: the en-
t
ergy mix effect ( ΔEes ); energy intensity effect ( ΔEeit ); economic
structure effect ( ΔEst ); labour effect ( ΔElt ), investment effect ( ΔElt ),
and constant term effect ( ΔEat ); as shown in the following formula.
t t
ΔEtot = ΔEes + ΔEeit + ΔEst + ΔEkt + ΔElt + ΔEat (4)
Each factor in the right hand side of Eq. (4) can be expressed as
follows:
416 B. Dong et al. / Energy Policy 97 (2016) 414–420

⎧0 if Oijt × Oij0 = 0 ; Table 1

⎪
⎪ The criteria of decoupling indicator.
t
ΔEes =⎨ Oijt − Oij0 ESijt
⎪∑ × LN ( 0 ) if Oijt × Oij0 ≠ 0. δE δG D Decoupling state
⎪ t 0
⎩ i, j LN (Oij ) − LN (Oij ) ESij (4a)
1 <0 >0 Dt < 0 Strong decoupling
2 >0 >0 0.8 ≥ Dt > 0 Weak decoupling
3 >0 >0 1.2 ≥ Dt > 0.8 Expansive coupling
⎧0 if Oijt × Oij0 = 0 ;
⎪
⎪
4 >0 >0 Dt > 1.2 Expansive negative
decoupling
ΔEeit = ⎨ Oijt − Oij0 EI t
⎪∑ 5 Dt < 0 Strong negative decoupling
× LN ( i 0 ) if Oijt × Oij0 ≠ 0. >0 <0
⎪ t 0
⎩ i, j LN (Oij ) − LN (Oij ) EOi 6 <0 <0 0.8 ≥ Dt > 0 Weak negative decoupling
(4b)
7 <0 <0 1.2 ≥ Dt > 0.8 Recessive coupling
8 <0 <0 Dt > 1.2 Recessive decoupling

⎧0 if Oijt × Oij0 = 0 ;
⎪
⎪
ΔEst = ⎨ Oijt − Oij0 St G0
⎪∑ × LN ( i0 ) if Oijt × Oij0 ≠ 0. Dt = × (E t−E 0)
⎪ t 0 E × (Gt −G 0)
0
⎩ i, j LN (Oij ) − LN (Oij ) Si (4c)
t
G0 ΔEes + ΔEeit + ΔEst + ΔEkt + ΔElt
= ×
E0
ΔGt
t
⎧0 if Oijt × Oij0 = 0 ; G0 ΔEes G0 ΔEeit G0 ΔEst G0 ΔEkt
⎪ = × + 0 × + 0 × + 0 ×
⎪ E0
ΔG t
E ΔG t
E ΔG t
E ΔGt
ΔEkt = ⎨ Oijt − Oij0 (K t )α 0 t
⎪∑ × LN ( 0 α ) if Oijt × Oij0 ≠ 0. G ΔEl t
⎪ t 0 + 0 × = Des + Deit + Dst + Dkt + Dlt
⎩ i, j LN (Oij ) − LN (Oij ) (K ) (4d) E ΔGt (6)
t t 0
Where, ΔG = G − G . According to the above Eq. (6), the decou-
pling indicator of energy consumption from GDP can be decom-
⎧0 if Oijt × Oij0 = 0 ;
⎪ posed into five decoupling indicator effects: energy mix decou-
⎪ t
pling effect ( Des ), energy intensity decoupling effect ( Deit ), economic
ΔElt = ⎨ Oijt − Oij0 (Lt ) β
⎪∑ × LN ( ) if Oijt × Oij0 ≠ 0. activity decoupling effect ( Dst ), labour decoupling effect ( Dlt ), and
⎪ t 0
⎩ i, j LN (Oij ) − LN (Oij ) (L0) β (4e) investment decoupling effect ( Dkt ).

2.3. Data description

⎧0 if Oijt × Oij0 = 0;
⎪
⎪ The research period in this paper starts in 1995 and ends in
ΔEat = ⎨ Oijt − Oij0 A
⎪∑ × LN ( ) if Oijt × Oij0 ≠ 0. 2012. All the energy data are collected from various issues of China
⎪ t 0
⎩ i, j LN (Oij ) − LN (Oij ) A Energy Statistical Yearbook (CESY, 1997, 2000, 2002–2014). The
(4f)
energy unit is standard coal consumption in 106 Mtce. This paper
only considers production energy consumption. Energy types used
Where, Oijt = A × ESijt × EIit × Sit × (K t )α × (Lt ) β . Because A is constant
in this paper includes electricity, heat, natural gas, oil products,
over the study period, ΔEat = 0. Thus only five factors are con- coal products, and other energy. Oil products include crude oil,
sidered in this paper. gasoline, kerosene, fuel oil, diesel oil, refinery gas, LPG, and other
Combining production function and LMDI presents a way to petroleum products. Coal products are composed of coal, other
extent the Kaya identity. However, there is the limitation of using washed coal, cleaned coal, coke, coke oven gas, briquettes, other
the C-D function in calculating the influencing factor. When cal- gas and other coking products.
culating the labour effect ( ΔElt ) and investment effect ( ΔElt ), there The unit of GDP is billion yuan in constant 1995 price, which is
exists a constant in formula (4d) and (4e), which may exaggerate collected from the Liaoning Statistical Yearbook (LNY, 2014). The
or abbreviate impact factor. whole economy of Liaoning Province is divided into three ag-
gregated industries. The primary industry refers to agriculture and
its related activities. The secondary industry sector only includes
2.2. Decoupling model industry and construction. The tertiary industry sector is com-
posed of transport, storage, post, wholesale, retail trade and hotel,
Following the definition presented by Tapio (2005), the de- restaurants. The labour input is measured by total number of
coupling index of energy consumption ( E ) from GDP ( G ), Dt , in a employed persons and its unit is 104 persons. The fixed asset in-
given period from a base year 0 to a target year t can be expressed vestment measured in billion yuan (1995 price) is the total in-
as the following formula: vestment in Liaoning Province.

Et − E 0
t E0
3. Results and discussion
D = t 0
G −G
G0 (5) 3.1. Analysis of energy consumption in Liaoning Province

According to the Tapio's definition, 8 kinds of decoupling state Since the start of economic reforms and opening-up in the late
t Et − E 0 t Gt − G 0 1970s, Liaoning Province has experienced spectacular economic
are listed in Table 1. where, δE = , δG = . So, the Eq. (5)
E0 G0 growth. Its GDP rose from 279.34 Billion Yuan in 1995 to 1696.32
can also be expressed as the following formula: Billion Yuan in 2012, representing an annual average growth rate
 B. Dong et al. / Energy Policy 97 (2016) 414–420
Fig. 1. Economic activity of Liaoning Province economy (GDP) and energy
consumption.
of 11.19%, as shown in Fig. 1. Along with the rapid economic de-
velopment, energy consumption increased from 60.82 Mtce in
1995–140.10 Mtce in 2012, representing an annual average growth
rate of 5.03%. Fig. 1 also shows that energy consumption in
Liaoning Province has undergone a rapid growth phase between
2001 and 2012. In 2012, energy consumption increased about
158.72% and that in 2001. The energy consumption of the sec-
ondary industry increased to 116.27 Mtce in 2012 from 56.52 Mtce
in 1995. As shown in Fig. 2, the share of the secondary industry in
total energy consumption declined to 82.9% in 2012 from 92.9% in
1995, which shows that energy consumption in Liaoning Province
is dominated by the secondary industry.
The tendency of energy structure of production energy con-
sumption in Liaoning Province over time is shown in Fig. 3. There
Fig. 2. The share of energy consumption by three industries over 1995–2012.
Fig. 3. The tendency of produce energy structure in Liaoning Province over 1995–
2012.
417
Fig. 4. The tendency of energy intensity in Liaoning Province over 1995–2012.
is a substitution between the increasing shares of the oil products
and electricity and a decreasing share of the coal products. The
share of coal products decreased from 59.05% in 1995 to 45.03% in
2012, whereas the share of oil products increased from 15.89% in
1995 to 31.69% in 2012. Among all energy types, oil products in-
creased the fastest. The electricity consumption increased to
18.33 Mtce in 2012 from 6.14 Mtce in 1995, and its share increased
to 13.08% from 10.67% during the study period. However, the share
of natural gas accounted for less than 5% over the study period.
Energy intensity for different industrial sectors is presented in
Fig. 4. The tendency of energy intensity for the secondary industry,
primary industry, and total economy declined in a wave-pattern
over the study period. However, the energy intensity for the ter-
tiary increased to 0.0324 kgce/Yuan in 2012 from 0.0299 Kgce/
Yuan in 1995. In terms of the annual decline rate, the secondary
industry is the fast. The energy intensity of the secondary industry
decreased from 0.406 Kgce/yuan in 1995–0.128 Kgce/yuan in 2012,
representing an annual change rate of 6.53%. The curve of energy
intensity of the secondary industry sector can be divided into two
phases: a rapid-decrease phase from 1995 to 2001, a steadily
wave-decrease phase between 2001 and 2012. Overall, the level of
energy intensity for the whole economy is similar with the sec-
ondary industry.
The changes of economic structure in Liaoning Province over
1995–2012 is presented in Fig. 5. The share of the primary industry
gradually decreased from 14.5% in 1995 to 8.7% in 2012. The share
of the secondary and tertiary industry increased in waved pattern
over the study period. During 1995–2012, the share of the sec-
ondary industry accounted for more than 47%. However, the share
of the secondary industry exceeded 52% since 2008, which shows
that Liaoning Province economy was mainly depend on the sec-
ondary industry. Fig. 2 also shows that the share of the tertiary
Fig. 5. The tendency of economic structure in Liaoning Province over 1995–2012.
418 B. Dong et al. / Energy Policy 97 (2016) 414–420

Table 2 continual increase of energy consumption over the period 1995–

Decomposition of produce energy consumption in Liaoning Province during 1995– 2012, as shown in Table 1. The accumulated effect is an increase of
2012, (Mtce).
9.168 Mtce. The declining trend of number of labour in 1997–1998
t t and 2002–2003 can explain why the labour effect played a positive
ΔEes ΔEei ΔEst ΔEkt ΔElt t
ΔEtot
impact on decreasing energy consumption in these four years, as
1995–1996 0.464 6.715 1.275 5.998 0.066 2.390 shown in Fig. 6.
1996–1997 4.215 5.538 0.056 6.339 1.097 3.863 Our study also shows that the economic structure effect in-
1997–1998 0.004 7.427 0.996 0.537 0.138 9.101 creased energy consumption in most years except 1996–1998,
1998–1999 0.004 6.931 0.169 0.393 0.518 5.855
2001–2002, 2004, 2009, and 2012. The accumulated effect is an
1999–2000 0.062 1.883 2.053 2.207 0.823 7.028
2000–2001 0.021 11.356 1.406 0.912 0.233 11.637 increase of 4.61 Mtce. Fig. 5 also shows that the share of the sec-
2001–2002 0.007 0.882 0.511 1.454 0.544 1.287 ondary industry decreased in 1996–1998, 2001–2002, 2004, 2009,
2002–2003 0.002 2.279 0.416 5.401 0.084 3.456 and 2012, which can explain why the economic structure effect
2003–2004 0.015 2.835 2.261 11.274 1.114 7.307
decreased energy consumption in this 8 years. Table 2 also in-
2004–2005 0.050 1.406 2.272 7.714 0.370 11.812
2005–2006 0.003 0.565 1.060 6.657 0.143 7.298 dicates that the energy structure effect played a very minor role in
2006–2007 0.003 4.111 0.699 15.194 1.071 12.856 impacting energy consumption, which only leaded to an accu-
2007–2008 0.031 14.921 3.627 38.683 0.432 27.852 mulated increase of 3.975 Mtce during the study period. The en-
2008–2009 0.003 0.303 0.532 14.519 2.043 13.312 ergy structure effect had a positive effect on the decrease of energy
2009–2010 0.016 7.777 2.970 15.345 1.010 11.565
2010–2011 0.066 9.640 0.913 35.311 1.359 28.009
consumption in five years: 1996, 1998, 1999, 2001 and 2009. As
2011–2012 0.001 3.539 2.534 6.143 1.848 1.919 shown in Fig. 3, the incline trend of the share of coal products can
1995–2012 3.975 79.766 4.610 143.970 9.168 81.957 explain why the energy structure effect decreased energy con-
sumption in this five year.
industry was less than 40% except 2001–2007.
3.3. Decoupling analysis
3.2. Decomposition analysis of produce energy consumption
According to the definition in Section 2.1, the decoupling in-
Based on the LMDI method in its additive form given in Section dicator between energy consumption and GDP for Liaoning Pro-
2.2, the decomposition results in produce sector of Liaoning Pro- vince over 1995–2012 is listed in Table 3. Because the δG is positive,
vince are presented in Table 2. The results show that the energy only four decoupling status appeared over the study period: ex-
intensity effect plays an important role in decreasing produce pansive negative decoupling, expansive coupling, weak decou-
energy consumption at the aggregate level. The aggregate change pling, and strong decoupling. Expansive negative decoupling only
of produce energy consumption contributed by energy intensity appeared in two years: 2000 and 2005, and the decoupling in-
effect is 79.766 Mtce. During the study period, the negative con- dicator were 1.876 and 1.500, respectively. The development in
tribution happened in three years: 2000, 2002, and 2005. The 2002, 2006, and 2009 presented expansive coupling. The energy
increase of energy intensity of the whole economy can explain consumption presented strong decoupling with GDP growth in
why the energy intensity played negative effect in these three four years: 1996, 1998, 1999, and 2001. Weak decoupling only
years, as shown in Fig. 4. appeared in the rest years.
As listed from Table 2, the investment effect is the critical factor Table 4 lists the decomposition results of decoupling indicator
in the growth of produce energy consumption in Liaoning Pro- for Liaoning Province over 1995–2012. The energy intensity de-
vince. The aggregate change of produce energy consumption coupling effect played a positive role in the development of de-
contributed by the investment effect is 143.97 Mtce. During the coupling except 2000, 2002, 2005, and 2009. However, the energy
study period, the investment effect makes the continual increase structure decoupling effect, economic structure decoupling effect,
of energy consumption except 1998 and 2009, as listed in Table 2. labour decoupling effect, and investment decoupling effect did not
As shown in Fig. 6, fixed assets investment in Liaoning Province play a positive role in the development of decoupling except
rose from 88.49 billion yuan in 1995 to 130.63 billion yuan in 2012, several years.
representing an annual average growth rate of 2.31%. Followed by Expansive negative decoupling only occurred in 2000 and
the investment effect, the labour effect is another critical factor in 2005, which can be explained by the effect of the energy intensity
the growth of produce energy consumption in Liaoning Province.
Except 1997–1998 and 2002–2003, the labour effect makes the Table 3
The decoupling state for Liaoning Province over 1995–2012.

δE δG D Decoupling state

1995–1996 0.038 0.086 0.439 Strong decoupling

1996–1997 0.003 0.089 0.037 Weak decoupling
1997–1998 0.049 0.083 0.595 Strong decoupling
1998–1999 0.039 0.082 0.477 Strong decoupling
1999–2000 0.167 0.089 1.876 Expansive negative decoupling
2000–2001 0.135 0.090 1.499 Strong decoupling
2001–2002 0.110 0.102 1.081 Expansive coupling
2002–2003 0.076 0.115 0.663 Weak decoupling
2003–2004 0.033 0.128 0.258 Weak decoupling
2004–2005 0.190 0.127 1.500 Expansive negative decoupling
2005–2006 0.149 0.142 1.052 Expansive coupling
2006–2007 0.106 0.150 0.705 Weak decoupling
2007–2008 0.020 0.134 0.152 Weak decoupling
2008–2009 0.123 0.131 0.938 Expansive coupling
2009–2010 0.095 0.142 0.667 Weak decoupling
2010–2011 0.051 0.122 0.416 Weak decoupling
Fig. 6. The tendency of investment and labour in Liaoning Province over 1995– 2011–2012 0.051 0.095 0.539 Weak decoupling
2012.
 B. Dong et al. / Energy Policy 97 (2016) 414–420 419

Table 4 decoupling. The economic structure decoupling effect and

Decomposition of decoupling indicator for Liaoning Province over 1995–2012. investment decoupling effect did not play a positive role in the
development of decoupling. However, the energy structure
Des Dei Ds Dk Dl
decoupling effect and labour decoupling effect played a re-
1995–1996 0.0851 1.2326 0.2341 1.1009 0.0121 lative small role in the development of decoupling.
1996–1997 0.0399 0.0525 0.0005 0.0600 0.0104
1997–1998 0.0003 0.4854 0.0651 0.0351 0.0090 According the main results received in this paper, the following
1998–1999 0.0003 0.5644 0.0138 0.0320 0.0422
policies should be considered: (1) Readjust and optimize the
1999–2000 0.0165 0.5027 0.5479 0.5892 0.2198
2000–2001 0.0027 1.4627 0.1811 0.1175 0.0300 economic structure. Since 2008, the share of the secondary in-
2001–2002 0.0061 0.7401 0.4288 1.2205 0.4571 dustry exceeded 52%. In the study period, the economic structure
2002–2003 0.0004 0.4371 0.0797 1.0363 0.0161 effect increased energy consumption in most years. Furthermore,
2003–2004 0.0005 0.1000 0.0797 0.3976 0.0393
Liaoning Province is China's important mining industry big pro-
2004–2005 0.0063 0.1786 0.2885 0.9795 0.0470
2005–2006 0.0004 0.0814 0.1528 0.9594 0.0206
vince and the heavy industry base. Thus, the government should
2006–2007 0.0001 0.2254 0.0383 0.8331 0.0587 accelerate changing the style of economic growth and advance
2007–2008 0.0002 0.0815 0.0198 0.2113 0.0024 new industrialization. (2) Adjust the internal structure of industry
2008–2009 0.0002 0.0213 0.0374 1.0226 0.1439 through investment. Our results shows that the investment effect
2009–2010 0.0009 0.4489 0.1714 0.8857 0.0583
2010–2011 0.0010 0.1432 0.0136 0.5246 0.0202
is the critical factor in the growth of produce energy consumption
2011–2012 0.0003 0.9936 0.7114 1.7247 0.5187 in Liaoning Province. In Liaoning, industrial economic growth is
still mainly rely on the growth of heavy industry. The development
of high energy consumption heavy industry will lead to more
decoupling, economic structure decoupling, and investment de- energy demand in the future. Therefore, the government should
coupling, as listed in Table 2. In 2002, 2006, and 2009, the oc- adjust industrial configuration and advance the process of in-
currence of expansive coupling is also attributed to the change of dustrialization of new pattern through investment. (3) Improve
the investment decoupling effect. The energy intensity decoupling the quality of the labour force to ensure a smooth transfer of la-
effect played a major role in the development of strong decoupling bour. The labour effect is another critical factor in the growth of
in 4 years: 1996, 1998, 1999, and 2001. The energy structure de- produce energy consumption in Liaoning Province. In order to
coupling effect and labour decoupling effect played a relative small meet the demand of the new type of industrialization to the la-
role in the development of decoupling over the study period 1995– bour force, the government should develop vocational training of
2012. different contents and levels so as to satisfy the various and high-
level needs of labour market. (4) Adjust and optimize energy
structure. The result indicates that the energy structure effect
4. Conclusions and policy implications played a very minor role in increasing energy consumption over
the study period. The share of natural gas accounted for less than
Along with the rapid economic development, Liaoning's energy 5% over the study period. Liaoning's government should make a
consumption has grown rapidly. The purpose of this paper was to plan to promote the development of natural gas, which will play a
investigate whether Liaoning economy was becoming less de- key role in the optimization of its energy structure. (5) Raise the
pendent on energy consumption. However, decoupling analysis is energy use efficiency and reduce the energy consumption. The
a best method to explore the relation between economic growth results show that the energy intensity effect plays an important
and energy consumption. Firstly, the generalized LMDI method is role in decreasing produce energy consumption at the aggregate
used to explore the driving forces governing production energy level. Compared with developed countries, there is much room for
consumption in Liaoning province. Then, the combination of Tapio improvement in energy efficiency. Thus the government should
decoupling indicator and generalized LMDI method is utilized to strengthen the utilization of new technologies and crafts by many
study what the decoupling status occurred in Liaoning Province financial subsidies and preferential policies and so on.
and why the decoupling status appeared. The main conclusions
drawn from the present study may be summarized as follows:
Acknowledgment
(1) Along with the rapid economic development, energy con-
sumption in Liaoning Province increased from 60.82 Mtce in
The authors gratefully acknowledge the Natural Science Foun-
1995–140.10 Mtce in 2012, representing an annual average
dation of China (71403266), China Postdoctoral Science Founda-
growth rate of 5.03%. Over the study period, the secondary
tion (2013M541752, 2015M580484, 2016T90517), Qing Lan Project
industry accounted for more than 82.9% of total energy con-
of Jiangsu Province, and the Fundamental Research Funds for the
sumption. In 2012, coal products, oil products, and electricity
Central Universities (2013W04).
accounted for 90% of total produce energy consumption.
(2) During the study period, the energy intensity for the sec-
ondary industry, primary industry, and total economy de-
References
clined. However, the energy intensity for the tertiary industry
increased over the study period.
Ang, B.W., 2004. Decomposition analysis for policymaking in energy: which is the
(3) The energy intensity effect plays an important role in de- preferred method? Energy Policy 32, 1131–1139.
creasing produce energy consumption over the study period. Ang, B.W., Liu, N., 2007. Handling zero values in the logarithmic mean Divisia index
However, the investment effect is the critical factor in the decomposition approach. Energy Policy 35, 238–246.
CESY (China) Energy Statistical Yearbook), 1997, 2000, 2002 2014. National Bureau
growth of produce energy consumption in Liaoning Province,
of Statistics of China, National Development and Reform Commission.
followed by the labour effect. Climent, F., Pardo, A., 2007. Decoupling factors on the energy-output linkage: The
(4) During the study period, only four decoupling status occurred: Spanish case. Energy Policy 35, 522–528.
expansive negative decoupling, expansive coupling, weak de- Diakoulaki, D., Mandaraka, M., 2007. Decomposition analysis for assessing the
progress in decoupling industrial growth from CO2 emissions in the EU man-
coupling, and strong decoupling. The energy intensity de- ufacturing sector. Energy Econ. 29, 636–664.
coupling effect played a positive role in the development of Freitas, L.C., Kaneko, S., 2011. Decomposing the decoupling of CO2 emissions and
420 B. Dong et al. / Energy Policy 97 (2016) 414–420
economic growth in Brazil. Ecol. Econ. 70, 1459–1469.
Juknys, R., 2003. Transition Period in Lithuania-do we move to sustainability. J.
Environ. Res. Eng. Manag. 4 (26), 4–9.
Lin, X., Polenske, K.R., 1995. Input–output anatomy of China’s energy use changes in
the 1980s. Econ. Syst. Res. 7 (1), 67–84.
LNY, 2014. Liaoning Statistical Yearbook. China Statistical Publishing House,, Beijing.
Ma, C.B., Stern, D.I., 2008. China’s changing energy intensity trend: a decomposition
analysis. Energy Econ. 30, 1037–1053.
OECD (Organization for Economic Co-operation and Development), 2010. Indicators
to Measure Decoupling of Environmental Pressure from Economic Growth.
Sustainable Development. SG/SD (2002) 1/Final. (Website): http://www.olis.
oecd.org/olis/2002doc.nsf/ LinkTo/sg-sd (2002) 1-final .
Ren, S.G., Hu, Z., 2012. Effects of decoupling of carbon dioxide emission by Chinese
nonferrous metals industry. Energy Policy 43, 407–414.
Su, B., Ang, B.W., 2012. Structural decomposition analysis applied to energy and
emissions: Some methodological developments. Energy Econ. 34, 177–188.
Tapio, P., 2005. Towards a theory of decoupling: degrees of decoupling in the EU
and the case of road traffic in Finland between 1970 and 2001. Transp. Policy
12, 137–151.
Von, W.E.U., 1989. Erdpolitik: Ökologische Realpolitik an der Schwelle zum Jahr-
hundert der Umwelt.
Wang, W.W., Liu, X., Zhang, M., Song, X.F., 2014. Using a new generalized LMDI
method to analyze China’s energy consumption. Energy 67, 617–622.
Zhang, M., Guo, F.Y., 2013. Analysis of rural residential commercial energy con-
sumption in China. Energy 52, 222–229.
Zhang, M., Liu, X., Wang, W.W., Zhou, M., 2013. Decomposition analysis of CO2
emissions from electricity generation in China. Energy Policy 52, 159–165.
Zhang, M., Wang, W.W., 2013. Decouple indicators on the CO2 emission-economic
growth linkage: the Jiangsu Province case. Ecol. Indic. 32, 239–244.
Zhang, Z.X., 2000. Decoupling China’s carbon emissions increase from economic
growth: an economic analysis and policy implications. World Dev. 28, 739–752.