Economy-Wide Energy Efficiency Using A Comprehensive Decomposition Method
Economy-Wide Energy Efficiency Using A Comprehensive Decomposition Method
Homepage: https://www.gjesm.net/
CASE STUDY
ARTICLE INFO A B ST R AC T
Article History:
This study compares the energy intensity performance in Indonesia to other South
Received 20 December 2019 East Asia countries such as Vietnam, Thailand, Singapore, The Philippines and
Revised 02 March 2020 Malaysia for the period from 1971 to 2016. For this goal, this research employs
Accepted 27 March 2020 a multiplicative Log Mean Divisia Index II method and Spatial-Temporal Index
Decomposition Analysis. The manufacturing sector and commercial sector played a
Keywords: key role in the region’s economic structures that accounted for around 60% to 80%
Association of Southeast Asian of the total economic output from 1971 to 2016. The contribution of manufacturing
Nations (ASEAN) sector increased quite significantly, from 8% in 1971 to a peak of around 31%
Economic in 2001, before it fell to 28% in 2016. On the other hand, the contribution of
Growth agriculture sector dropped from 49% in 1971 to approximately 17% in 2016. It is
Intensity demonstrated in this research that the aggregate trend of the changes of energy
Log Mean Divisia Index intensity in these countries in the past forty-five years has been decreasing. For
Indonesia, aggregate energy intensity rose steadily by an average of 3% per year
from 1971 to 1999, more than doubling over this period, while from 1999 to 2001
energy intensity fell by 1% per annum on average, falling by 17% overall in 2016.
Overall, in terms of structure and industry effects on aggregate energy intensity, all
these countries showed a shift in industry value added to more energy-intensive
industries which also offset by falling within-industry energy intensity. However, the
analysis shows that both element of this trend was most pronounced in Indonesia.
41 3 3
*Corresponding Author:
Email: [email protected]
Phone: +628111 78 6661
Fax: +628111 78 6661
Note: Discussion period for this manuscript open until October 1, 2020 on GJESM website at the “Show Article.
D. Setyawan
386
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
improvements (intensity/technology effects) and (1996) examined the carbon intensity of the power
structural changes (structural effects). By classifying sector in 12 Asian countries and found that fuel
the interaction amongst these two effects, this study changes were the main driver affecting changes
attempts to explain the main driving force behind in carbon intensity. Schleich et al. (2001) explored
recent energy efficiency improvements of each Germany’s emission reductions in 1990 and argued
ASEAN countries over time. In addition, this study also that population growth, GDP and energy supply mix
attempts to examine whether the decline in energy were the main drivers for the emission changes. Davis
intensity in one country is similar within the same et al., (2003) investigated carbon intensity in the USA
sector across other countries. This paper examines from 1996 to 2000 and suggested that the declining
the energy intensity performance in Indonesia carbon intensity was driven by the warming climate.
compared to other selected ASEAN countries from Bhattacharyya and Ussanarassamee (2004) observed
1970 to 2016. Although decomposition analysis has energy and carbon intensities in the manufacturing
been widely used in studies of energy-related issues sector in Thailand from 1981 to 2000 and argued
in many countries, there is a lack of research into the that energy intensity and industry structure were the
decomposition of energy in Indonesia, particularly main factors driving the changes in carbon intensity.
when comparing Indonesia to other countries. The By investigating 14 EU countries from 1990 to 2003,
selected ASEAN countries for analysis in this study Diakoulaki and Mandaraka (2007) examined carbon
are Indonesia, Malaysia, the Philippines, Singapore, emission changes and suggested five main factors:
Thailand, and Vietnam while Brunei Darussalam, energy intensity, output, fuel mix, structure and
Cambodia, Myanmar, and Lao PDR have not been utility mix. Besides LMDI method, other studies also
included due to lack of data. The aims of this study attempt to measure the determinants of energy
are essentially two. First, this paper aims at examining intensity by applying different methods, such as Data
the development of energy efficiency and discovering Envelopment Analysis (DEA) and Stochastic Frontier
energy intensity changes amongst selected ASEAN Analysis (SFA). The DEA method has been employed
countries. Second, this study captures both temporal in many studies to examine the relationship between
improvements within individual countries and spatial energy efficiency and economic performance which
changes amongst selected countries employing the formulates an index of Total-Factor Energy Efficiency
model developed by Ang et al. (2016). This study (Honma and Hu, 2008, 2013, 2014; Hu and Kao, 2007;
objective expects to provide a robust description of Hu and Wang, 2006; Xiaoli et al., 2014). In terms of
the overtime performance of the selected ASEAN SFA, there are only a limited number of studies
country since it incorporates the abovementioned employing this method to examine the impact of
spatial and temporal approaches within a single energy efficiency policy measures to energy efficiency
analysis structure. The study investigates the driving level (Adom et al., 2018; Filippini and Hunt, 2011,
forces affecting energy intensity performances 2012; Filippini et al., 2014). In LMDI method, there
of Indonesia to other south East Asia countries, are two effects that are conventionally measured
including, Vietnam, Thailand, Singapore, The in many energy intensities studies (Ang, 2004 and
Philippines and Malaysia for the period from 1971 to 2015), namely: 1) structural effect utilised to measure
2016. This study has been carried out in Melbourne, changes in energy consumption as a result of changes
Australia during 2018 – 2019. in the activity mix by sub-sectors. The structure effect
of the productive sectors will be used to measure
MATERIALS AND METHODS the changes in energy intensity that result from a
Many scholars employ the decomposition transition of economic sectors such as from less
analysis to measure the driving forces of energy energy-intensive to a more energy-intensive sector
intensity (Hasanbeigi et al., 2013; Ke et al., 2012; Su (for example, changes from agriculture sector to
and Ang, 2015; Wang et al., 2018; Wood, 2009). This industry sector). This effect can determine changes
method is also an effective tool to observe factors in the energy intensity resulting from changes in the
that influence changes in carbon emissions, which composition of economic sectors, and 2) intensity
also may provide analyses of the effects of related effect employed to measure changes in energy
policy measures (Ang, 2004). Shrestha and Timilsina intensity caused by the improvement of technical
387
Assessing economy-wide energy efficiency
388
elements:
specifies
the thestructural
performance reference effect,
analysis countryintensity
involves by n effect
using and(C1
a weighted
countries activity effect
toaverage
Cn) (Ang,
andof all be
will 2004
countriesandin
evaluated 2015;
all IEA,years
forthe
year 2018).
0 and of
However,
specified
year T, with as assumptions
study this study
period focused
in order
of k to on the
provide
sectors. Cua progress
neutral
will of energy
as theofefficiency
description
be taken the country
benchmark development,(Cu is itthe
performance.
region Asis
elements:
important
the structural
to
performance effect,involves
onlyanalysis
isolate intensity
the effect and
parameters
n countries on(C1activity
the effect
tostructural
Cn) and (Ang, be2004
effect
will andand 2015;
intensity
evaluated IEA, 2018).
effect
for year 0 and as
benchmark country). The ratio changes in energy intensity between two countries (Country 1
However,
discussed inasAngthisandstudy (2001)focused on the et progress ofSince
energy efficiency development,(Cu is itthe is
year
and T, with
Country 2/ C1 andLiu
assumptions C2) of and0, Ang
k sectors.
in year Cu al.
will(2003).
is indirectly be taken
constructed asthis study,
theEqs.
as 5, 6having
benchmark no preference
and 7.region in
importantthe
selecting
benchmark to only isolate
study
country). period
The ratio the parameters
andchanges on the
countryin reference,
energy structural
following
intensity effect etand
Ang two
between intensity
al.countries effect
(2016), (Country
this study as
1
discussed
specifies in
the Ang and
reference Liu (2001)
country and
by Ang
using
Globalet
a al. (2003).
weighted
J. Environ. Since
average
Sci.
and Country 2/ C1 and C2) in year 0, is indirectly constructed as Eqs. 5, 6 and 7. this
of
Manage., study,
all having
countries
6(3): 385-402, inno
allpreference
the
Summer years
2020 in
of
selecting the⁄ study
specified study periodperiod
in order andto country
provide areference, following ofAng
neutral description the et al. (2016),
country this study
performance. As
specifies the reference country by using a weighted average of all countries in all the (5)
years of
the performance ⁄ analysis
⁄ period involves n countries (C1 to Cn) and will be evaluated for year 0 and
The
specified structure
study and in intensity
order to effects
provide a are computed
neutral description
year T, with assumptions of k sectors. Cu will be taken as the benchmark region (Cu generate
of the countrychanges in energy
performance. As intensity. At the lowest of
(5)is the
as
the Eq.
benchmark6 andcountry).
performance ⁄ 7. analysis The involves
ratio n countries
changes in energy(C1 intensity
to Cn) and sectoral
will
between be two disaggregation
evaluated for year
countries 0 or
(Country and 1individual process, energy
The structure
year T, with 2/ and intensityof
assumptions effects are computed
k sectors. Cu will be as taken
Eq. 6 and 7. benchmark region (Cu is the
as as
the
and Country C1 and C2) in year 0, is indirectly constructed intensity
Eqs. 5, 6 and can7. simply be measured as the change in
benchmark country). The ratio changes in energy intensity between two countries (Country 1
The structure and intensity effects are computed as Eq. 6 and 7. energy consumed per unit of output. Nevertheless,
and Country 2/ C1 and C2) in year (∑0, is indirectly constructed ) as Eqs.
this 5,measurement
6 and 7. might not adequately explain the
⁄
(6) energy intensity at (5) the (6)aggregate level as there are
⁄ (∑ )
(∑ ) other driving forces, (5)
such as structural shifts, that
⁄ can lead to economic (6) changes in energy intensity
The structure and intensity effects (∑ are computed as Eq. 6 and 7.
(∑ )) (Shahiduzzaman and Alam, 2013). For instance, when
(7
The structure and intensity effects are computed as Eq. 6 and 7. the structure of the economy shifts over a period
(∑ ) of time from the less (7 )energy-intensive agriculture
(∑
(∑ ))
(7) sector to the more energy-intensive (6)
) manufacturing
(∑ )
(∑
(∑ )) sector, then energy intensity may increase without
Where, and any decline in energy efficiency. (6) These indicators of
∑ (∑ )∑
Where, E C1,0
E
Cu (∑ and ) energy efficiency have been developed and utilized
E E
C 2,0 Cu
L
E∑ E
k
,
C1,0
k
Cu L , k
∑
C 2,0
k
Cu for measuring, observing (7 and evaluating energy
For the ratio
Where, w difference
C1,0 −Cu
= of a country’s
(∑ w energy
and
C 2,0 −Cu
=
Eintensity
E
)E (C1) between
two time periods (year) 0
performance across countries. Energy efficiency is
k
E C1,0
E Cu k
E C 2,0 Cu
∑ L E , Eas Eqs.
and year T), is formulated
k k
8, 9 and 10.∑ L E , )E (7
(∑
k k
k C1,0 Cu
C 2,0 Cu
usually twoestimated with either physical or monetary
k
For the ratio difference of a country’s energy intensity (C1) between time periods (year ) 0
For the ratio difference(∑of a country’s energy
and year T), is formulated as Eqs. 8, 9 and 10.
⁄ between two time periods (year 0 and
) indicators. One of the monetary based indicators that
intensity (C1)
Where, and are mostly used to estimate(energy efficiency is energy
year T), is formulated
⁄ ∑ as Eqs. 8, 9 and 10. ∑ 8)
Where, ⁄ and intensity, where in this term ( the energy efficiency is
andC1,T ∑ ∑ measured by how much 8) energy is required or needed
For the I ⁄difference
ratio of a country’s energy intensity (C1) between two time periods (year 0
per unit of output (in the economic context, it is
I C1,year
and
T
T),
and is I Cu
formulated as Eqs. (∑8, 9 and 10. )
the=ratio
ForC1,0 C 1,0difference of a country’s energy intensity (C1) (8) between
measuredtwo time by periods
dividing(yearthe 0total energy consumption
I I (9)
and year T), is formulated
I Cu
as Eqs. (∑8, 9 and 10. ) with total GDP). Overall, the International Energy
⁄ (∑ ) (
Agency (IEA) defines energy (9) efficiency as an effort to
8)
And ⁄ (∑ control energy consumption ( and to attain economic
(∑ ))
and
⁄
growth. Data in this 8)study is gathered from two
databases: The United (10)Nations Statistics Division
and (∑
(∑ )
(∑ )) (UNSD) datasets and (10) International Energy Agency
(9) (9)
(∑
(∑
)
))
(IEA). The UNSD supports databases in international
(∑ industrial statistics. Although this database provides
Where, and (9)
∑ (∑ )∑ a substantial coverage of statistics across industries,
Where, (∑ and ) it only comprises output/activity data and does not
incorporate energy use(10)
∑ ∑
(∑ ) data. Both output and energy
(∑ 5 ) consumption data are (10)essential for this research to
(10)
(∑ 5 ) observe the level of energy efficiency development
Where, and across the selected countries. Therefore, this study
∑ ∑ employed energy use data sourcing from the IEA that
Where, and includes energy balance and energy consumption
E C1,T E Cu E C1,0 EkCu
L kC∑1,T , kCu L kC1,0 , ∑
E E E E Cu data across each sector of the economy. The GDP
Where, wkC1,T −Cu = and wkC1,0−Cu =
5 E C1,0
EkC1,T EkCu
∑ kL E C1,T , E Cu ∑ kL EkC1,0
E Cu
, kCu data in this study include the four main sectors of the
E
5 economy for the six ASEAN countries: manufacturing,
Changes to energy intensity are defined by many transportation, commercial and agriculture. The total
driving factors including selection of production energy consumption data is obtained from the energy
process, price of energy, technology advancement balances of OECD and non-OECD economies data sets.
and innovation, fuel mix and government policy As with many other cross-country studies measuring
measures (Sun, 1998). Shifting within energy-intensive GDP, this study utilises the Value Added by Economic
sectors or changes in the economic structure can also Activity, at constant 2010 prices in US dollars.
389
D. Setyawan
RESULTS AND DISCUSSION to increase, although, the growth rate was slower
Comparative analysis of the energy-intensity amongst compared to the early 1990s. All economic activity
ASEAN countries in the key end-use sectors of the ASEAN-6 economy
Fig. 1 shows the energy-intensity changes in the six increased at an annual average rate of greater than
ASEAN countries from 1971 to 2016. Overall, in ASEAN 4% from 2011 to 2016 (Table 1). The manufacturing,
countries, the Total Final Energy Consumption (TFC) transport and commercial sectors contributed the
per GDP or defined as Energy Intensity, fluctuated most to growth in the total industrial value-added.
during the period of this study. Indonesia’s energy While economic activity increased over the period of
intensity was average compared to the other ASEAN study in all sectors, rates of economic growth varied
countries, where it started increasing moderately and across the sectors. Annual growth in activity in the
peaked in 1999, followed by a decreasing trend of manufacturing, transport and commercial sectors
aggregate energy intensity for the rest of the period. was considerably greater than the rate of growth in
the agriculture sector. The decline in the share of
Economic Structure (sectoral share to GDP, 1971– total economic activity of the agriculture sector and
2016) the increasing shares of the manufacturing, transport
Indonesia’s industry has grown significantly over and commerce sectors indicate a change in the
the past forty-five years. The total value of industrial structure of ASEAN-6 economies over the period of
output increased from US$61.6 billion in 1971 to study.
US$801.9 billion in 2016 at 2010 prices, with an overall Table 1 summarises ASEAN countries’ value-
annual average growth rate of 8.95%. In the 1970s to added shares and annual growth rate of value-added
the beginning of 1990s, industrial growth averaged from 1971 to 2016.
approximately 7% per year. This phenomenal period Indonesia shifted from a predominately agricultural
of growth was followed by a short period during the economy to a predominately manufacturing economy
economic crisis from 1997 to 1998, where industrial in quite short time period (Jacob, 2005). Since the
growth dropped sharply by negative 1% from 1997 to 1990s, the domination of manufacturing sector has
2001. The commercial sector suffered a severe drop exceeded the contribution of the agriculture sector.
of negative 2% growth annually from 1997 to 2001. The contribution of manufacturing sector increased
However, after the 2000s, industrial growth started quite significantly, from 8% in 1971 to a peak of
390
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
Table 1: Gross value-added shares and annual growth rate of ASEAN-6
Table 1: Gross value-added shares and annual growth rate of ASEAN-6
around 31% in 2001, before it fell to 28% in 2016. On to a mainly manufacturing economy demonstrates a
the other hand, the contribution of agriculture sector significant development in its economic policies. This
dropped from 49% in 1971 to approximately 17% in structural shift is the key indicator of the successful
2016. The greatest contribution in the economy came transition to industrialisation in Indonesia.
from the commercial sector, which was relatively
steady at around 40% on average during the study Composition of total energy consumption by sector
period. Meanwhile, the transportation sector had Total energy consumption in ASEAN-6 accounted
the smallest contributor to Indonesia’s economy. The for around 276 million tonnes of oil equivalent (MToE)
structural changes in Indonesia from an agricultural in 2016. In 2016 (Table 2), total energy consumption
391
Assessing economy-wide energy efficiency
Table 2. Energy Consumption shares and annual growth rate of ASEAN-6
Table 2. Energy Consumption shares and annual growth rate of ASEAN-6
in the manufacturing and transportation sectors Energy consumption in Indonesia increased from
accounted for more than 80% of all consumption 4,648 KTOE in 1971 to 92,946 KTOE in 2016, with
in ASEAN-6 countries. The trend of total energy an average annual growth rate of 7.18%. For the
consumption from 1971 to 2016 clearly demonstrates same period, total energy consumption in both
the dominance of these two sectors’ share compared manufacturing and transportation sector also
to the remaining sectors. The increase in energy increased from 1,625 KTOE and 2,692 KTOE in 1971
consumption from manufacturing and transportation to 37,732 and 47,249 KTOE in 2016 with an average
is potentially due to advances in technology, while annual growth rate of 8.32% and 6.74%, respectively.
the increase in the commercial sector is a result of Indonesia’s manufacturing and transportation
increasing incomes due to economic development. sector were responsible for the largest share of the
392
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
Energy Intensity
Country Sectors (KTOE/ million 2010 USD prices)
1971 1981 1991 1997 2001 2011 2016
Manufacturing 0.32 0.38 0.29 0.25 0.27 0.22 0.17
Transportation 1.56 1.18 1.05 1.12 1.24 0.62 0.51
Indonesia Commercial 0.00 0.01 0.01 0.01 0.02 0.02 0.02
Agriculture 0.01 0.01 0.02 0.02 0.04 0.02 0.02
Aggregate EI 0.08 0.12 0.13 0.14 0.16 0.13 0.12
Manufacturing 1.41 1.95 1.63 1.20 1.00 1.06 1.00
Transportation 2.13 0.79 1.14 1.63 1.70 2.11 1.89
Vietnam Commercial 0.00 0.02 0.03 0.04 0.05 0.05 0.05
Agriculture 0.00 0.02 0.02 0.03 0.03 0.03 0.03
Aggregate EI 0.28 0.31 0.26 0.27 0.28 0.36 0.37
Manufacturing 0.48 0.28 0.24 0.28 0.26 0.25 0.27
Transportation 1.19 1.17 1.26 1.41 1.01 0.85 0.78
Thailand Commercial 0.01 0.01 0.02 0.03 0.03 0.04 0.03
Agriculture 0.07 0.07 0.09 0.10 0.10 0.10 0.08
Aggregate EI 0.16 0.13 0.16 0.20 0.18 0.18 0.17
Manufacturing 0.07 0.07 0.04 0.04 0.09 0.11 0.12
Transportation 0.48 0.26 0.19 0.12 0.10 0.09 0.07
Singapore Commercial 0.01 0.01 0.02 0.01 0.01 0.01 0.01
Agriculture 0.00 0.01 0.01 0.02 0.03 0.00 0.00
Aggregate EI 0.05 0.06 0.05 0.03 0.04 0.04 0.04
Manufacturing 0.17 0.15 0.16 0.20 0.15 0.14 0.11
Transportation 1.70 1.01 1.10 1.59 1.18 0.59 0.62
Philippines Commercial 0.01 0.02 0.02 0.03 0.03 0.03 0.03
Agriculture 0.04 0.03 0.02 0.02 0.02 0.01 0.02
Aggregate EI 0.14 0.11 0.11 0.16 0.13 0.09 0.09
Manufacturing 0.91 0.49 0.33 0.28 0.30 0.23 0.21
Transportation 1.76 0.87 0.96 0.87 0.93 0.66 0.66
Malaysia Commercial 0.02 0.02 0.03 0.03 0.04 0.04 0.03
Agriculture 0.00 0.00 0.01 0.03 0.01 0.03 0.01
Aggregate EI 0.21 0.16 0.17 0.19 0.21 0.15 0.15
country’s end-use energy demand accounting for 2 shows the summary of ASEAN-6 countries’ energy
around 80 per cent (Table 2). Both the manufacturing consumption shares and an annual growth rate of
and transportation sectors are the highest energy- energy consumption from 1971 to 2016.
consuming sectors in Indonesia across the study Like other developing countries worldwide, the
period, where both sectors share approximately large transportation energy consumption in these
90% of energy use across all sectors in 2016. Table ASEAN-6 countries is mainly due to rapid population
393
D. Setyawan
growth and urbanisation. Here the growing numbers manufacturing and transportation sector in Malaysia
of vehicles, increased urbanisation and traffic in 2016 dropped to one-fourth and less than a half,
congestion were the major causes of high energy respectively, compared to its level in 1971.
consumption in the transportation sector. This
tendency has also been apparent in all other ASEAN Sub-periods decomposition
countries (ACE, 2017). This study examined ASEAN-6 energy intensity
into sub-periods decomposition analysis. The analysis
Energy intensity was conducted for seven sub-periods: 1971 to 1981,
Energy intensity reflects not only how much energy 1981 to 1991, 1991 to 1996, 1996 to 1998, 1998 to
is utilised in the economy but also the changes in 2001, 2001 to 2011 and 2011 to 2016 (Fig. 2). These
energy consumption across sectors. Aggregate energy sub-periods analyses aim to provide information
Intensity in Indonesia has fluctuated moderately in a about the changes that occurred in the ASEAN-6
range of 0.08 to 0.16 oil equivalent per million dollars economy during the study period; specifically,
(constant 2010 $US prices) from 1971 to 2016 (see before and after the Financial Crisis. Indonesia had
Table 3). The highest energy intensity in Indonesia a significant structural change throughout the first
came from the transportation sector with 1.56 in three sub-periods (1971 to 1981, 1981 to 1991 and
1971 that decreased to 0.51 in 2016. Following the 1991 to 1996), where the structural effect surged
transportation sector, the manufacturing sector was the aggregate energy intensity. During the period of
second highest at 0.32 in 1971 and declined to 0.17 in crisis (1996 to 1998 and 1998 to 2001), the energy
2016. While the commercial and agriculture sectors intensity increased by 14% and 7%, respectively. Both
had the smallest energy intensity ranging from 0 to energy intensity increases were due to the increase
0.04 from 1971 to 2016. Table 3 shows the summary in intensity effect. However, in the last two periods
of ASEAN-6 countries energy intensity from 1971 to of 2001 to 2011 and 2011 to 2016, the aggregate
2016. energy intensity in Indonesia fell quite significantly
From 1997 to 2016, the intensity of the two to around 19% and 13%, respectively, compared to
major energy-using sectors (manufacturing and the base years of 2011 and 2016. Indonesia’s energy
transport) in Indonesia dropped significantly, intensity trend confirmed the previous result of
although the aggregate energy intensity had not Voigt et al. (2014) which showed that the structural
declined substantially. In 1997, the transportation effect peaked from 1999 to 2000 and decreased
and manufacturing sector energy intensity decreased thereafter, shaping the overall trend of Indonesia’s
from 1.12 and 0.25 to 0.51 and 0.17, respectively. energy intensity. Similarly, to Indonesia, Malaysia
However, the aggregate energy intensity only reduced also had a significant structural change in the first
from 0.14 in 1997 to 0.12 in 2016. The reason for three sub-periods (1971 to 1981, 1981 to 1991 and
this trend is that the level of the energy intensity 1991 to 1996). During the period of crisis, aggregate
of transport sector is three to four times that of energy intensity in Malaysia increased by 1% and 8%,
the manufacturing sector, therefore the big shift respectively. However, in the last two sub-periods
to transport in output has increased the aggregate of 2001 to 2011 and 2011 to 2016, energy intensity
energy intensity, despite of the decline in the intensity declined by around 25% and 2%, respectively.
of transport sector. Overall, there was a significant Singapore had the lowest aggregate energy
drop in the manufacturing and transportation sectors intensity during the period 1981 to 1991 and 1991 to
aggregate energy intensity in most ASEAN-6 countries. 1996 due to the high intensity effect. In Singapore,
For instance, in Indonesia, the manufacturing sector the role of structural effect was not as significant as
energy intensity in 2016 declined by 50% compared intensity effect across the sub-periods analysis except
to the 1971 level. Indeed, the transportation energy in 1971 to 1981, where the structural effect brought
intensity of Indonesia decreased even further up energy intensity to 147% of that prevailed in 1971,
compared to the manufacturing sector as in 2016 it while the intensity effect led to a fall in aggregate
dropped to one third compared to its level in 1971. energy intensity to 70%. During the period of crisis
These significant decreases not only occurred in (1996 to 1998 and 1998 to 2001) energy intensity
Indonesia but also in Malaysia. Energy intensity in the increased by 1% and 25%, respectively, compared
394
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
(D)
(A)
(B) (E)
(C) (F)
395
D. Setyawan
to the 1996 and 1998 base levels. However, in individual country in the ASEAN-6 by employing a
the last period of 2011 to 2016, energy intensity Spatial-Temporal Index Decomposition Analysis (ST-
decreased by 11% compared to the 2011 base level. IDA). The purpose of the spatial-temporal graph
Thailand’s energy intensity fell by 18% from 1971 (as seen in Fig. 3) is to show a better picture of
to 1981, but the next two periods — 1981 to 1991 each ASEAN-6 performance across countries over
and 1991 to 1996 — the energy intensity increased a given period. It (Fig. 3) describes a multi-country
by around 20% to 23%. Both increases in energy performance and captures how the structural
intensity were due to the increase in the structural effect and the intensity effect changes over time
effect and intensity effect. However, during the amongst each ASEAN-6 country. As this subsection
crisis period 1996 to 1998 and 1998 to 2001, the aims to provide the performance and interpret the
energy intensity declined by 3% to 4%, due to the evolution of intensity effect and structural effect of
decreasing intensity effect. The decrease in energy each ASEAN-6 countries before and after the 1997
intensity also occurred until the end of sub-periods economic crisis, thus it analyses the variations of the
of (2001 to 2011 and 2011 to 2016). In the beginning aggregate energy intensity for all ASEAN-6 countries
period 1971 to 1981, Vietnam’s energy intensity for three consecutive years including 1973, 1997 and
increased by 9% compared to the 1971 base level, in 2016. The Financial Crisis was considered as part
due to an increase in the intensity effect. But in the of the study in order to interpret the evolution of
period 1981 to 1991, the energy intensity decreased energy intensity at an economy level in all ASEAN-6
by 15% compared to the 1981 base level. During countries during the study period. Following the
the crisis period of (1996-1998 and 1998-2001), ST-IDA approach developed by Ang et al., (2016),
energy intensity increased to around 2% as a result the reference region is constructed based on the
of an increase in the structural effect. The energy weighted averages of the energy intensity and
intensity peaked in the period 2001 to 2011 for value-added shares of the ASEAN-6 countries for
around a 31% increase compared to the 2001 base the selected three-year period. The reference region
level, due to the increase in both structural and serves as a benchmark for ASEAN-6 countries. As
intensity effect. However, at the end of the 2011 to shown in Fig. 3, the changes in aggregate energy
2016 period, energy intensity decreased about 2% intensity are described by the moves of points and
compared to the 2011 base level; this occurred due arrows along with the plot diagram that depicts the
to a decrease in the intensity effect. Energy intensity changes in intensity effect and structure effect. The
in the Philippines from 1971 to 1981 decreased results of decomposition for every country for 1971,
by around 22% compared to the 1971 base level, 1997 and 2016 are indicated by dots that are joined
where this due to a decrease in intensity effect. by arrows. For the reference country, the production
From 1981 to 1991, energy intensity only increased share values, and energy intensity are calculated
by 2%, however, from 1991 to 1996, energy intensity employing the weighted averages of the six ASEAN
surged by around 35% compared to the 1991 base countries over the three years period. The origin
level. This was due to an increase in the intensity value of (1,1) describes the reference region with an
effect. During the Financial Crisis, energy intensity aggregate energy intensity of 0.14 KToE per million
only increased by 3% compared to the 1996 base US dollars (KToE/106 USD) at constant 2010 prices
level. In the last three periods: 1998 to 2001, 2001 to in US dollars. This study utilises a multiplicative
2011 and 2011 to 2016, energy intensity decreased decomposition, therefore, the intensity and
by 13%, 31% and 2%, respectively, due to the fall structure effects are conveyed in ratio estimations,
intensity effect. where the measures in the y-axis and x-axis imply
the ratio change from the benchmark country. Thus,
Comparing regional aggregate energy intensity a value of 1.1 means that the effect is determined
The aim of this sub-section is to measure to be 10% greater as compared to the benchmark
regional disparities that can simultaneously capture country, while on the other hand, a value of 0.9
both temporal changes and spatial differences implies that the effect is 10% lower as considered to
in energy efficiency developments within each the reference country.
396
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
Fig. 3 can be divided into four quadrants, namely along the horizontal axis over time. Almost all
Quadrant I, II, III and IV. From this Figure, a positive ASEAN-6 countries have shifted from Quadrant IV to
improvement in either intensity or structural effects another Quadrant, except Vietnam. Vietnam stayed
are represented by the country moving towards a in Quadrant IV, showing its economy comprises of
smaller value along an axis over a period of time, less intensive industry and less efficient in energy
which is related to the decreasing of aggregate usage compared to the rest of ASEAN-6 countries.
energy intensity over time. On the other hand, a Last, Quadrant II is in the bottom right corner. This
country shifting towards a greater value along an describes the countries with a higher structure
axis indicates a negative development, which effect, but less intensity effect compared to the
contributes to an increase in aggregate energy ASEAN-6 average. There are two countries located
intensity over time. Quadrant III is set in the bottom in this quadrant: Indonesia (in 2016) and Singapore
left corner and represents the best quadrant (the (in 1997 and 2016). It is interesting to note that
most energy efficient countries), where the Singapore went through less intensive industry (in
countries in this quadrant represent fewer intensive 1971) to more intensive industry (in 1997), but in
industries and more efficient energy usage compared 2016, Singapore was approaching Quadrant III
to the benchmark country. During the three-year showing industry in Singapore become less intensive
study period there are only two countries that are and more energy efficient. From Fig. 3, all of the
located in Quadrant III: Singapore (in 1971) and the ASEAN-6 countries have improved their aggregate
Philippines (in 2018). The aggregate energy intensity energy intensity (AEI) over time. The AEI of the
in these countries showed a lower than the ASEAN countries influenced by the structural and
aggregate energy intensity of the reference country. intensity factors. The graph showed that the
On the other hand, Quadrant I is located in the top developing countries develop their economies from
right corner which represents the worst quadrant less intensive energy sectors to more intensive
(or the least energy efficient nations). The countries energy sectors, for instance from agriculture to the
in this quadrant have a higher structure effect manufacturing sector. The structural effects have
(having a more intensive industry in the economic improved over time; however, the magnitude is not
structure) and a higher intensity effect (using less as great as the intensity factors. This situation
efficient energy usage technology) compared to the implies that the government economic policies are
reference country. There are two countries significant in determining the result of the overall
positioned in Quadrant I: Malaysia (in 1997) and energy efficiency over time in the ASEAN-6. The
Thailand (in 2016). Additionally, Quadrant IV is overall industrialisation process of the ASEAN-6
positioned in the top left which depicts those countries can be observed from the changes in its
countries with less intensive industries and less economic structure in all the ASEAN-6 economies
efficient energy usage compared to the reference over the 45 years of analysis. Four trends can be
country. Most of the ASEAN-6 countries had some observed from the ASEAN-6 economy in terms of
time of aggregate energy intensity in this quadrant, structural changes (Table 1). First, the commercial
except Singapore, since Singapore was more sector played a substantial role in the overall
developed compared to the rest of the ASEAN-6 economic structures that accounted for more than
countries. This situation is consistent with the fact 40% of the aggregate economic structure of the
that most of the ASEAN-6 countries are developing ASEAN-6 from 1971 to 2016. Second, the role of the
countries; therefore, their economy uses less energy manufacturing sector has increased almost two-fold
(using less energy-intensive sectors of the economy) in the ASEAN-6 countries, except for Singapore and
in the beginning of their economic development. the Philippines, whereas in Indonesia the share
However, along with the advancement of its from this sector has increased almost four-fold.
economy, their economic structure shifting from Third, the declining share of the agricultural sector
less energy-intensive to a more energy-intensive to the total economic output of most of the ASEAN-6
sector economy over time. The shifting of economic countries, except Singapore. The share of the
structure can be seen from the increasing value agriculture sector was prominent at the beginning
397
D. Setyawan
398
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
because of structural shifts in the economy and new 17% overall. While this change in the trend after the
investment in the economic sector. From 2000 to Financial Crisis is also evident in the three countries
2016, the aggregate value-added was doubled, total mentioned, it is most pronounced in Indonesia and
energy consumption rose but less than the aggregate becomes a central theme of this analysis. By 2016
value-added growth, noting that the energy intensity Indonesia’s aggregate energy intensity was towards
improved (or decreased). Other ASEAN countries the bottom of the range of the five larger developing
also experienced a similar trend, whereas the energy members of ASEAN-6, is below that of Thailand and
intensity has improved over time. The improvement Malaysia and well below that of Vietnam, but above
in intensity effect in all ASEAN-6 countries has that of the Philippines. Each of these countries has
become the primary driving force for the aggregate an energy intensity much higher than Singapore.
energy intensity to decline, while the magnitude of As observed in the analysis, the main energy
the structural effect is limited. Based on the energy consumer in all ASEAN-6 countries comes from the
intensity patterns of ASEAN-6, Indonesia and manufacturing and transportation sector, where
Malaysia demonstrated a greater impact on these two sectors account for more than 70% of
structural changes than other ASEAN countries. The total national energy usage. In Indonesia these two
structural changes had a more significant impact on sectors consumed more than 80% of total energy
increasing the aggregate energy intensity in consumption during the study period. In addition
Indonesia and Malaysia than in the other ASEAN-6 to its consumption share, both manufacturing
countries. In other words, this shows that these and transportation energy intensity in Indonesia
countries made substantial industrialisation process had dropped significantly compared to its level in
by moving from less energy-consuming sectors such 1971. In 2016, the manufacturing energy intensity
as agriculture to more energy-intensive sectors such declined to a half and the transportation decreased
as manufacturing. up to one-third compared to its level in 1971.
From the above findings, it can be observed that
CONCLUSION Indonesia experienced a substantial structural shift
in its economy from less energy-consuming sectors
This study decomposed the changes in aggregate (that is, agriculture) to more energy-intensive
energy intensity in the ASEAN-6 countries for sectors (that is, manufacturing). In terms of sectoral
the period 1971 to 2016. For this goal, this study energy intensity reduction, the manufacturing and
employed a multiplicative LMDI-II method and ST- transportation sectors were the two highest sectors
IDA. This study demonstrated that the aggregate that drove the aggregate energy intensity to decline.
trend of the changes of ASEAN energy intensity The significant drop in both transportation and
fluctuated moderately for all ASEAN-6 countries, manufacturing sector energy intensity needs to be
where variations in energy intensity appeared further investigated.
in these countries. As an overall trend, energy
intensity in Singapore was the lowest and markedly AUTHOR CONTRIBUTIONS
stable during the study period, while Vietnam
had the highest and more fluctuating energy D. Setyawan performed all of the process in
intensity. Indonesia, Malaysia, the Philippines and developing this paper, including, compiled the
Thailand have more moderate energy intensity data, interpreted the data, the literature review,
trend compare to other ASEAN-6 countries. Two experimental design, analysed, prepared the
distinct periods are evident in terms of trends in manuscript text, manuscript edition, and manuscript
aggregate energy intensity in Indonesia, and in finalisation for journal publication.
some other ASEAN countries (Thailand, Malaysia
and the Philippines). For Indonesia, aggregate ACKNOWLEDGMENT
energy intensity rose steadily by an average of 3% The author would like to thank to Professor
per year from 1971 to 1999, more than doubling Peter Sheehan and Dr. Steven Parker in the Victoria
over this period, while from 1999 to 2001 energy University, Melbourne, Australia for their valuable
intensity fell by 1% per annum on average, falling by insights and contributions to the study. The author
399
D. Setyawan
also wishes to express gratitude to the Fiscal Policy LMDI-II Logarithmic Mean Divisia Index II
Agency, Ministry of Finance of the Republic of
Indonesia for providing facilities and to the Australian MBoE Million Barrels of Oil Equivalent
Awards Scholarship for providing funding that have MToE Million tonnes of oil equivalent
made this research endeavour possible. Organisation for Economic Co-
OECD
operation and Development
CONFLICT OF INTEREST SFA Stochastic Frontier Analysis
The authors declare that there is no conflict of Spatial-Temporal Index
interest regarding the publication of this manuscript. ST-IDA
Decomposition Analysis
In addition, the ethical issues, including plagiarism,
TFC Total Final Energy Consumption
informed consent, misconduct, data fabrication and/
or falsification, double publication and/or submission, TFEE Total-Factor Energy Efficiency
and redundancy have been completely observed by ToE Tons of Oil Equivalent
the authors. UNSD United Nations Statistics Division
400
Global J. Environ. Sci. Manage., 6(3): 385-402, Summer 2020
32(9): 1131-1139 (9 pages). IMF, (2017). Seeking sustainable growth short-term recovery,
Ang, B.W., (2015). LMDI decomposition approach: A guide for long-term challenges. International Monetary Fund,
implementation. Energy Policy. 86: 233-238 (6 pages). Washington, DC.
Ang, B.W.; Su, B.; Wang, H., (2016). A spatial–temporal Jacob, J., (2005). Late industrialization and structural change:
decomposition approach to performance assessment in Indonesia, 1975–2000. Oxford Develop. Stud., 33(3-4):
energy and emissions. Energy Econ., 60: 112-121 (10 pages). 427-451 (25 pages).
Bhattacharyya, S.C.; Ussanarassamee, A., (2004). Ke, J.; Price, L.; Ohshita, S.; Fridley, D.; Khanna, N.Z.; Zhou, N.;
Decomposition of energy and CO2 intensities of Thai Levine, M., (2012). China’s industrial energy consumption
industry between 1981 and 2000. Energy Econ., 26(5): trends and impacts of the Top-1000 Enterprises Energy-
765-81 (17 pages). Saving Program and the Ten Key Energy-Saving Projects.
Davis, W.B.; Sanstad, A.H.; Koomey, J.G., (2003). Contributions Energy Policy. 50: 562-569 (8 pages).
of weather and fuel mix to recent declines in US energy and OECD, (2019). Economic Outlook for Southeast Asia, China
carbon intensity. Energy Econ., 25(4): 375-96 (22 pages). and India 2019: Towards Smart Urban Transportation, The
Diakoulaki, D.; Mandaraka, M., (2007). Decomposition Organisation for Economic Co-operation and Development.
analysis for assessing the progress in decoupling industrial Ozturk, I., (2010). A literature survey on energy–growth
growth from CO2 emissions in the EU manufacturing nexus. Energy Policy. 38(1): 340-349 (10 pages).
sector. Energy Econ., 29(4): 636-64 (29 pages). Schleich, J.; Eichhammer, W.; Boede, U.; Gagelmann,
Filippini, M.; Hunt, L.C., (2011). Energy demand and energy F.; Jochem, E.; Schlomann, B.; Ziesing, H.J., (2001).
efficiency in the OECD Countries: A stochastic demand Greenhouse gas reductions in Germany—lucky strike or
frontier approach. Energy J., 32(2): 59-80 (22 pages). hard work? Climate Policy. 1(3): 363-380 (18 pages).
Filippini, M.; and Hunt, L.C., (2012). US residential energy Shi, X., (2016). The future of ASEAN energy mix: A SWOT
demand and energy efficiency: A stochastic demand analysis. Renewable and Sustainable Energy Rev., 53: 672-
frontier approach. Energy Econ., 34(5): 1484-1491 (8 680 (9 pages).
pages). Shahiduzzaman, M.; Alam, K., (2013). Changes in energy
Filippini, M.; Hunt, L.C.; Zoric, J., (2014). Impact of energy efficiency in Australia: A decomposition of aggregate
policy instruments on the estimated level of underlying energy intensity using logarithmic mean Divisia approach.
energy efficiency in the EU residential sector. Energy Policy. Energy Policy. 56: 341-51 (11 pages).
69: 73-81 (9 pages). Shrestha, R.M.; Timilsina, G.R., (1996). Factors affecting CO 2
Hasanbeigi, A.; Price, L.; Fino C.C.; Lu, H.; Ke, J., (2013). intensities of power sector in Asia: A Divisia decomposition
Retrospective and prospective decomposition analysis of analysis. Energy Econ., 18(4): 283-293 (11 pages).
Chinese manufacturing energy use and policy implications. Su, B.; Ang, B.W., (2015). Multiplicative decomposition of
Energy Policy. 63: 456-466 (11 pages). aggregate carbon intensity change using input–output
Honma, S.; Hu, J.L., (2008). Total-factor energy efficiency of analysis. Appl. Energy. 154: 13-20 (8 pages).
regions in Japan. Energy Policy. 36 (2): 821-833 (13 pages). Sun, J.W., (1998). Changes in energy consumption and energy
Honma, S.; Hu, J.L., (2013). Total-factor energy efficiency intensity: A complete decomposition model. Energy Econ.,
for sectors in Japan. Energy Sources, Part B: Econ. Plann. 20(1): 85-100 (16 pages).
Policy. 8(2): 130-136 (7 pages). Voigt, S.; De C.E.; Schymura, M.; Verdolini, E., (2014). Energy
Honma, S.; Hu, J.L., (2014). Industry-level total-factor energy intensity developments in 40 major economies: Structural
efficiency in developed countries: A Japan-centered change or technology improvement?. Energy Econ., 41: 47-
analysis. Appl. Energy., 119: 67-78 (12 pages). 62 (16 pages).
Hu, J.L.; Kao, C.H., (2007). Efficient energy-saving targets for Wang, J.; Hu, M.; Rodrigues, J.F.D., (2018). An empirical
APEC economies. Energy Policy. 35(1): 373-382 (10 pages). spatiotemporal decomposition analysis of carbon intensity
Hu, J.L.; Wang, S.C., (2006). Total-factor energy efficiency of in China’s industrial sector. J. Cleaner Prod., 195: 133-44
regions in China. Energy Policy. 34: 3206-3217 (12 pages). (12 pages).
IEA, (2013). Southeast Asia Energy Outlook, International Wood, R., (2009). Structural decomposition analysis of
Energy Agency. Australia’s greenhouse gas emissions. Energy Policy.
IEA, (2015). Southeast Asia Energy Outlook 2015, Directorate 37(11): 4943-4948 (6 pages).
of Global Energy Economics, International Energy Agency, World Bank, (2019). Data Bank, World Bank.
France. Xiaoli, Z.; Rui, Y.; Qian, M., (2014). China’s total factor energy
IMF, (2015). Adjusting to lower commodity prices. efficiency of provincial industrial sectors. Energy. 65: 52-61
International Monetary Fund, Washington, DC. (10 pages).
401
D. Setyawan
COPYRIGHTS
©2020 The author(s). This is an open access article distributed under the terms of the Creative Commons
Attribution (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, as long
as the original authors and source are cited. No permission is required from the authors or the publishers.
DOI: 10.22034/gjesm.2020.03.08
url: https://www.gjesm.net/article_38759.html
402