Journal of Cleaner Production 143 (2017) 278e287

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Analysis of energy related carbon dioxide emission and reduction

potential in Pakistan
Boqiang Lin a, *, Izhar Ahmad b, c
a
Collaborative Innovation Center for Energy Economics and Energy Policy, China Institute for Studies in Energy Policy, School of Management, Xiamen
University, Xiamen, Fujian, 361005, PR China
b
The School of Economics, China Center for Energy Economics Research, Xiamen University, Xiamen Fujian, 361005, PR China
c
Directorate of Colleges, Higher Education, Archives and Libraries Department, Government of Khyber Pakhtunkhwa, Pakistan

a r t i c l e i n f o a b s t r a c t

Article history: Climate change is one of the most dangerous and complex issues human being has ever encountered.
Received 5 July 2016 Pakistan is one of the highly vulnerable countries to the effects of climate change. CO2 emission from the
Received in revised form combustions of fossil fuels is usually considered as the major factor of climate change. This study
19 December 2016
attempted to analyze the energy related CO2 emissions in Pakistan for a sample period of 1990e2014.
Accepted 20 December 2016
Available online 24 December 2016
The LMDI (Logarithmic Mean Divisia Index) method is applied to extended Kaya identity to decompose
the change in emissions into pre-determined factors. According to our analysis, the increase in GDP per
capita and populations are the major factors responsible for the increase in energy related CO2 emissions.
Keywords:
CO2 emissions
Carbon intensity contributes to the reduction of emissions. Energy intensity and fuel substitution has
LMDI method mixed and unstable effect on the reduction of emission. The decomposed effects are also used in pre-
Emissions reduction potential dicting future CO2 emission for the period 2015e2025. Based on the predicted results, the reduction
Pakistan potential of CO2 emissions in Pakistan is estimated, using special designed scenario analysis. The findings
show that emissions will reach 251.5 Mt CO2 in 2025 as per BAU (Business as usual) scenario. The
reduction potential for the year 2025 is estimated as 28.94 Mt CO2 and 55.02 Mt CO2 as per moderate and
aggressive emission reduction scenario, respectively. The findings show that carbon tax, energy price
reforms, diversification of energy supply in favor of cleaner energy and energy conservation are critical to
materialize the emission reduction potential.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction Climate change affects the functioning and structure of the

ecosystem, and has negative impact on species and their habitats
Climate change refers to the enhanced concentration of Green and also adversely affects water availability, food security and hu-
House Gases (GHGs), especially carbon dioxide (CO2), in the at- man health. This is one of the most dangerous and complex envi-
mosphere which causes global warming. The GHGs includes water ronmental issues man has ever created. In order to mitigate the
vapor, carbon dioxides, methane, nitrous oxides and ozone. The CO2 effects of climate change, the global community is striving to limit
emissions produced by human activities come from the combus- the global mean temperature to 2
C till the end of this century. The
tion of carbon based fuels (coal, oil and natural gas) vis-a
-vis pledges, otherwise known as the Intended Nationally Determined
deforestation, soil erosion, animal agriculture etc. In year 2010, 68% Contributions (INDCs), submitted by individual countries to the
of global GHGs emission is attributed to the energy sector. This 21st Conference of the Parties (COP21) on climate change in
includes a share of 90% for carbon dioxide, 9% methane and 1% December 2015, are critical to achieving this goal. Therefore, to
nitrous oxide and others gases (IEA, 2015). significantly deal with CO2 emission, research on individual coun-
tries in the light of the submitted INDCs is immensely important.
Pakistan is among the countries which are most vulnerable to
* Corresponding author.Collaborative Innovation Center for Energy Economics the effects of climate change (Table 1). Kreft et al. (2016) report
and Energy Policy, China Institute for Studies in Energy Policy, School of Manage- Pakistan among the few countries that are recurrently affected by
ment, Xiamen University, Xiamen, Fujian, 361005, PR China. extreme weather events. Since 2010, the country has experienced
E-mail addresses: [email protected], [email protected] (B. Lin),
[email protected] (I. Ahmad).
unprecedented and extreme climate events, resulting in an

http://dx.doi.org/10.1016/j.jclepro.2016.12.113
0959-6526/© 2016 Elsevier Ltd. All rights reserved.
 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287 279

economic loss of US$6 billion (CPEIR, 2015). Although Pakistan The paper is an interesting contribution to the empirical liter-
contributes less than one percent (0.8%) of global carbon emission, ature of CO2 emission mitigation particularly for Pakistan and other
the Government of Pakistan is committed to respond to climate transitional economies in general, specifically in the context of
change, both by adaptation methods as well as by reducing GHGs COP21. The study provides empirical guidelines to compare and
emission. The energy, transport, agriculture & livestock, forestry, contrast the pledges (INDCs in COP21) and potentials of CO2
town planning and industrial sectors are the key areas that needed emissions reduction in the developing countries. For this purpose,
to be intervened in order to mitigate the effects of climate change. the study applied LMDI (Logarithmic Mean Divisia Index) method
Pakistan submitted the INDC to the 21st Conference of Parties of decomposition to the extended Kaya identity. This analysis is
(COP21) on climate change in Paris on 29th November 2015. The further extended to design specific scenarios to answer the ques-
INDC found its roots in the Pakistan Vision (2014) and the National tions (1) What are the major factors that influence the change in
Climate Change Policy (2012). energy related CO2 emissions (2) What is the degree of influence of
Over the last decade Pakistan has been undergoing a severe each factor? (3) What will be the CO2 emissions in future and how
energy crisis in its history. In 2014, Pakistan primary energy supply much is the reduction potential? (4) What should be the policy
was 66.8 million toe (Tons of oil equivalent). The energy supply implications of CO2 emissions mitigations? We believe the study is
consists of oil (32.5%), gas (48.2%), LPG (0.5%), coal (6%), hydro- the first of its nature to thoroughly analyze the past, present and
electricity (11%), and nuclear electricity (1.7%) (Pakistan Energy future CO2 emission with recommendation to materialize emission
Yearbook, 2014). Almost two decades ago, Pakistan electricity reduction potential. Besides, answering the above mentioned
supply mostly consists of hydro energy. However, due to rapid questions will not only support Pakistan energy policy but may also
industrialization and economic expansion, the government adop- be helpful for transition economies seeking accelerated economic
ted short term policies and set up thermal power plants to increase growth. The rest of the paper is as follows:
electricity supply. This policy has severe implications regarding Section 2 reviews the literature related to decomposition of
balance of payments, energy security and environmental sustain- emission using LMDI in developing countries. Section 3 consists of
ability. Besides, the government is still struggling to deal with the data source and methods applied for data analysis. Section 4 re-
energy crisis as the electricity shortfall exceeds 3000 MW in May ports the empirical results and discusses it in details. Section 5
2016. concludes the paper with some policy recommendations.
Pakistan Vision (2014) has been approved with the objectives of
increasing economic growth vis-a
-vis climate change mitigation 2. Literature reviews
and reduction in energy poverty. In order to improve the balance of
payments and energy security, successive governments in Pakistan With the problem of climate change becoming severe, the
since the last decade started to exploit the country's abundant coal decomposition of CO2 emissions into its determining factors has
resources. The $46 billion worth China Pakistan Economic Corridor attracted significant attention in academic research. The IDA (index
(CPEC) mostly consists of energy projects ($33.8 billion). In the first decomposition analysis) assesses the major factors responsible for
phase, the energy projects consist of coal power projects (capacity CO2 emissions change and presents policy implications for emis-
7560 MW) and renewable energy projects (capacity 2790 MW). The sions mitigations (Ang, 2004). A number of decomposition
coal and renewable energy projects are worth $8.8 billion and $6.4 methods have been applied in the literature. Laspeyres index and
billion respectively. Since coal has high carbon content, a higher Divisia index are the most commonly applied methods derived
share of coal projects relative to renewable energy projects has from economics and statistics respectively (Ang and Zhang, 2000).
severe environmental implications, which would be a significant The Laspeyres index, holding other components constant, isolates
setback for climate change cause. the main effect of each component of the aggregated indicator and
The lackluster performance in the past, present and the future compares it with a fixed base year. The divisia index method esti-
outlook of Pakistan energy structure make the situation even more mates each individual effect through formulations in terms of
critical to global climate change leadership. Consequently, Pakistan weighted average logarithmic changes of the relevant variables.
did not specify a concrete carbon reduction plan in its pledge to the The studies of (Ang and Choi, 1997) and (Ang et al., 1998)
COP21. Moreover, the carbon mitigating pledge is subject to introduced the LMDI (Logarithmic Mean Divisia Index) under the
“affordability, provision of international climate finance, transfer of umbrella of IDA. This method was conveniently applied in cross
technology and capacity building” (INDC, 2015). However, this country/regional energy & environmental related analysis and was
enraged the climate experts at home. They believe Pakistan could regarded as the preferred method due to its theoretical founda-
have pledge for solid targets without undermining the develop- tions, adaptability and easy to use characteristics etc. (Ang, 2004;
ment goals. It is further believed that the pledge i.e. (INDC, 2015). is Ang and Zhang, 1999; and Zhang and Ang, 2001).
a far cry even from a more substantive draft summary forwarded Ang and Liu (2001) find a new decomposition method having
for the approval of the Prime Minister, which pledge to reduce the the desirable properties of perfect decomposition and consistency
emissions by 5% or 18% (if supported by rich nations) compared to in aggregations. This method was named as LMDI I (Log-Mean
2012 emissions level, in 2030 (Ebrahim, 2015). This paradoxical Divisia Index Method 1) and the earlier method was later named as
backdrop of pledge & potential of reducing the emissions motivate LMDI II. Regarding the applicability and flexibility, the LMDI I has
us to thoroughly analyze CO2 emissions from the energy sector of several advantages over other perfect decomposition methods i.e.
Pakistan. adjustment for zero and negative values (Ang et al., 2003; Ang and

Table 1
Pakistan vulnerability to climate change. Source: National Climate Change Policy, 2012.

1 Substantial increase in the frequency and intensity of extreme weather events, unpredictable rains leading to severe rivers' water overflow
2 Anticipated depletion of Hindu Kush Himalaya (HKH) glacier intimidating water inflows into Indus River System (IRS).
3 Increased incursion of saline water into Indus delta due to rise of sea-level.
4 Increased temperature leading to reduce productivity of agriculture.
5 These are the potential threats to water, food, energy and national security.
280 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287

Liu, 2007). Besides energy conservation and efficiency, the LMDI I mineral product industry into predetermined factors. The de-
method has also been utilized in empirical investigation of carbon terminants were further used to forecast future CO2 emissions in
footprint of energy consumption, water footprint of crop produc- the respective sectors. The intensity effect has negative sign, rep-
tion and industrial wastewater emissions etc. (Geng et al., 2014; resenting energy efficiency gains. Sonnenschein and Mundaca
Wang et al., 2013; Xu et al., 2015). This literature review reports (2016) study the decarbonization under the green growth strate-
the most recent studies, conducted in developing countries on the gies for South Korea. Using both decomposition and econometric
energy related CO2 emissions using LMDI I.1 methodologies for the period 1971e2012, the findings show that
Paul and Bhattacharya (2004) conducted a comprehensive the impacts of the important driving factors of energy and carbon
assessment of the aggregate and sectoral CO2 emissions in India intensities have even worsened in the initial years of green growth
from 1980 to 1996. The results showed that economic growth has strategies. The econometric analysis shows GDP per capita as the
the largest positive effect on aggregate as well as major sectoral CO2 major factor of GHGs emissions. Robaina-Alves et al. (2016) ana-
emissions. CO2 emissions in the transport and industrial sector are lyzes the CO2 emissions from Portugal tourism industry for the
decreasing due to improved energy efficiency and efficient fuel period of 2000e08, by dividing the industry into five sub-sectors.
substitution. The results further showed a relatively greater impact Using the LMDI approach of decomposition, the study concluded
of energy intensity on energy related CO2 emissions compared to that tourism activity has the strongest effect in increasing emis-
the pollution coefficient. Wang et al. (2005) analyzed the aggre- sions. Carbon and energy intensity are effective in reducing emis-
gated CO2 emissions for China from 1957 to 2000. The findings sions while energy substitution has a mixed effect on CO2 emission.
showed that energy intensity, fossil fuel substitution and renew- In case of Pakistan, some studies examine the economy from
able energy penetration are significant to reduce aggregate CO2 environmental perspective in different sectors. Yousuf et al. (2014)
emissions. Lise (2006) in his analysis on Turkey's economy for the has calculated the baseline emission factor, for the first time in
period 1980e2003 shows that economic expansion is the largest Pakistan, by determining the fossil fuel consumption in power
driving force of CO2 emissions, followed by carbon intensity and generation, their net efficiencies, energy output and carbon emis-
structure of the economy. The study could not find evidence of sions from each fuel source. The study claims that the baseline
decoupling of CO2 emissions and economic growth over the sample emission factor is beneficial in determining the per mega watt
period. Liu et al. (2007) show that industrial activity and energy hours emissions and is able to determine the cleaner project. The
intensity exhibit an overwhelming positive and negative effect, study mainly focuses on alternative and renewable energies to
respectively, on China industrial CO2 emission from 1998 to 2005. mitigate carbon emission. Ortolano et al. (2014) evaluates the
Ma and Stern (2008) inculcate biomass in the extended Kaya adoption of cleaner production measures as well as firms compli-
identity and focused on the fall and resurgence of China's aggregate ance with Pakistan environmental standard in Pakistan leather
CO2 emissions since the mid-1990s. The findings show that the tanneries and textile firms. A survey of 80 tanneries and firms
positive effect of population growth is diminishing over time. Oh concluded that most of the firms adopted the cleaner production
et al. (2010) employed the LMDI method on five energy intensive measures. The size and involvement of firms with foreign cus-
sectors and 7 sub-sectors in South Korea. The analysis was made in tomers are encouraging the adoption of cleaner production and the
terms of fuel mix, energy intensity, structural change and economic establishment of environmental management standard. Khan et al.
growth. The findings show economic growth to be a dominant (2016) examine the relationship among energy consumption, air
factor for increasing CO2 emissions across all the sectors. Fuel mix pollution, water resources and natural resource rent. Results indi-
drives the reduction of CO2 emissions in all the sectors, except the cate that energy consumption and water resources are positively
energy supply sector. Energy intensity was the prominent driving affecting the air pollution in the short as well as long run. The
factor of the reduction in CO2 emissions for all sectors except for the natural resource rent is; however, relatively less effective to affect
several manufacturing sub-sectors. Tunc et al. (2009) subdivided the air pollution. Khan and Jamil (2015) is the only stud in Pakistan
the Turkish economy into agriculture, industry and services sector. to decompose the aggregate CO2 emissions for the period
The energy sources used in these sectors are then aggregated into 1990e2012. The study finds that activity effect is the largest driving
solid fuels, petroleum, natural gas and electricity. The study in- force of changes in CO2 emissions, followed by structural and in-
vestigates the effects of macroeconomic policies on CO2 emissions tensity effects.
through changes in the shares of industries and use of different While identifying the dearth of literature on the decomposition
energy sources for the period 1970e2006. The results show that the analysis of energy related CO2 emissions in Pakistan, this study tries
main driving force of CO2 emissions is economic activity. Intensity to contribute to the literature on developing countries in general
effect significantly reduces CO2 emissions for a sub-sample, while and Pakistan in particular. The study attempted to decompose the
the structure effect does not. Akbostancı et al. (2011) decomposed change in Pakistan fossil fuel related CO2 emissions into pre-
CO2 emissions from Turkish industries for the period 1957e2001. determined factors based on Kaya identity. The results are further
Industrial activity and energy intensity are the driving factors of used in special designed scenario analysis to forecast future CO2
change in CO2 emissions. Among the sources of energy, coal is the emission and the reduction potential of CO2 emission in Pakistan.
determining factor while among the different sectors, the iron and To the best of our knowledge, this study is the first of its nature in
steel sector is the dominant sector with respect to CO2 emissions. the energy literature of Pakistan.
Wang et al. (2013) estimated the GHGs inventory in China's cement
industry and identified the main driving factors using LMDI method 3. Materials and methods
of decomposition. The study concluded that cement and clinker
production activities are the factors increasing GHGs emission 3.1. Variables and data source
while energy intensity is effective to mitigate emissions. The study
provides some policy recommendation to effectively reduce GHGs The study is based on annual observation covering the period
emissions. Lin and Mubarak (2013) and Lin and Ouyang (2014) 1990e2014. The variables are selected on the basis of the Kaya
decomposed CO2 emissions in China textiles and nonmetallic Identity (Kaya, 1990). Thus the variables include CO2 emission, FFE
(Fossil Fuels Energy consumption), TOE (Total Energy Consump-
tion), GDP (Gross Domestic Product), Pop (Population).
1
We use LMDI instead of LMDI I in the remaining paper. CO2 emissions from fossil fuels consumption are calculated by
 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287 281

the formula presented in equation (1). fossil fuels, EI stands for energy intensity, AC stands for over all
activities or “affluence” and Pop stands for total population.
X
8
CO2 ¼ Ei $CFi $CCi $COFi $ð44=12Þ (1)
i¼1 3.3. Methods

wherein, We adopted the LMDI method to study the change in CO2

emissions from the combustion of fossil fuels as this is a perfect
CO2 /fossil fuels related carbon dioxide emissions decomposition approach with no residual. It can efficiently handle
i/ represent the ith energy. (Fossil fuels energy includes, motor zero and negative values and time reversal test can also be applied
gasoline, aviation gasoline, kerosene, diesel oil, residual fuel oil,
to it (Ang and Liu, 2007). The change in CO2 emission ðDCO2 Þ from
LPG, natural gas, coal)
Pakistan fossil fuel consumption between a base year 0 and an end
CFi / Conversion factor from physical unit to kjoule (the
year 1 can be decomposed into the following effects.
calculation takes the appendix 7.4 of Pakistan Energy Yearbook
(2014)).
CIeff / Carbon intensity effect, which is the change in carbon
CCi / Carbon content coefficient of the ith fuel (Collected from
emissions per unit of fossil fuels consumption.
IPCC, 2006)
Subeff /Substitution effect, which is the change in emissions
COFi / Carbon oxidation factor
due to the change in the share of fossil fuel in total energy
44=12/ Conversion factor from carbon to carbon dioxide
consumption.
EIeff /Energy intensity effect, which is the change in emissions
It is stated that CFi $CCi  COFi $ð44=12Þ is the coefficient of CO2
due to the change in total energy consumption per unit of GDP.
emission for the ith energy. The coefficient is usually assumed
ACeff / Activity effect, which is the change in emission due to
constant overtime. In order to avoid double counting of CO2
change in GDP per capita.
emission from power generation sector, electricity generation from
Popeff /Population effect, which is the change in emissions due
hydro and nuclear sources is used as proxy for electricity con-
to change in total population of Pakistan.
sumption and is therefore regarded as clean.2 Consequently, elec-
tricity generation can also be treated as renewable energy in this
Since the LMDI is a method which totally decomposes the
case. A description of all the variables is presented in Table 2. The change into the related factors and there is no question of residual
calculation of CO2 emissions series has some limitations. These
factor, therefore change in CO2 emission can be written as in
include the current values of energy content, conversion rates of equation (5).
sources of fossil fuels from physical unit to kjouls and emissions
factors in defining the emissions performance of fossil fuels in the DCO2 ¼ CO2 ð1Þ  CO2 ð0Þ ¼ CIeff þ Subeff þ EIeff þ ACeff þ Popeff
time series analysis.
(5)
3.2. Model construction The effects in turn can be calculated from the following
formulae in equations (6)e(10) based on the LMDI method:
The decomposition of fossil fuel induced CO2 emissions into pre-
determined factors dates back to a series of studies undertaken in CIeff ¼ W$ln½CIð1Þ=CIð0Þ (6)
1980s, where Kaya proposed an identity and decomposed CO2
emission into related factors. This is known as the Kaya Identity, Subeff ¼ W$ln½Subð1Þ=Subð0Þ (7)
presented in equation (2).3

CO2 TOE GDP EIeff ¼ W$ln½EIð1Þ=EIð0Þ (8)

CO2 ¼ $ $ $Pop (2)
TOE GDP Pop
ACeff ¼ W$ln½ACð1Þ=ACð0Þ (9)
Equation (2) established the relationship between CO2 emis-
sions and the determining factors of CO2 emissions change such as
CO2 emission per unit of energy consumption, energy consumption Popeff ¼ W$ln½Popð1Þ=Popð0Þ (10)
efficiency, level of economic development and population. Equa-
tion (2) is subsequently expanded as: where

COt2 EFF t TOEt GDP t CO2 ð1Þ  CO2 ð0Þ

COt2 ¼ $ $ $ $Popt (3) W¼ (11)
EFF t TOEt GDP t POP t ln½CO2 ð1Þ=CO2 ð0Þ

The Kaya identity in equation (3) focuses on CO2 emission from Based on equations (6)e(10), the CO2 emissions change in
the combustion of fossil fuels. Equation (3) can be also written as Pakistan fossil fuel consumption can be computed. Our sample
observation ranges from 1990 to 2014. However, for easy data
CO2 ¼ CI$Sub$EI$AC$Pop (4) management, we also sub-divide the data into 5 year interval as
1990e1994, 1995e1999, 2000e2004, 2005e2009, 2010e2014 (Lin
Where CI stands for carbon intensity, Sub stands for substitution of and Mubarak, 2013; Lin and Ouyang, 2014). The time interval is
reasonable because the interval for national economic planning in
Pakistan is also 5 years.
2
Since electricity consumption is regarded as clean; therefore, the two phrases Since one of the objectives of the study is to estimate the
fossil fuels related and energy related CO2 emission is used interchangeably in this
reduction potential of CO2 emissions, we further expand and
paper.
3
Some economist believes that cobb douglas function produce better results in
construct the decomposition model. Equation (5) implies that the
analysis of CO2 emissions. While we acknowledge this, such consideration is left as change in CO2 emission from base year 0 to end year 1 is the sum of
an avenue to the future research. all effects. Therefore, given the CO2 emission in the base year 0 and
282 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287

Table 2
Description of variables.

Variables Unit Data source Mean Standard deviation

Carbon Dioxide Emission ðCO2 Þ Million tons Calculated as per Eq. (1) 125.404 33.99
Fossil Fuel ConsumptionðEFFÞ Tera Jouls Pakistan Economic Surveys (2006; 2015) 1770972 491035
Total Energy Consumption ðTOEÞ Tera Jouls Pakistan Economic Surveys (2006; 2015) 2045851 548284.5
National Income (GDP) PKR million WDI, 2015 6979050 1916745
Population (Pop) Millions WDI, 2015 148.844 21.847

the carbon intensity effect ðCIeff Þ, substitution effect ðSubeff Þ, energy factor can be quantified.
intensity effect ðEIeff Þ, activity effect ðACeff Þ, and population effect
ðPopeff Þ, we can forecast the CO2 emission in the target year 1 as per 4. Results and discussions
equation (12).
4.1. Carbon dioxide emissions change during the time interval
CO2 ð1Þ ¼ CO2 ð0Þ þ CIeff þ Subeff þ EIeff þ ACeff þ Popeff (12)
The LMDI decomposition analysis quantifies the impact of the
Assuming that a; b, g; d and q are the rate of change in carbon determinants of fuel related CO2 emissions by degree. In this paper,
intensity, fuel substitution, energy intensity, overall activity, and we split the study sample from 1990 to 2014 into five time intervals.
population respectively from base year 0 to target year 1, therefore, The rationale behind this split is that (1) Pakistan national eco-
nomic planning which drives economic development is based on
CIð1Þ ¼ CIð0Þ$ð1 þ aÞ (13)
five year time intervals i.e. annual 5 years plan, and medium term
development framework. (2) The effects of each determinant of
Subð1Þ ¼ Subð0Þ$ð1 þ bÞ (14) fossil fuel induced CO2 emissions among the sub-samples may be
captured and compared.
EIð1Þ ¼ EIð0Þ$ð1 þ gÞ (15) The results show that significant changes takes place in CO2
emissions from the combustions of fossil fuels in Pakistan from
ACð1Þ ¼ ACð0Þ$ð1 þ dÞ (16) 1990 to 2014. Identifying and quantifying the reasons for emissions
changes are usually helpful in designing policy related to CO2
emission mitigations. Table 3 and Fig. 1 show that overall activity/
Popð1Þ ¼ Popð0Þ$ð1 þ qÞ (17)
GDP per capita effect ðACeff Þ and population effect ðPopeff Þ are
Equation (4) employs that CO2 emission for the base year 0 is significantly affecting the increase in CO2 emission across all the
sub-periods/time intervals. The overall activity effect ðACeff Þ is in
CO2 ð0Þ ¼ CIð0Þ$Subð0Þ$EIð0Þ$ACð0Þ$Popð0Þ (18) compliance with the earlier results across the sub-periods for
And the CO2 emission for the target year 1 is Pakistan (Khan and Jamil, 2015). The population effect ðPopeff Þ is,
somehow, in contrast with the findings in China, where the effect is
CO2 ð1Þ ¼ CIð1Þ$Subð1Þ$EIð1Þ$ACð1Þ$Popð1Þ (19) significant but falling over time (Wang et al., 2005). Across the sub-
periods both the activity and population effect, with some excep-
Combining equation (19) and equations (13)e(17), we get
tions, are increasing over time. The carbon intensity effect ðCIeff Þ is
equation (20)
negatively affecting the changes in CO2 emissions across all the
CO2 ð1Þ ¼ CO2 ð0Þ$ð1 þ aÞ$ð1 þ bÞ$ð1 þ gÞ$ð1 þ dÞ$ð1 þ qÞ (20) time intervals except for the period 1995e1999. The energy in-
tensity ðEIeff Þ and fuel substitution ðSubeff Þ have mixed and unsta-
Putting equation (20) into equation (6), we get
ble effects on CO2 emissions changes across all the sub-periods. The
CIeff ¼ Z$lnð1 þ aÞ (21) carbon intensity effect and fuel substitution effect are in compli-
ance with the carbon intensity and substitution effect in China food
where industry (Lin and Lei, 2015).
The results in Table 3 provide some insight into the policy im-
CO2 ð0Þ$½ð1 þ aÞ$ð1 þ bÞ$ð1 þ gÞ$ð1 þ dÞ$ð1 þ qÞ  1 plications. Across the entire sample of 1990e2014, the results of the
Z¼ (22) sub-periods 2000e04 and 2010e14 are the most important. During
ln½ð1 þ aÞ$ð1 þ bÞ$ð1 þ gÞ$ð1 þ dÞ$ð1 þ qÞ
these two sub-periods, carbon intensity effect ðCIeff Þ, substitution
In a similar fashion, we obtain the expression for substitution effect ðSubeff Þ and energy intensity effects ðEIeff Þ are negatively
effect ðSubeff Þ, energy intensity effect ðEIeff Þ, activity effect ðACeff Þ affecting the change in CO2 emissions. Interestingly, both sub-
and population effect ðPopeff Þ, presented in equation (23)e(26). samples are the preceding periods of global oil shocks of 1999/
2003 and 2008, where price of petroleum products increased
Subeff ¼ Z$lnð1 þ bÞ (23) significantly.
During the period 2010e14, the negative effects of carbon in-
EIeff ¼ Z$lnð1 þ gÞ (24) tensity ðCIeff Þ, fuel substitution ðSubeff Þ, and especially energy in-
tensity effect ðEIeff Þ on CO2 emissions have overridden the positive
ACeff ¼ Z$lnð1 þ dÞ (25) effects of activity effect ðACeff Þ and population ðPopeff Þ. As a result,
only in 2010e14 did the overall CO2 emissions from the combus-
Popeff ¼ Z$lnð1 þ qÞ (26) tions of fossil fuels decreased by 0.0291 Mt (Table 3). The sub-
period was the preceding period of the global financial crisis,
Thus, based on the above model from equation (21)e(26), the mostly induced by the oil price shocks. Since Pakistan mostly rely
future CO2 emission can be forecasted and the contribution of each on thermal source of energy, the surge in oil prices discouraged the
 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287 283

growing energy consumption. There was also evidence of fuel contribution for Pakistan emission increased from 0.297 Mt in 1991
substitution from petroleum products to gas (Lin and Ahmad, to 1.99 Mt in 2014. The average annual growth rate was 22.79%. On
2016). The overall activity effect ðACeff Þ also declined in this sub- the other hand, the component of energy intensity effectðEIeff Þ has
period. Thus, low energy consumption, increased energy effi- mixed effect on the changes in fossil fuel related CO2 emissions.
ciency (rapid decline in energy intensity) and fuel substitution have Earlier in the literature, Tunç et al. (2009) also found mixed and
played a significant role in reducing overall CO2 emissions in unstable results of energy intensity for Turkish economy. The en-
Pakistan during the period 2010e2014. ergy intensity in our analysis changes from 1.243 Mt in 1991
This is also similar for the period 2000e04, however, with lesser to 9.493 Mt in 2014. In the initial years, the contribution of energy
intensity. Although the overall CO2 emissions did not reduce during intensity to CO2 emissions is negative as hydro energy was a
the period 2000e04, the growth rate was slower. The increase in dominant factor in the energy supply mix. However, the power
CO2 emissions in the period 2000e04 is less than the increase in policy of 1993 had placed more focus on thermal power, making the
1995e1999, where the major driver was fuel substitution (Table 3). contribution of energy intensity effectsðEIeff Þ to be positive. Since
In the remaining sub-periods where no oil shock occurred, activity the oil shock-induced global financial crisis in 2007e2009, energy
effect ðACeff Þ and population effect ðPopeff Þ on CO2 emissions was intensity has witnessed a rapid declining trend, and therefore the
substantial and could not be offset by fuel substitution effect contribution of energy intensity effectðEIeff Þ to changes in CO2
ðSubeff Þ, carbon intensity effects ðCIeff Þ and energy intensity effect emissions become negative (Fig. 2).
ðEIeff Þ. The above discussion implies that energy substitution, The structure or fuel substitution effect is also not stable over
conservation and energy price reforms can have important policy the sample period. The fuel substitution effect ðSubeff Þ ranges
implications for mitigating energy related CO2 emissions in from 0.830 Mt in 1991 to 1.713 Mt in 2014. This results comply
Pakistan. These results have also been drawn for Pakistan by studies with the earlier results drawn by Khan and Jamil (2015) for
Lin and Ahmad (2016); UNDP (2015) and Yousaf et al. (2014). Pakistan. Besides, Lin and Mubarak (2013) found mixed and un-
stable results of substitution effect for Chinese textile industry. The
4.2. Aggregated effects of CO2 emissions changes contribution of structure effect to change in CO2 emissions was
negative from 1991 to 1996. This was the period where gas con-
There is a risk of missing out some important information if we sumption in Pakistan, especially in the transport and household
restrict ourselves to the analysis based on time intervals. Therefore, sector, got significant boost and was substituted for petroleum
to avoid such problems, we further estimate the aggregate effects of products. However, the increase in the contribution of petroleum
the contributing factors of fossil fuel-induced CO2 emissions in products to power generation after the power policy of 1995
Pakistan. This calculation for the period 1991e2014 is also based on significantly offset the negative contribution of gas consumption to
equations (6)e(10), using 1990 as the base year. The results are changes in CO2 emissions. Therefore, the contribution of substitu-
shown in Fig. 2. tion effect ðSubeff Þ to changes in CO2 emissions becomes positive
In Fig. 2, CO2 emissions from Pakistan fossil fuels consumption from 1997 onward.
increased from 71.09 Mt in 1990 to 168.27 Mt in 2014, registering a The completion of coal power projects in the first phase of CPEC
growth rate of 136.69%. The figure shows that once again the overall will increase the share of coal in the energy supply mix. Therefore,
activity effect ðACeff Þ and population effect ðPopeff Þ have the the substitution effect ðSubeff Þ in the future will remain positive and
greatest impacts on the increase in CO2 emissions for the period may increase in magnitude. The situation would be even more
1991e2014. Both effects ðACeff &Popeff Þ are increasing over time critical with respect to the mitigating of CO2 emissions. In order to
during the whole sample period. These results are once again in mitigate the environmental effects of coal power projects, the
compliance with Pakistan's earlier results for period 1971e2008 government needs to increase the share of renewable energy in the
(Attari and Attaria, 2011). Studies like Lise (2006); Ma and Stern second phase. Furthermore, there is need to handle the coal pro-
(2008) and Sonnenschein and Mundaca (2016) also found same jects with modern technology. China will transfer its energy supply
results for Turkey, China and South Korea, respectively. The activity mix from coal-dominated to renewable supply mix. Therefore,
effect ðACeff Þ increased from 1.54 Mt in 1991 to 47.52 Mt in 2014. China needs to export its clean coal technology to Pakistan. Under
This increase indicates a growth rate of 2981%. The annual average the CPEC platform, Pakistan gets the best opportunity to import
growth rate of changes in CO2 emissions attributed to the overall renewable and clean coal technologies from China. The import of
activity effect ðACeff Þ is 119.25% during 1991e2014. The population ultra-supercritical coal power technology and solar power tech-
effect increased from 1.996 Mt in 1991 to 59.42 Mt in 2014, with a nology from China may be the best way to exploit abundant coal
growth rate of 2876.0%. and renewable resources. In this case, CO2 emission would be a
Intensity effects ðCIeff &EIeff Þ have mixed impacts on changes in good deal less per unit of energy than imported oil-fired power
plants, which will improve energy security and environmental
CO2 emissions in Pakistan. Meanwhile, the major contributor to the
sustainability (Lin and Ahmad, 2016a).
decline in CO2 emissions is carbon intensity effect ðCIeff Þ Robaina-
While acknowledging that the energy saving or conservation as
Alves et al. (2016) found carbon intensity is effective in reducing
the primary remedy to reduce the emission, recently the global
CO2 emission for Portugal tourism industry. The negative

Table 3
Decomposition of CO2 emission changes from fossil fuels consumption in Pakistan.

Period Emissions change CIeff Subeff EIeff ACeff Popeff

1990e1994 19.42213 0.10005 1.00660 4.22373 5.80853 8.48330

1995e1999 16.27219 0.02962 2.82470 1.59386 1.44730 10.37669
2000e2004 12.22526 1.8892 4.05465 2.63433 11.5215 9.28193
2005e2009 26.04664 0.53264 2.58972 0.14714 12.23969 11.60274
2010e2014 0.02907 0.22217 3.84006 21.2497 11.30724 13.97565
1990e2014 97.18011 1.98917 1.71388 9.49345 47.52474 59.42411
284 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287

30

20

CO2 emissions (Million tons)

10

0
1990-1994 1995-1999 2000-2004 2005-2009 2010-2014

-10

-20

-30
Years

CI effect Sub effects EI effects

AC effects Pop effects CO2 emissions change

Fig. 1. Decomposition of CO2 emission changes from fossil fuels consumption in Pakistan.

120
CO2 emissions (Millions tons)

100
80
60
40
20
0
1990

2003
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002

2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
-20
Years

ΔC Δ_CI Δ_Sub Δ_EI Δ_AC Δ_Pop

Fig. 2. Aggregate effects of contributors to fossil fuel related CO2 emissions in Pakistan.

strategies to reduce the emission also includes the direct method of follows:
carbon capture and storage and the utilization of alternative ma- I. BAU (Business as usual or baseline scenario): This scenario is
terials (Benhelal et al., 2013). Under the platform of CPEC, once based on the historical development trend of each variable. The
again Pakistan can take significant advantage from the increased average growth rates of each variable for the sample period
Chinese investment to develop technology for exploiting alterna- 1990e2014 are chosen as the baseline scenario. Although a baseline
tive energy resources and carbon capturing and storage. emissions factor scenario is available in the study of Yousaf et al.
(2014), we avoid using it as that scenario represents only the
4.3. Emissions reduction potential electricity sector of Pakistan. The BAU scenario reflects the possible
variation trend. However, the purpose is not to provide precise
4.3.1. Estimates of energy related CO2 emission estimates for specific economic conditions but to make clarifica-
As per equation (4), CO2 emissions for the base year 2014 can be tions on the major factors that contribute to future CO2 emissions.
calculated. If the rate of change of CO2 emissions per unit of fossil The other two scenarios are based on the BAU scenario, which can
fuel consumption ðaÞ, fossil fuel share in the total energy con- also be called the reference scenarios.
sumption ðbÞ, total energy consumption per unit of GDP ðgÞ, GDP II. Scenario A: This is the higher emission scenario, where the
per capita ðdÞ, and the population ðqÞ for the forecasting year are growth rates of each variable are two percentages higher than that
exogenous, future energy-related CO2 emissions in Pakistan can be of the growth rates in the BAU scenario.
calculated based on equations (16)e(21). III. Scenario B: This is a lower emissions scenario and the growth
We estimate the future fossil fuels related CO2 emission in rates of each variable are two percentages lower than that of the
Pakistan for the years 2015e25. This time period is important BAU scenario.
because Pakistan government has set up a number of targets The advantage of the above mentioned scenario design is that
regarding growth, energy and climate change mitigations (Vision, we can have a relatively comprehensive analysis of the possible
2014). The INDCs submitted to COP21 also finds its roots in the results of the CO2emission, so that different policies on energy
Visions 2025. While estimating the future energy related CO2 utilization and CO2 emissions can be compared. The hypotheses
emission for Pakistan, we set up three scenarios with the objectives based on our designed scenario analysis are presented in Table 4.
to provide comprehensive comparison of results. These scenarios Based on the hypothesis, the simulation results of CO2 emission
are designed following the studies of Lin and Ouyang (2014), Ou from Pakistan fossil fuel consumption are presented in Fig. 3. As
et al. (2010) and Zhang et al. (2012). The three scenarios are as shown in the figure, CO2 emission from energy consumption in
 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287 285

Pakistan will keep increasing. This is because the energy demand in 2030 (Zeb, 2015) was also reasonable.
Pakistan will increase in the future and the consumption structure
will probably remain stable during the process of economic 5. Conclusions and policy recommendations
development. Furthermore, the likelihood of a critical energy
structure with respect to fossil fuel related CO2 emissions mitiga- 5.1. Conclusions
tion is high, if the new coal power projects do not use modern
technologies. Under the BAU scenario, the CO2 emissions from Being one of the most vulnerable countries to the effects of
Pakistan fossil fuel consumption are predicted as 209.51 Mt in 2020 climate change, the government of Pakistan is keen to mitigate the
and 251.5 Mt in 2025. The emissions will reach 320.03 Mt and effects of climate change. Therefore, Pakistan submitted the INDC
257.36 Mt in 2025 as per scenario A and scenario B respectively. (Intended Nationally Determined Contributions) to the 21st UN
Conference of Parties (COP21) on climate change in Paris in 2015.
This study is an attempt to analyze the change in CO2 emissions
4.3.2. Sensitivity analysis
from the combustions of fossil fuels (one of the major cause of
While we did the simulation for fossil fuel related CO2 emission
climate change) in Pakistan for the period 1990e2014. The objec-
using the scenario analysis, it is believed that sensitivity analysis is
tive is to thoroughly analyze CO2 emissions in light of the submitted
also a powerful tool while presenting the simulation results
INDC and to provide policy recommendations to mitigate it. The
(Kleijnen, 2010). There is little difference between scenario and
LMDI (Log Mean Divisia Index) method is applied to decompose
sensitivity analysis. The sensitivity analysis refers to the process of
changes in CO2 emissions into pre-determined factors based on the
tweaking one variable to check how sensitive the model is to
Kaya identity. The effects of these factors are then utilized to
change in the specific variable. On the other hand, scenario analyses
forecast fossil fuels related CO2 emissions and estimate the reduc-
involve listing of all the variables and changing the value of each
tion potential until 2025.
variable for each scenario, as we did in Table 4.
The results show that population and GDP per capita are the
We did our sensitivity analysis by increasing the value of each
leading factors positively affecting increase in CO2 emissions while
individual variable (while keeping others constant) by 2% from the
carbon intensity is negatively affecting increase in CO2 emissions.
value of BAU value for the simulated period of 2015e2025. Fig. 4
The effects of energy intensity and fossil fuels substitution are
shows that the simulated CO2 emission is more sensitive to the
mixed. The positive effect of GDP per capita and population is
2% increase in population and GDP per capita as compared to car-
substantial compared to the negative effect of carbon intensity and
bon intensity, fuel substitution and energy intensity. This is also
the mixed effects of energy intensity and structural/fossil fuel
been evident from Fig. 2.
substitution over the sample period. Therefore, CO2 emissions are
increasing throughout the sample period, except for the sub-period
4.3.3. Estimates of emission reduction potential of 2009e2014. In this sub-period, the negative effects of carbon
In order to estimate future emissions reduction potential, we intensity, energy intensity and substitution effect outweighs the
further design two scenarios. These scenario analyses are followed positive effect of GDP per capita and population; therefore, the total
from the study of Lin and Ouyang (2014, 2014a). The scenarios CO2 emissions decreases. The forecasting results show that CO2
include the moderate emissions' reduction scenario and the emissions will be 251.5 Mt in 2025 under the business as usual
aggressive emissions' reduction scenario. Under the moderate (BAU) scenario. It further demonstrates that the reduction potential
emissions reduction scenario, CO2 emissions from the combustions of CO2 emissions is 28.94 Mt and 55.02 Mt respectively under the
of fossil fuels in Pakistan will transform from scenario-A to BAU moderate and aggressive emission reduction scenario in 2025.
scenario. In other words, CO2 emissions reduction for a particular
year under the moderate emissions reduction scenario can be ob- 5.2. Policy recommendations
tained by subtracting the CO2 emission for that year under BAU
scenario from the CO2 emissions under scenario-A. The CO2 emis- Being the leading factors positively affecting CO2 emissions, the
sions will transform from Scenario-A to Scenario-B under the current high growth rate of population and the target of achieving
aggressive emissions reduction scenario. high GDP growth in the Vision, 2014 will constitute a major chal-
Utilizing the results in Fig. 3, the reduction potential of CO2 lenge to the role of energy sector reforms in reducing CO2 emis-
emissions from the combustions of fossil fuels in Pakistan, as per sions. The results imply that in addition to population control, the
the designed scenarios, is computed and reported in Fig. 5. Under following measures related to the energy sector are necessary to
the moderate emissions reduction scenario, the potential of fossil mitigate the emissions.
fuel related CO2 emissions reduction is 12.82 Mt in 2020 and The effectiveness of carbon intensity to reduce emissions calls
24.97 Mt in 2025. The potentials are 28.94 Mt and 55.02 Mt under for the promulgation of carbon tax, which has been shown as an
the aggressive emissions reductions scenario in 2020 and 2025, effective way to reduce future energy related CO2 emissions in
respectively. Table 5 shows the reduction potential in percentage Pakistan (Saleem et al., 2015). To reduce the energy intensity, en-
with respect to emission under BAU scenario and emissions in 2005 ergy conservation should be the major strategy compared to energy
and 2012. The entries in Table 5 show that Pakistan could have structure adjustment. This is considered to be a cheaper and
make a solid pledge for emissions reduction in the submitted INDC. feasible way to reduce energy related CO2 emissions. Besides,
It further implies that the draft summary that was submitted for promoting an environment of energy conservation, reforming the
the approval of Prime Minister to reduce emission from 5 to 18% by energy structure in favor of reducing energy intensity, investing in
the substitution of energy with capital and labor, reforming of
energy prices and substituting renewable and domestic energy for
Table 4
Hypothesis of variables (%).
imported and GHGs emitting source of energy are vital ways to
achieve this objective (Lin and Ahmad, 2016). Diversification of the
Scenario CI Sub EI AC Pop
energy supply mix and promotion of a cleaner energy structure are
BAU Scenario 0.0721 0.0739 0.3214 1.7909 2.2205 effective to contain the acceleration of CO2 emissions. To modify the
Scenario A 0.1778 0.2739 0.1214 1.9909 2.4405 fossil fuel dominating energy structure, the consumption of oil &
Scenario B 0.2721 0.126 0.5214 1.5909 2.0205
gas, especially in the power generation sector, should be replaced
286 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287

300

CO2 emissions (Million Tons)

250

200
Actual emissions
150
BAU Scenario
100
Scenario A

50 Scenario B

0
1990 2000 2010 2020
Years

Fig. 3. Actual and estimated fossil fuel related CO2 emissions in Pakistan.

260

250

240
CO2 in BAU
230
Increase in CI
220
Increase in Sub
210
Increase in EI
200 Increase in AC
190 Increase in Pop
180

170
2015 2017 2019 2021 2023 2025

Fig. 4. Sensitivity of forecasted CO2 to the individual increase in the variables.

60
CO2 emmissions (Millios Tons)

50

40

30
Moderate Scenario
20
Aggressive scenario
10

years
Fig. 5. Reduction potential of fossil fuel related CO2 emissions under moderate and aggressive emissions reduction scenarios.

Table 5
CO2 emissions reduction potential in percentage.

Years %Age of BAU emissions %Age of emission in 2012 (166.3 Mt CO2) %Age of emission in 2005 (144.26 Mt CO2)

Moderate scenario Aggressive scenario Moderate scenario Aggressive scenario Moderate scenario Aggressive scenario

2020 6.12 11.91 7.7116 15.014 8.889 17.307

2025 11.5 21.88 17.402 33.084 20.06 38.137
 B. Lin, I. Ahmad / Journal of Cleaner Production 143 (2017) 278e287 287

by cleaner source of energy such as hydro, solar, wind, etc. resources in Pakistan. J. Clean. Prod. 112, 1375e1385.
Kleijnen, J.P., 2010. Sensitivity analysis of simulation models: an overview. Procedia-
Energy pricing reform is a long term strategy for mitigating
Social Behav. Sci. 2 (6), 7585e7586.
energy related CO2 emissions. Renewable energy will be competi- Kreft, S., Eckstein, D., Dorsch, L., Fischer, L., 2016. Global Climate Risk Index. Who
tive with fossil fuels in Pakistan, if fossil fuels price reflects its true Suffers Most from Extreme Weather Weather-related Loss Event in 2014 and
cost. There is need for the rationalization of fuel tariffs and sub- 1995-2014? (Accessed 26 March 2016). http://germanwatch.org/en/download/
13503.pdf.
sidies as well as government support for renewable energy in the Lin, B., Ahmad, I., 2016. Technical change, inter-factor and inter-fuel substitution
form of feed-in-tariff. To address the high cost of renewable energy, possibilities in Pakistan: a trans-log production function approach. J. Clean.
support in the form of increasing renewable energy subsidies is a Prod. 126, 537e549.
Lin, B., Ahmad, I., 2016a. Energy substitution effect on transport sector of Pakistan
common scheme in many countries (Lund, 2007). Thus, acknowl- based on trans-log production function. Renew. Sustain. Energy Rev. 56,
edging Pakistan's carbon reduction pledge which is conditional on 1182e1193.
international climate leadership support (INDC, 2015), imple- Lin, B., Lei, X., 2015. Carbon emissions reduction in China's food industry. Energy
Policy 86, 483e492.
mentation of the above mentioned recommendations would surely Lin, B., Mubarak, M., 2013. Decomposition analysis: change of carbon dioxide
encourage the international community to transfer climate finance emission in Chinese textile industry. Renew. Sustain. Energy Rev. 26, 389e396.
and technology to Pakistan. Lin, B., Ouyang, X., 2014. Analysis of energy-related CO2 (carbon dioxide) emissions
and reduction potential in the Chinese non-metallic mineral products industry.
Energy 68, 688e697.
Acknowledgements Lin, B., Ouyang, X., 2014a. Electricity demand and conservation potential in the
Chinese nonmetallic mineral products industry. Energy Policy 68, 243e253.
Lise, W., 2006. Decomposition of CO2 emissions over 1980e2003 in Turkey. Energy
The paper is supported by the Grant for Collaborative Innovation Policy 34 (14), 1841e1852.
Center for Energy Economics and Energy Policy (No: 1260- Liu, L.C., Fan, Y., Wu, G., Wei, Y.M., 2007. Using LMDI method to analyze the change
Z0210011), Xiamen University Flourish Plan Special Funding (No: of China's industrial CO2 emissions from final fuel use: an empirical analysis.
Energy Policy 35 (11), 5892e5900.
1260-Y07200), and Newcastle University Joint Strategic Partnership Lund, H., 2007. Renewable energy strategies for sustainable development. Energy
Fund. 32 (6), 912e919.
Ma, C., Stern, D.I., 2008. Biomass and China's carbon emissions: a missing piece of
carbon decomposition. Energy Policy 36 (7), 2517e2526.
References National Climate Change Policy, 2012. Ministry of Climate Change. Government of
Pakistan, Islamabad.
Akbostancı, E., Tunç, G.I., _ Türüt-Aşık, S., 2011. CO2 emissions of Turkish Oh, I., Wehrmeyer, W., Mulugetta, Y., 2010. Decomposition analysis and mitigation
manufacturing industry: a decomposition analysis. Appl. Energy 88 (6), strategies of CO2 emissions from energy consumption in South Korea. Energy
2273e2278. Policy 38 (1), 364e377.
Ang, B.W., 2004. Decomposition analysis for policymaking in energy: which is the n, S.A., 2014. Cleaner pro-
Ortolano, L., Sanchez-Triana, E., Afzal, J., Ali, C.L., Rebello
preferred method? Energy policy 32 (9), 1131e1139. duction in Pakistan's leather and textile sectors. J. Clean. Prod. 68, 121e129.
Ang, B.W., Choi, K.H., 1997. Decomposition of aggregate energy and gas emission Ou, X., Zhang, X., Chang, S., 2010. Scenario analysis on alternative fuel/vehicle for
intensities for industry: a refined Divisia index method. Energy J. 18 (3), 59e73. China future road transport: life-cycle energy demand and GHG emissions.
Ang, B.W., Liu, F.L., 2001. A new energy decomposition method: perfect in Energy Policy 38 (8), 3943e3956.
decomposition and consistent in aggregation. Energy 26 (6), 537e548. Pakistan Economic Survey, 2006. Ministry of Finance, Government of Pakistan,
Ang, B.W., Liu, F.L., Chew, E.P., 2003. Perfect decomposition techniques in energy Islamabad.
and environmental analysis. Energy Policy 31 (14), 1561e1566. Pakistan Economic Survey, 2015. Ministry of Finance, Government of Pakistan,
Ang, B.W., Liu, N., 2007. Handling zero values in the logarithmic mean Divisia index Islamabad.
decomposition approach. Energy Policy 35 (1), 238e246. Pakistan Energy Yearbook, 2014. Government of Pakistan, Islamabad. Hydrocarbon
Ang, B.W., Zhang, F.Q., 1999. Inter-regional comparisons of energy-related CO2 Development Institute. Ministry of Petroleum & Natural Resource.
emissions using the decomposition technique. Energy 24 (4), 297e305. Paul, S., Bhattacharya, R.N., 2004. CO2 emission from energy use in India: a
Ang, B.W., Zhang, F.Q., 2000. A survey of index decomposition analysis in energy decomposition analysis. Energy Policy 32 (5), 585e593.
and environmental studies. Energy 25 (12), 1149e1176. Robaina-Alves, M., Moutinho, V., Costa, R., 2016. Change in energy-related CO2
Ang, B.W., Zhang, F.Q., Choi, K.H., 1998. Factorizing changes in energy and envi- (carbon dioxide) emissions in Portuguese tourism: a decomposition analysis
ronmental indicators through decomposition. Energy 23 (6), 489e495. from 2000 to 2008. J. Clean. Prod. 111, 520e528.
Attari, M.I.J., Attaria, S.N., 2011. The decomposition analysis of CO2 emission and Saleem, M.M., Khan, M.B., Sarwar, U.B., 2015. The impact of carbon tax policies on
economic growth in Pakistan India and China. Pak. J. Commer. Soc. Sci. 5 (2), future carbon emissions from Pakistan's energy sector using the Times
330e343. Modeling Framework. Explore Islamabad J. Soc. Sci. 1 (6), 193e198.
Benhelal, E., Zahedi, G., Shamsaei, E., Bahadori, A., 2013. Global strategies and po- Sonnenschein, J., Mundaca, L., 2016. Decarbonization under green growth strate-
tentials to curb CO2 emissions in cement industry. J. Clean. Prod. 51, 142e161. gies? The case of South Korea. J. Clean. Prod. 123, 180e193.
CPEIR, 2015. Pakistan-climate Public Expenditure and Institutional Review (CPEIR). Tunç, G.I., Türüt-Aşık, S., Akbostancı, E., 2009. A decomposition analysis of CO2
UNDP, Islamabad (Accessed 15 June 2016). http://www.pk.undp.org/content/ emissions from energy use: Turkish case. Energy Policy 37 (11), 4689e4699.
dam/pakistan/docs/Environment%20&%20Climate%20Change/UNDP% UNDP, 2015. Study on the Assessment of CO2 Emissions from Truck Freight Trans-
20Climate%20Report%20V10.pdf. port in a Low Carbon Scenario. Pakistan sustainable transport project,
Ebrahim, T.Z., 2015. Pakistan offers Nothing to Paris Climate Summit (Accessed 03 Islamabad.
April 2016). https://www.thethirdpole.net/2015/11/18/pakistan-offers-nothing- Vision, 2025, 2014. Pakistan 2025, One Nation-one Vision. Planning Commission,
to-paris-climate-summit. Ministry of Planning, Development & Reforms. Government of Pakistan,
Geng, Y., Wang, M., Sarkis, J., Xue, B., Zhang, L., Fujita, T., Yu, X., Ren, W., Zhang, L., Islamabad.
Dong, H., 2014. Spatial-temporal patterns and driving factors for industrial Wang, C., Chen, J., Zou, J., 2005. Decomposition of energy-related CO2 emission in
wastewater emission in China. J. Clean. Prod. 76, 116e124. China: 1957e2000. Energy 30 (1), 73e83.
IEA, 2015. CO2 emission from Fuels Combustions: Highlights. IEA statistics, Inter- Wang, Y., Zhu, Q., Geng, Y., 2013. Trajectory and driving factors for GHG emissions in
national energy agency, OECD, Paris. the Chinese cement industry. J. Clean. Prod. 53, 252e260.
INDC, 2015. Pakistan- Intended Nationally Determined Contribution, Submitted to WDI, 2015. World Development Indicators. World Bank, Washington DC (Accessed
COP21 (Accessed 02 June 2016). http://www4.unfccc.int/submissions/indc/ 07 April 2016). http://data.worldbank.org/country/pakistan.
Submission%20Pages/submissions.aspx. Xu, Y., Huang, K., Yu, Y., Wang, X., 2015. Changes in water footprint of crop pro-
IPCC, 2006. Guidelines for National Greenhouse Gas Inventories. Inter-govern- duction in Beijing from 1978 to 2012: a logarithmic mean Divisia index
mental panel on climate change. Volume 2, Chapter 1: 23-24. (Accessed 07 decomposition analysis. J. Clean. Prod. 87, 180e187.
April 2016). https://www.ipcc.ch. Yousuf, I., Ghumman, A.R., Hashmi, H.N., Kamal, M.A., 2014. Carbon emissions from
Kaya, Y., 1990. Impact of Carbon Dioxide Emission on GNP Growth: Interpretation of power sector in Pakistan and opportunities to mitigate those. Renew. Sustain.
Proposed Scenarios. In: Paris: Paper Presented to the IPCC Energy and Industry Energy Rev. 34, 71e77.
Subgroup, Response Strategies Working Group. Paris. Zhang, F.Q., Ang, B.W., 2001. Methodological issues in cross-country/region
Khan, A., Jamil, F., 2015. Energy Related CO2 Emissions in Pakistan: a Decomposition decomposition of energy and environment indicators. Energy Econ. 23 (2),
Analysis Using LMDI. S3H Working Paper Series Number 8. National University 179e190.
of Science and Technology (NUST), Islamabad, Pakistan (Accessed 07 April Zhang, Q., Ishihara, K.N., Mclellan, B.C., Tezuka, T., 2012. Scenario analysis on future
2016). http://www.nust.edu.pk/INSTITUTIONS/Schools/S3H/Research/Pages/ electricity supply and demand in Japan. Energy 38 (1), 376e385.
Economics-Publications.aspx. Zeb, A., 2015. Pakistan to Pledge 5 % Emissions Reduction (Accessed 03 April 2016).
Khan, M.M., Zaman, K., Irfan, D., Awan, U., Ali, G., Kyophilavong, P., Naseem, I., 2016. https://www.thethirdpole.net/2015/09/29/pakistan-to-pledge-5-emission-
Triangular relationship among energy consumption, air pollution and water reduction.