- Email: [email protected]
China's carbon emissions embodied in (normal and processing) exports and their driving forces, 2006–2012
China’s carbon emissions embodied in (normal and processing) exports and their driving forces, 2006–2012 Bin Su, Elspeth Thomson PII: DOI: Reference:
S0140-9883(16)30243-2 doi: 10.1016/j.eneco.2016.09.006 ENEECO 3436
To appear in:
Energy Economics
Received date: Revised date: Accepted date:
25 April 2016 24 August 2016 6 September 2016
Please cite this article as: Su, Bin, Thomson, Elspeth, China’s carbon emissions embodied in (normal and processing) exports and their driving forces, 2006–2012, Energy Economics (2016), doi: 10.1016/j.eneco.2016.09.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT China’s Carbon Emissions Embodied in (Normal and Processing) Exports and Their Driving Forces, 2006-2012
IP
T
Bin Su *, Elspeth Thomson
SC R
Energy Studies Institute, National University of Singapore, Singapore
NU
Abstract
This paper constructed a time-series extended input-output dataset (2006-2012) to
MA
analyze China’s carbon emissions embodied in both normal and processing exports at a detailed sectoral level. The structural decomposition analysis (SDA) was further applied to shed light on the driving forces behind the changes in their embodied emissions over the
D
entire time period. This empirical study confirms the importance of using the extended model
TE
for analyzing the trade-related embodiment, especially for processing exports. The embodied emissions in both normal and processing exports first increased from 2006 to 2008, then
CE P
dropped during the global financial crisis (2008-2009), and then rose again after 2009. The embodied emissions as a percentage of total CO2 emissions were quite stable before and after the global financial crisis, at around 24% over the 2006-2008 period, and 18% over the 2010-
AC
2012 period. From 2006 to 2012, emission intensity played the key role in reducing the embodied emissions (around 595 Mt CO2), while the total export effect contributed the most to the increase in embodied emissions (around 552 Mt CO2). Similar analysis can be applied to other indicators, such as energy, water, GHG emissions, pollutants and materials.
Keywords: Input-output analysis; Emissions embodied in trade; Structural decomposition analysis; Processing exports; China
*
Corresponding author. Tel.: +65-6601-2075; Fax: +65-6775-1831. E-mail address: [email protected] ; [email protected] (B. Su). —1—
ACCEPTED MANUSCRIPT 1. Introduction In November 2014, the General Office of the State Council (GOSC, 2014) announced ―China’s Energy Development Strategy Action Plan (2014-2020)‖, which provided the
T
guidelines for China’s energy developments during the 12th (2011-2015) and 13th (2016-
IP
2020) Five Year Plans. As the world’s largest energy consumer and CO2 emitter, China is doing all it can to conserve energy and reduce its CO2 emissions. At the COP21 conference
SC R
recently held in Paris (30 Nov – 11 Dec 2015), the countries of the world agreed to limit the temperature increase below 2°C. However, based on the national climate actions plans (intended nationally determined contributions - INDCs) submitted by the governments before
NU
COP21, these actions will not be sufficient to achieve this target.
Over the last decade, hundreds of studies have been carried out examining how
MA
international trade affects countries’ domestic emissions, or ―emissions embodied in trade‖ and result in ―carbon leakage‖ between developed and developing countries through trade. Since 1990, due to globalization, merchandise trade value (exports plus imports) increased
D
from USD 7,090 billion in 1990 to USD 32,732 billion in 2008 (WTO, 2015), while
TE
embodied emissions increased from 4.3 Gt CO2 to 7.8 Gt CO2 over the same period (Peters et al., 2011). China exported more than 20% of its total CO2 emissions after year 2000 (Su and
CE P
Ang, 2013; Sato, 2014). China’s energy efficiency improvements, especially those carried out in the 11th Five Year Plan (2006-2010) period greatly helped reduce the embodied emissions in China’s exports. However, the demand for China’s products continues to increase. Thus it
AC
is important for China to monitor this demand and understand how it affects China’s energy use and resulting CO2 emissions. To date there have been many studies on China’s embodied emissions at the national level, such as Weber et al. (2008), Su et al. (2010), Lin and Sun (2010), Chen and Chen (2010, 2011), Yan and Yang (2010), Xu et al. (2011) and Ren et al. (2014). Most of these were based on the environmental input-output (I-O) framework (Leontief, 1970; Miller and Blair, 2009). A review of them and their estimates for China can be found in Su and Ang (2013), Sato (2014) and Hawkins et al. (2015). Some recent studies look into China’s embodied emissions in both interregional and international trade, such as Su and Ang (2010), Feng et al. (2012), Guo et al. (2012), Meng et al. (2013), Chen et al. (2013), Su and Ang (2014a), Zhang et al. (2014), Liu et al. (2015) and Zhang and Tang (2015). In addition, there have been studies which use the structural decomposition analysis (SDA) to understand the driving forces behind the embodied emission changes. A review of SDA studies can be found at Su and Ang —2—
ACCEPTED MANUSCRIPT (2012a) and recent studies include Su and Ang (2012b, 2013, 2014b, 2015, 2016), Xu and Dietzenbacher (2014), Xia et al. (2015), and Zhang and Tang (2015). There is another interesting feature in China’s international exports. Figure 1 shows that
T
about half of China’s exports are processing exports, meaning exports of end products made
IP
from imported assembling/processing intermediate inputs, and which are exempted from Chinese tariffs. The emissions embodied per dollar of processing exports are found to be
SC R
much lower than emissions embodied per dollar of normal exports (Su et al., 2013). Differentiating the normal and processing exports requires the construction of the ―new‖ extended input-output tables and models. Such analysis is far more complicated than the
NU
national and multi-regional embodied emission analysis for China. Some studies that used an extended framework include Dietzenbacher et al. (2012), Su et al. (2013), Weitzel and Ma
MA
(2014), Xia et al. (2015) and Jiang et al. (2015). Their results are summarized in Table 1. Among them, only two papers (Su et al., 2013; Xia et al., 2015) further applied the SDA analysis to study the driving forces behind the embodied emission changes in normal and
D
processing exports.
TE
As revealed in the report by WTO and IDE-JETRO (2011), some forms of processing trade can be found in over 130 countries. Due to data constraints, the studies shown in Table 1,
CE P
generally use one or two years of data in the analysis. Currently, no study reports the timeseries estimate of the embodiment in a country’s normal and processing exports. From Table 1, the sector classification level is also found to vary from 28 to 104 sectors. Recent studies (e.g. Su et al., 2010) indicate that sector aggregation has significant impacts on the embodied
AC
emission estimates, especially at the sectoral level. Generally speaking, a higher sector disaggregation level is preferred in empirical studies (Su et al., 2010; Lenzen, 2013; Bouwmeester and Oosterhaven, 2013; De Koning et al., 2015). For SDA studies, it is also important to use the shorter time interval to reduce the potential impacts from temporal aggregation (Su and Ang, 2012b). Very few of the studies on China’s energy and emissions use the time-series dataset in SDA analysis. This paper is an attempt to construct time-series (2006-2012) estimates of China’s embodied emissions in normal and processing exports at the detailed 135-sector level. With these estimates, we can further investigate the driving forces behind the embodied emission changes using the SDA, and also discuss the contributions to the emission efficiency improvements by sector and by export types. Section 2 of the paper explains the estimation of emissions embodied in trade using the extended I-O framework, and the driving forces behind the embodied emission changes using the additive SDA framework. The numerical results of —3—
ACCEPTED MANUSCRIPT the empirical study on China’s embodied emissions from 2006 to 2012 are presented in Section 3. The final section summaries the paper’s main findings and conclusions.
T
2. Extended I-O Framework
IP
2.1 Emissions Embodied in Normal and Processing Exports
In order to account for processing trade in embodied emission studies, the traditional I-
SC R
O table compiled by the National Department of Statistics must be disaggregated into the extended I-O table. Su and Ang (2013) further indicate that non-competitive imports assumptions should be included to avoid overestimating the embodied emissions in trade.
NU
Table 2 shows the structure of the extended I-O table with processing trade and noncompetitive imports. This structure was first introduced by Chen et al. (2001) for the
MA
purposes of analyzing China’s domestic value added and employment induced by its exports. Recently, this extended I-O structure has been used in embodied emission studies (e.g. Dietzenbacher et al., 2012; Su et al., 2013; Weitzel and Ma, 2014; Xia et al., 2015; Jiang et
D
al., 2015).
TE
The economic extended I-O model with processing trade can be formulated as:
Z dp yd yne Add 1 y 0 pe 0
CE P
x y pe Z dd y pe 0
Adp x y pe yd yne 0 y pe y pe
(1)
where x is the vector of total outputs, Z dd is the matrix of domestic intermediate demands for domestic use and normal exports, Z dp is the matrix of domestic intermediate demands for
AC
processing exports, yd is the vector of domestic final demand, yne is the vector of normal exports, y pe is the vector of processing exports, Add Z dd diag ( x y pe ) domestic
production
coefficients
for
domestic
use
and
1
normal
is the matrix of exports,
and
Adp Z dp diag ( y pe ) is the matrix of domestic production coefficients for processing 1
exports. Rearranging Eq. (1) leads to the following equation for I-O analysis with processing trade: 1
x y pe I Add Adp yd yne I y pe y pe 0 I Add 1 I Add 1 Adp yd yne 0 I y pe Ldd Adp yd yne L dd I y pe 0 —4—
(2)
ACCEPTED MANUSCRIPT where Ldd I Add represents the domestic Leontief inverse matrix for domestic use and 1
normal exports. With the emission intensity vectors f d and f e representing the CO2 emissions per unit
T
of value of industry output for domestic use/normal exports and processing exports
IP
respectively, the total amount of CO2 emissions from production can be formulated as
L f p ' dd 0
fd '
L f p ' dd 0
Ldd Adp yd yne 0 I 0 0 y pe
NU
SC R
fd '
x y pe f p ' y pe Ldd Adp yd yne I y pe
Ctot Ctot ,d Ctot , p f d '
(3)
MA
f d ' Ldd yd f d ' Ldd yne f d ' Ldd Adp f p ' y pe Cd Cne C pe
D
where Ctot ,d f d ' x y pe is the total CO2 emissions for domestic use and normal exports,
TE
Ctot , p f p ' y pe is the total CO2 emissions for processing exports, Cd f d ' Ldd yd is the emissions embodied in domestic final demand, Cne f d ' Ldd yne is the emissions embodied in
CE P
normal exports, and C pe f d ' Ldd Adp f p ' y pe is the emissions embodied in processing exports.
AC
2.2 Structural Decomposition Analysis of Embodied Emission Changes With the extended I-O model in Table 2 and embodied emissions in Eq. (3), the changes of embodied emissions in normal exports and processing exports between t1 and t2 ( t1 t2 ) t2 t1 C pe can be formulated as Cne Cnet2 Cnet1 and C pe C pe . Using the simplest four-factor
decomposition identity, those changes of embodied emissions can be decomposed into four sub-effects, i.e.
Cne Cnet2 Cnet1 t2 t2 Ltdd2 Ltdd2 Adp yne t f ' f ' f d1 ' I 0 0 f t2 ' Ldt2 ynet2 f t1 ' Ldt1 ynet1 t2 d
t2 p
1 Ltdd f ' 0
t1 p
t1 t1 1 yne Ltdd Adp I 0
t2 t1 t1 t1 t1 f t2 ' Ldt2 snet2 ytot , ne f ' Ld sne ytot , ne
Ceint,ne (t1 , t2 ) Clstr,ne (t1 , t2 ) Cdstr,ne (t1 , t2 ) Cdtot,ne (t1 , t 2 ) —5—
(4)
ACCEPTED MANUSCRIPT C pe C tpe2 C tpe1 t2 Lt2 Ltdd2 Adp 0 t f pt2 ' dd y t2 f d 1 ' I pe 0 t2 t1 f t2 ' Ldt2 y pe f t1 ' Ldt1 y pe
t1 1 0 Ltdd Adp t I y pe1
Lt1 f pt1 ' dd 0
(5)
T
f dt2 '
IP
t2 t1 t1 t1 t1 f t2 ' Ldt2 s pet2 ytot , pe f ' Ld s pe ytot , pe
Ceint, pe (t1 , t2 ) Clstr,pe (t1 , t2 ) Cdstr,pe (t1 , t2 ) Cdtot,pe (t1 , t2 )
SC R
where n is the number of sectors in the traditional I-O table, f t fi t vector of emission intensity at time t , Ldt Ldt ,ij 2 n1
is the extended
is the extended domestic Leontief
and s pet s pet , j
NU
inverse matrix at time t , snet snet , j
2 n2 n
2 n1
2 n1
are the extended vectors of
t t normal and processing export structures at time t , ytot , ne and ytot , pe are the total amounts of
MA
normal exports and processing exports at time t ; Ceint, (t1 , t2 ) is the emission intensity effect,
Clstr, (t1 , t2 ) is the Leontief structure effect, Cdstr, (t1 , t2 ) is the final demand (export)
TE
between t1 and t2 ( t1 t2 ).
D
structure effect, and Cdtot, (t1 , t2 ) is the total final demand (export) effect in the period
Using the LMDI-I approach (Ang, 2005; Ang et al., 2010; Su and Ang, 2012, 2014), the
CE P
effects in Eqs. (4-5) can be calculated as
f i t2 Ceint, (t1 , t2 ) L(C , C ) ln t1 Ceint, ,i (t1 , t2 ) i, j fi i
(6)
Ldt2,ij Clstr, (t1 , t2 ) L(C , C ) ln t1 L i, j d ,ij
Clstr, ,ij (t1 , t2 ) i, j
(7)
s t,2j Cdstr, (t1 , t2 ) L(C , C ) ln t1 s i, j ,j
Cdstr, , j (t1 , t2 ) j
(8)
AC
t2 ,ij
t2 ,ij
t2 ,ij
t1 ,ij
t1 ,ij
t1 ,ij
t2 ytot Cdtot, (t1 , t2 ) L(C , C ) ln t1 , y i, j tot , t2 ,ij
t1 ,ij
(9)
t where L(a, b) (a b) (ln a ln b) is the logarithmic mean function, C t,ij fi t Ldt ,ij s t, j ytot , is
the embodied CO2 emissions from sub-category (i, j ) , Ceint, ,i (t1 , t2 ) is sector i ’s emission intensity effect, Clstr, ,ij (t1 , t2 ) is Leontief structure effect by sub-element Ld ,ij , Cdstr, , j (t1 , t2 ) is sector j ’s final demand structure (exports) effect. Detailed comparisons of different decomposition approaches in the context of additive SDA can be found in Su and Ang (2012). —6—
ACCEPTED MANUSCRIPT
3. Empirical Study 3.1 Data and Assumptions
T
The 2007 Chinese I-O tables (NBS, 2009), with 135 sector classifications, were used to
IP
construct the extended I-O model given in Table 2.1 Based on the IPCC (2006) guidelines2, the CO2 emissions from 44 sectors in 2007 were estimated from the data published in the
SC R
China Energy Statistical Yearbooks (NBS, 2007-2013a). The ―Data Treatment Scheme 2‖ proposed in Su et al. (2010) was used to disaggregate the energy data to match the detailed IO classifications in the I-O table to ensure comparability. Thus the traditional environmental
NU
I-O tables for China in 2007 have 135 sectors.
The first step was to estimate the Chinese extended I-O tables as outlined in Table 2.
MA
These are not officially published. We followed the procedures of minimizing the information loss function in the disaggregation proposed in Su et al. (2013) to construct the extended I-O tables.3 The datasets used in the disaggregation included the share vector of
D
processing trade (GAC, 2007-2013; Koopman et al., 2012) and economic indicators (such as
TE
output and value added) of China’s state-owned/state-holding industrial enterprises and foreign invested enterprises (NBS, 2007-2013a). The detailed formula used in the
CE P
disaggregation can be found in Appendix A in Su et al. (2013). The emission intensities for processing exports were assumed to be the same as the representative processing trade regions (such as Guangdong, Fujian and Jiangsu Provinces) 4 in China. The detailed steps for 1
AC
Some studies (such as Lau et al., 2010; Dietzenbacher et al., 2012; Xia et al., 2015) further differentiate the domestic use and normal exports into two parts. Unfortunately, their datasets and techniques used in the disaggregation are not clearly addressed. In this paper, we follow the same two-part structure used by Koopman et al. (2012), Su et al. (2013) and Jiang et al. (2015). 2
Alternatively, Chinese specific emission factors can be used in estimating China’s carbon emissions. See Liu et al. (2015) for example. 3
In the literature, there are some disaggregation methods such as Lenzen (2011) and Lindner et al. (2012). However, they are not designed specifically for estimating the extended I-O tables with processing trade. Other disaggregation methods for such extended I-O tables include Koopman et al. (2012) and Lau et al. (2010). 4
Based on regional statistical yearbooks, the processing trade from these three representative account for around 64.2% of China’s total processing exports and 68.3% of China’s total processing imports. In 2007, the aggregate CO2 intensity (CO2 per total output) for the manufacturing sectors in Guangdong, Fujian and Jiangsu Provinces were 4.82 kg CO2/1000 RMB, 9.18 kg CO2/1000 RMB and 7.36 kg CO2/1000 RMB, while the national average CO2 intensity for manufacturing was 11.36 kg CO2/1000 RMB. Following the same approach in Su et al. (2013), we assume the direct emission intensity of processing exports the same as the average emission intensity of major processing trade regions in China. Similar approach is used in Su et al. (2013) to estimate the emission intensity for processing exports in 1997 and 2002. Comparing with the 2007 estimates reported in Jiang et al. (2015), the total emission intensity of processing exports is 0.089 tone CO 2 per 1000 RMB (0.062-0.069 in Jiang et al., 2015), while that of normal exports is 0.232 tone CO2 per 1000 RMB (0.205-0.239 in Jiang et al., 2015). —7—
ACCEPTED MANUSCRIPT estimating the sectoral emissions at China’s regional level can be found in Appendix A in Su and Ang (2010). The final environmental extended I-O tables for China in 2007 have 270 sectors.
T
The second step was to estimate the time-series of emission intensity vector f t ,
IP
t domestic Leontief inverse matrix Ldt , and processing exports vectors ynet and y pe for the
SC R
period 2006-2012. The time-series of sectoral CO2 emissions for 2006-2012 were estimated on the basis of energy data published in the China Energy Statistical Yearbooks (NBS, 20072013a). The time-series of sectoral value added and total output values for 2006-2012 were obtained from the China Statistical Yearbooks (NBS, 2007-2013b) and China Industry
NU
Economy Statistical Yearbooks (NBS, 2007-2013c). The time-series for sectoral processing trade values for 2006-2012 were compiled from the China Customs Statistical Yearbooks
MA
(GAC, 2007-2013) and China Trade and External Economic Statistical Yearbooks (NBS, 2007-2013d). All the values were deflated to the constant 2007 price using the price index given in China Price Statistical Yearbooks (NBS, 2007-2013e).
D
The domestic production coefficient matrix in 2007 was updated to estimate the time-
TE
series of the domestic production coefficient matrix and domestic Leontief inverse matrix for 2006-2012. The following assumptions were used in estimating each sector’s secondary
CE P
domestic input coefficients: (a) the summation of domestic inputs coefficient equals to one minus value added coefficient and imports coefficient, (b) the major energy input coefficient (e.g. coal, oil, gas and electricity) is calculated using the physical energy consumption times
AC
the energy price indicator dividing by the sectoral output, and (c) the relative structure of the remaining inputs coefficients are the same as the structure in 2007.
3.2 China’s Emission Embodiment in 2007 We used the traditional and extended I-O models in Eq. (3) to estimate China’s carbon emission embodiments in 2007. The total amount of CO2 emissions was allocated to the emissions embodied in various types of final demand, including domestic final demand and exports. Domestic final demand includes five sub-categories: rural consumption, urban consumption, government consumption, gross fixed capital formation and inventory change; exports include normal exports and processing exports. The embodiment results are shown in Table 3. Both traditional and extended I-O models show that China’s carbon emissions were mainly driven by investment and exports. In 2007, around 39.79% (traditional I-O model) or —8—
ACCEPTED MANUSCRIPT 41.47% (extended I-O model) of total CO2 emissions were embodied in gross fixed capital formation, while around 27.67% (traditional I-O model) or 24.73% (extended I-O model) of total CO2 emissions were embodied in international exports. For three consumption
T
categories, urban consumption contributed more than 13%, and government consumption and
IP
rural consumption accounted for more than 5% and 4% of total CO2 emissions, respectively. The emissions embodied in rural and urban consumption are indirect carbon emissions. When
SC R
the CO2 emissions from direct energy use by rural and urban households shown in the last two rows of Table 3 are included, rural household consumption accounts for more than 9% while urban household consumption accounts for more than 16% of total CO2 emissions.
NU
The differences between the results using two I-O models are given in the last column of Table 3. The negative value means that when using the traditional I-O model, the
MA
embodiment estimates are under-estimated; the opposite means the over-estimation of the carbon embodiment. The finding is consistent with the finding in previous studies (e.g. Dietzenbacher et al., 2012 and Su et al., 2013) that the emissions embodied in processing
D
exports are over-estimated by more than 70% (617.73/356.09-1=73.48%) while the
TE
remaining embodiments are under-estimated. Compared with the results using detailed sector classifications for 1997 and 2002 discussed in Su et al. (2013), the differences between the
CE P
two sets of embodiment estimates become smaller. The main reason is that some processing services have already been deducted from the standard I-O tables in the compilation of the 2007 China I-O tables.5 Using the extended I-O model, only 24.73% of China’s total CO2 emissions are embodied in its international exports, and only 21.68% of the embodiments in
AC
exports come from processing exports. Detailed comparisons of the embodiment estimates using different models at the sectoral level are shown in Table 4.6 For ease of illustration, we aggregate the 135 sectoral results into 28 sectors. When measuring the relative changes from the estimates using the extended I-O model, the under-estimate of the embodiment in domestic final demand varies from 2% to 11%, the over-estimate of the embodiment in normal exports varies from 2% to 19%, while that in processing exports varies from 37% to 398%. The major absolute changes 5
Processing trade includes ―process and assembling‖ and ―process with imported materials‖ in China’s Customs Statistics. Both are included in the export column in the previous Chinese I-O tables (2002 and before). In 2009, the Department of Statistics adjusted the trade data by removing the ―process and assembling‖ part from exports and imports when publishing the Chinese I-O tables 2007. Detailed explanations can be found in the Appendix of NBS (2009). 6
The sectoral results shown in Table 4-6 and Figure 4 are available for 135 sectors. The relationship between the 135-sector and 28-sector level classification is shown in Table A.1 of Appendix A. Detailed results at the 135-sector level are given in the attached data file. —9—
ACCEPTED MANUSCRIPT are from the embodiment estimates for processing exports in the ―S18-Electric equipment and machinery‖ (41.27 Mt CO2) sector and the ―S19-Telecommunication equipment, computer and other electronic equipment‖ (63.86 Mt CO2) sector, and those for domestic final demand
IP
T
in the ―S25-Construction‖ (-61.70 Mt CO2) sector and ―S28-Services‖ (-30.61 Mt CO2). 3.3 China’s Emissions Embodied in Processing Trade, 2006-2012
SC R
With the estimated time-series data and extended I-O model, the embodied emissions in normal and processing exports for the 2006-2012 period can be obtained using Eq. (3). The results in aggregate are given in Figure 2 and those at the sectoral level are shown in Tables 5
NU
and 6. For ease of illustration, we aggregated the 135 sectoral results into a 28-sector level in Tables 5 and 6.
MA
From Figure 2, the overall trend shows that the embodied emissions in both normal and processing exports first increased (2006-2008), then dropped during the final crisis (20082009), and then rose again after 2009. The emissions embodied in normal exports amounted
D
to about 1,000-1,300 Mt CO2, while those embodied in processing exports were about 250-
TE
350 Mt CO2. When comparing these total embodied emissions in exports with national total CO2 emissions, the percentage was quite stable before and after the global financial crisis.
CE P
The embodied emissions in trade accounted for around 24% for the 2006-2008 period and some 18% for the 2010-2012 period. Some other interesting patterns emerge from the results at the sectoral level. For example, in Tables 5 and 6 it can be seen that the drops in embodied emissions in 2009 were
AC
mainly from the ―S14-Metal smelting and pressing‖ sector for normal exports (accounting for 47% of the total drop in 2008-2009) and the ―S19-Telecommunication equipment, computer and other electronic equipment‖ sector for processing exports (accounting for 34% of the total drop in 2008-2009). On the other hand, some sectors’ embodied emissions increased in 2009, such as for the ―S13-Non-metal minerals products‖ sector for normal exports and the ―S17-Transport equipment‖ sector for processing exports. These embodiment movements can be explained by various factors, including the emission intensity, production structure, exports structure and total exports. This was learned through the SDA analysis discussed in the next subsection.
3.4 Structural Decomposition Analysis of Embodied Emission Changes, 2006-2012 To better understand the driving forces behind the embodied emission changes, we applied the SDA analysis using the formula in Eqs. (6-9) for period 2006-2012. The chaining — 10 —
ACCEPTED MANUSCRIPT approach was adopted to minimize the temporal aggregation effect in the SDA results (Su and Ang, 2012b). The decomposition results for normal exports and processing exports are shown in Figure 3(a) and 3(b).
T
The overall patterns behind the driving factors throughout the period for embodiments
IP
in both normal and processing exports are very similar. Among the four driving factors discussed in Eqs. (4-5), the emission intensity effect played the key role in reducing the
SC R
embodied emissions, while the total export effect contributed the most to the increase in embodied emissions. Of particular note was the drop in embodied emissions in 2009, which was due mainly to the decline in total exports during the financial crisis. With the recovery in
NU
the global economy starting from 2010, China’s total exports steadily increased. After 2010, the emissions embodied in normal exports became stable but those from the processing
MA
exports started to decline in 2012 due to continuous emission efficiency improvements and a slight slowdown in the increase of processing exports. From Eq. (3), the emission intensity effect for embodiment changes in normal exports
D
was determined by the emission intensity improvements ( f d ) in the domestic use/normal
TE
exports production only. However, the emission intensity effect for embodiment changes in the processing exports came from both emission intensity improvements in the domestic
CE P
use/normal exports production ( f d ) and in the processing exports production ( f p ). Their contributions to total emission intensity effect for processing exports are shown in Figure 4. It was found that more than 70% of the contribution came from improvements in the domestic
AC
use/normal exports production.
The emission intensity effects by sector for normal exports and processing exports in each year are shown in Figure 5(a) and 5(b). From the 28-sector level, the largest three contributing sectors were ―S12-Chemicals‖, ―S13-Non-metal minerals products‖ and ―S14Metal smelting and processing‖. Their emission efficiency improvements accounted for more than 50% and 80% of the total improvements in the 2006-2011 and 2011-2012 periods, respectively. Indeed, the ―S14-Metal smelting and processing‖ sector alone contributed more than 70% and 60% of total emission efficiency improvements in the 2011-2012 period for embodiments in normal exports and processing exports, respectively. There was also some increase in emission intensity found in some sectors and time periods, such as the ―S3-Crude petroleum and natural gas products‖ sector, ―S21-Other manufacturing products‖ sector and ―S26-Transport and warehousing‖ sector in 2011 and 2012.
— 11 —
ACCEPTED MANUSCRIPT 4. Discussion and Conclusions This paper constructed time-series extended I-O datasets (2006-2012) to analyze China’s embodied emissions in both normal and processing exports at a detailed sectoral
T
level and applied SDA analysis to understand the driving forces behind the changes of their
IP
embodiment over the whole time period. The empirical results confirm the importance of using the extended I-O model for analyzing the trade-related embodiment, especially for
SC R
processing exports which account for about half of China’s total exports. The period under analysis covers all of the 11th Five Year Plan (2006-2010) period, the global financial crisis (2008-2009) and the first two years of the 12th Five Year Plan (2011-2012). Over the whole
NU
period, energy efficiency measures played the most important role in reducing the emissions and offsetting the increasing demand around the world for China’s exports.
MA
The climate negotiations in Paris in December 2015 highlighted that fact that countries have common but differentiated responsibilities when it comes to climate change. Globalization has made the relations among all countries closer than before through
D
international trade. The phenomenon of ―carbon leakage‖ through embodied emissions in
TE
trade will remain, and very likely increase in absolute volume. Thus it is very important to evaluate the benefits and losses from such leakages, especially when establishing the national
CE P
reduction targets in the INDCs. A key finding in this paper is that China contributed around 1,000 Mt of CO2 emissions each year from 2006 to 2012 through its exports (especially the normal exports) to numerous countries. The emission efficiency improvements in China from
AC
2006 to 2012 helped reduce emissions by about 2,469 Mt CO2 embodied emissions in exports. As developed countries are major importers of embodied emissions from developing countries, it is critical for developed countries to help reduce the carbon missions in developing countries through emission efficiency improvements. This can be achieved through technology transferring and paying the price of carbon when consuming the final products/services in developed countries. Emission efficiency improvements mainly occur through improvements in energy efficiency, optimizing the energy mix through using more renewable energy, and carbon capture and storage. China made significant progress in its 11th Five Year Plan (Price et al., 2011) and 12th Five Year Plan (Li and Wang, 2012). The energy efficiency efforts were focused primarily on the energy intensive sectors. The detailed sectoral direct and embodied emission results give an overview of the situation for the whole economy. By combining the efficiency potential studies and cost benefit analysis for individual sectors, it would be possible to design effective energy and climate policies for China. Our analysis also indicates — 12 —
ACCEPTED MANUSCRIPT that China’s demand is driven by investment and international exports. Although processing exports account for around half of China’s total exports, the energy/emissions caused by the processing exports are much less than those caused by the normal exports. The time series
T
estimates derived in this paper can be further used to analyze the contribution of processing
IP
trade to national aggregate emission intensity reduction (Su and Ang, 2015). In late 2015, after examining the achievements of seven pilot emission trading schemes,
SC R
the Chinese central government announced plans to establish a national emissions trading scheme in 2017. This market mechanism will reduce the mitigation costs for everyone. When it is launched, China will have the largest carbon market in the world, exceeding that of the
NU
EU and US-California. However, due to the large trade volume with other countries, carbon pricing in China will (directly or indirectly) influence the goods/service prices in other world
MA
countries through embodied emissions in exports. The possible impacts can be estimated by multiplying the embodied emissions by the carbon price rate (Su and Ang, 2014)7. Since China already plays an important role in the regional/global supply chain of various products,
D
it would be interesting to expand the existing national-level analysis on different trade types
TE
(normal and processing trade) to regional or world-level analysis for specific industries or even products.
CE P
Due to data constraints, there are very few studies focusing on embodied emissions and SDA studies using time-series data, and capturing both normal and processing exports. This paper uses only the Chinese Input-Output Tables 2007 and annual statistics data to estimate the time-series embodied emissions in 2006-2012. Although Chinese Input-Output Tables
AC
2012 has been published, the detailed custom data 2012 (HS-6 digit level) are not available to estimate the extended Input-Output Tables as shown in Table 2. If these datasets become available in future, the embodied emissions for 2008-2011 can also be estimated using the backward approach and Input-Output Tables 2012. This is an area that deserves further research, not only for China as a whole but also for specific regions in China and for other countries. A similar analysis could also be applied to other indicators, such as energy, water, GHG emissions, pollutant and materials, though there are uncertainties in data treatment and model assumptions that need special attention in empirical studies.
7
For a comprehensive analysis of carbon pricing, more complicated simulation models, such as the computable general equilibrium model discussed in Liang et al. (2016), are required. — 13 —
ACCEPTED MANUSCRIPT Acknowledgements This study is partially supported by the National Natural Science Foundation of China
IP
T
(nos. 71473010 and 71573186).
SC R
Appendix A. Sector Classification
Table A.1 Sectors at the 28-sector level and the number of sectors at the 135-sector level within each sector at the 28-sector level, China (2007)
NU
Sector Name
CE P
TE
D
MA
Agriculture Coal mining, washing and processing Crude petroleum and natural gas products Metal ore mining Non-ferrous mineral mining Manufacture of food products and tobacco processing Textile goods Wearing apparel, leather, furs and related products Sawmills and furniture Paper and paper products, printing and reproduction Petroleum processing, coking and nuclear fuel processing Chemicals Non-metal minerals products Metals smelting and pressing Metal products Common and special equipment Transport equipment Electric equipment and machinery Telecommunication equipment, computer and other electronic equipment Instruments, meters, cultural and office machinery Other manufacturing products Production and supply of electricity and heating power Gas production and supply Water production and supply Construction Transport and warehousing Wholesale and retail trade Services
AC
28-Sector Level ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
— 14 —
135-Sector Level 5 1 1 2 1 14 5 2 2 3 2 9 9 6 1 9 4 5 6 2 2 1 1 1 1 10 4 26
ACCEPTED MANUSCRIPT References
AC
CE P
TE
D
MA
NU
SC R
IP
T
Ang, B.W., 2005. The LMDI approach to decomposition analysis: A practical guide. Energy Policy 33 (7), 867-871. Ang, B.W., Mu, A.R., Zhou, P., 2010. Accounting frameworks for tracking energy efficiency trends. Energy Economics 32 (5), 1209-1219. Bouwmeester, M.C., Oosterhaven, J., 2013. Specification and aggregation errors in environmentally extended input-output models. Environmental and Resource Economics 56 (3), 307-335. Chen, G.Q., Chen, Z.M., 2010. Carbon emissions and resources use by Chinese economy 2007: A 135-sector inventory and input-output embodiment. Communications in Nonlinear Science and Numerical Simulation 15, 3647-3732. Chen, G.Q., Chen, Z.M., 2011. Greenhouse gas emissions and natural resources used by the world economy: Ecological input-output modelling. Ecological Modelling 222 (14), 2362-2376. Chen, G.Q., Guo, S., Shao, L., Li, J.S., Chen, Z.M., 2013. Three-scale input-output modelling for urban economy: Carbon emission by Beijing 2007. Communications in Nonlinear Science and Numerical Simulation 18 (9), 2493-2506. Chen, X., Cheng, L.K., Fung, K.C., Lau, L.J., 2001. The estimation of domestic value-added and employment induced by exports: An application to Chinese exports to the United States. Working Paper, Department of Economics, Stanford University, June 2001. De Koning, A., et al., 2015. Effect of aggregation and disaggregation on embodied material use of products in input-output analysis. Ecological Economics 116, 289-299. Dietzenbacher, E., Pei, J., Yang, C., 2012. Trade, production fragmentation, and China’s carbon dioxide emissions. Journal of Environmental Economics and Management 64, 88-101. Feng, K., et al., 2012. Outsourcing CO2 within China. Proceedings of the National Academy of Sciences of the United States of America 110 (28), 11654-11659. GAC, 2007-2013. China Customs Statistical Yearbook 2006-2012. General Administration of Customs of China, China Customs Press, Beijing. GOSC, 2014. Energy Development Strategy Action Plan (2014-2020). General Office of the State Council in China, 19 November 2014. Guo, J., Zhang, Z., Meng, L., 2012. China’s provincial CO2 emissions embodied in international and interprovincial trade. Energy Policy 42, 486-497. Hawkins, J., Ma, C., Schilizzi, S., Zhang, F., 2015. Promises and pitfalls in environmentally extended input-output analysis for China: A survey of the literature. Energy Economics 48, 81-88. IPCC, 2006. 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate Change. Jiang, X.M., Guan, D.B., Zhang, J., Zhu, K., Green, C., 2015. Firm ownership, China’s export related emissions, and the responsibility issue. Energy Economics 51, 466-474. Koopman, R., Wang, Z., Wei, S.J., 2012. Estimating domestic content in exports when processing trade is pervasice. Journal of Development Economics 99 (1), 178-189. — 15 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Lenzen, M., 2013. Aggregation versus disaggregation in input-output analysis of the environment. Economic Systems Research 23 (1), 73-89. Leontief, W.W., 1970. Environmental repercussions and the economic structure: An inputoutput approach. The Review of Economics and Statistics 52 (3), 262-271. Lau, L.J., 2010. Input-occupancy-output models of the non-competitive type and their application – An examination of the China-US trade surplus. Social Sciences in China, 31 (1), 35-54. Li, J., Wang, X., 2012. Energy and climate policy in China’s twelfth five-year plan: A paradigm shift. Energy Policy 41, 519-528. Liang, Q.M., Wang, T., Xue, M.M., 2016. Addressing the competitiveness effects of taxing carbon in China: domestic tax cuts versus border tax adjustments. Journal of Cleaner Production 112, 1568-1581. Lin, B., Sun, C., 2010. Evaluating carbon dioxide emissions in international trade of China. Energy Policy 38, 613-621. Lin, S., Legault, J., Guan, D., 2012. Disaggregating input-output models with incomplete information. Economic Systems Research 24 (4), 329-347. Liu H., Liu, W., Fan, X., Zhou, W., 2015. Caron emissions embodied in demand-supply chains in China. Energy Economics 50, 294-305. Liu, Z., et al., 2015. Reduced carbon emission estimates from fossil fuel combustion and cement production in China. Nature 524, 335-338. Meng, B., Xue, J., Feng, K., Guan, D., Fu, X., 2013. China’s inter-regional spillover of carbon emissions and domestic supply chains. Energy Policy 61, 1305-1321. Miller, R.E., Blair, P.D., 2009. Input-Output Analysis: Foundations and Extensions (2nd edition). New York, Cambridge University Press. NBS, 2009. 2007 Input-Output Table of China. China Statistics Press, Beijing. NBS, 2007-2013a. China Energy Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013b. China Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013c. China Industry Economy Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013d. China Trade and External Economic Statistical Yearbooks 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013e. China Price Statistical Yearbook 2006-2012. China Statistics Press, Beijing. Peters, G.P., Minx, J.C., Weber, C.L., Edenhofer, O., 2011. Growth in emission transfers via international trade from 1990 to 2008. Proceedings of the National Academy of Sciences of the United States of America 108 (21), 8903-8908. Price, L., et al., 2011. Assessment of China’s energy-saving and emission-reduction accomplishments and opportunities during the 11th Five Year Plan. Energy Policy 39, 2165-2178. Ren, S., Yuan, B., Ma, X., Chen, X., 2014. The impact of international trade on China’s industrial carbon emissions since its entry into WTO. Energy Policy 69, 624-634. Sato, M., 2014. Embodied carbon in trade: A survey of the empirical literature. Journal of Economic Survey 28 (5), 831-861. — 16 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Su, B., Ang, B.W., 2010. Input-output analysis of CO2 emissions embodied in trade: The effects of spatial aggregation. Ecological Economics 70 (1), 10-18. Su, B., Ang, B.W., 2011. Multi-region input-output analysis of CO2 emissions embodied in trade: The feedback effects. Ecological Economics 71, 42-53. Su, B., Ang, B.W., 2012a. Structural decomposition analysis applied to energy and emissions: Some recent developments. Energy Economics 34 (1), 177-188. Su, B., Ang, B.W., 2012b. Structural decomposition analysis applied to energy and emissions: Aggregation issues. Economic Systems Research 24 (3), 299-317. Su, B., Ang, B.W., 2013. Input-output analysis of CO2 emissions embodied in trade: Competitive versus non-competitive imports. Energy Policy 56, 83-87. Su, B., Ang, B.W., 2014a. Input-output analysis of CO2 emissions embodied in trade: A multi-region model for China. Applied Energy 114, 377-384. Su, B., Ang, B.W., 2014b. Attribution of changes in the generalized Fisher index with application to embodied emission studies. Energy 69, 778-786. Su, B., Ang, B.W., 2015. Multiplicative decomposition of aggregate carbon intensity change using input-output analysis. Applied Energy 154, 13-20. Su, B., Ang, B.W., 2016. Multi-region comparisons of emission performance: The structural decomposition analysis approach. Ecological Indicators 67, 78-87. Su, B., Ang, B.W., Low, M., 2013. Input-output analysis of CO2 emissions embodied in trade and the driving forces: Processing and normal exports. Ecological Economics 88, 119125. Su, B., Huang, H.C., Ang, B.W., Zhou, P., 2010. Input-output analysis of CO2 emissions embodied in trade: The effects of sector aggregation. Energy Economics 32 (1), 166-175. United Nations, 1993. Handbook of National Accounting: Integrated Environmental and Economic Account. Studies in Methods, United Nations, New York. Weber, C.L., Peters, G.P., Guan, D., Hubacek, K., 2008. The contribution of Chinese exports to Climate Change. Energy Policy 36, 3572-3577. Weitzel, M., Ma, T., 2014. Emissions embodied in Chinese exports taking into account of the special export structure of China. Energy Economics 45, 45-52. WTO, IDE-JETRO, 2011. Trade patterns and global value chains in East Asia: From trade in goods to trade in task. WTO and IDE-JETRO Publications. WTO, 2015. International Trade Statistics 2015. World Trade Organization, Switzerland. Xia, Y., Fan, Y., Yang, C.H., 2015. Assessing the impact of foreign content in China’s exports on the carbon outsourcing hypothesis. Applied Energy 150, 296-307. Xu, M., Li, R., Crittenden, J.C., Chen, Y., 2011. CO2 emissions embodied in China’s exports from 2002 to 2008: A structural decomposition analysis. Energy Policy 39 (11), 73817388. Xu, Y., Dietzenbacher, E., 2014. A structural decomposition analysis of the emissions embodied in trade. Ecological Economics 101, 10-20. Yan, Y., Yang, L., 2010. China’s foreign trade and climate change: A case study of CO 2 emissions. Energy Policy 38, 350-356. Zhang, Y., Tang, Z., 2015. Driving forces of carbon embodied in China’s provincial exports. Energy Economics 51, 445-454. — 17 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Zhang, Z., Guo, J., Hewings, G.J.D., 2014. The effects of direct trade within China on regional and national CO2 emissions. Energy Economics 46, 161-175.
— 18 —
ACCEPTED MANUSCRIPT
D
MA
NU
SC R
IP
T
List of Figures
AC
CE P
TE
Figure 1. China’s normal and processing exports, 1981-2012
Figure 2. China’s emissions embodied in normal and processing exports, 2006-2012
— 19 —
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
(a) Normal Exports
(b) Processing Exports Figure 3. Results of the additive SDA of China’s embodied emission changes, 2006-2012
— 20 —
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 4. Contribution of emission intensity effect for embodiment changes in processing
AC
CE P
TE
D
exports, 2006-2012
— 21 —
(a) Normal Exports
AC
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
— 22 —
(b) Processing Exports
AC
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 5. Sectoral emission intensity effects by sector, 2006-2012 (Mt CO2)
— 23 —
ACCEPTED MANUSCRIPT List of Tables Table 1. Summary of the estimates of China’s CO2 emissions embodied in processing trade
2002(b)
28
1997
104
2002
104
2007
42
2002 Xia et al. (2015)
Jiang et al. (2015) [2]
TE
D
2007
MA
Su et al. (2013)
Weitzel and Ma (2014)
2007
T
28
IP
2002(a) [1]
CO2 Embodied in Exports in Mt (% of Total CO2 Emissions) Total Normal Processing Exports Exports Exports 429 358 71 (12.6%) (10.5%) (2.1%) 464 368 96 (13.6%) (10.8%) (2.8%) 396.7 349.0 47.7 (12.6%) (11.1%) (1.5%) 563.6 495.0 68.6 (15.8%) (13.9%) (1.9%) 1,630 1,264 366 (28.3%) (21.9%) (6.4%) 403 329 74 (15.5%) (12.6%) (2.8%) 1,284 1,071 213 (27.2%) (22.7%) (4.5%) 1,596 1,309 287 (25.9%) (21.2%) (4.7%)
NU
Dietzenbacher et al. (2012)
Number of Sectors
SC R
Authors
Year of Studied
32 32 42
AC
CE P
Notes: [1] Dietzenbacher et al. (2012) utilize two different approaches, i.e. (a) separate coefficients and (b) identical coefficients, for estimating the emission coefficients of normal and processing exports productions. [2] Jiang et al. (2015) further differentiate the contributions from the Chinese owned enterprises (COEs) and foreign-invested enterprises (FIEs) for each exporting type.
— 24 —
ACCEPTED MANUSCRIPT
MA
NU
SC R
IP
T
Table 2. Structure of extended I-O table with processing trade and non-competitive imports
Table 3. Summary of the estimates of China’s emission embodiments in aggregate, 2007
Domestic Final Demands
Traditional I-O Model (1) 4,243.11 (63.89%) 270.71 (4.08%) 878.17 (13.22%) 348.75 (5.25%) 2,624.74 (39.79%) 102.74 (1.55%) 1,837.62 (27.67%) 1,219.89 (18.37%) 617.73 (9.30%)
D
CE P
Rural Consumption
TE
Unit: Mt CO2
Urban Consumption
AC
Government Consumption
Gross Fixed Capital Formation Inventory Change Exports Normal Exports Processing Exports Household (Direct Energy Use)
Extended I-O Model (2) 4,438.51 (66.83%) 284.22 (4.28%) 925.67 (13.94%) 366.88 (5.52%) 2,754.31 (41.47%) 107.44 (1.62%) 1,642.21 (24.73%) 1,286.13 (19.37%) 356.09 (5.36%)
560.66 (8.44%)
Rural Household
350.74 (5.28%)
Urban Household
209.92 (3.16%)
— 25 —
Difference (3) = (1)-(2) -195.40 (-2.94%) -13.51 (-0.20%) -47.50 (-0.72%) -18.12 (-0.27%) -111.57 (-1.68%) -4.70 (-0.07%) 195.40 (2.94%) -66.24 (-1.00%) 261.64 (3.94%)
ACCEPTED MANUSCRIPT
Table 4. Comparison of the estimates of China’s sectoral emission embodiments, 2007
Total
SC R
IP
T
Estimates of Embodiment using Extended I-O Model (Mt CO2) Domestic Normal Processing Final Exports Exports Demand 129.90 6.75 0.08 5.21 6.91 0.00 0.86 2.89 0.03 1.69 1.31 0.09 0.09 1.34 1.12 222.57 18.55 4.05 10.97 140.85 17.99 87.11 58.87 14.67 28.12 27.63 5.15 9.37 16.66 19.30 11.32 10.20 1.63 48.22 97.11 20.05 36.13 127.61 20.60 30.33 235.74 13.05 44.01 72.68 18.13 309.09 90.01 19.55 250.55 40.78 14.85 119.83 56.26 37.04 52.58 30.78 128.18 6.41 17.07 19.72 42.45 23.81 0.82 53.44 1.48 0.00 7.58 0.00 0.00 8.04 0.00 0.00 2,051.13 14.08 0.00 128.08 88.55 0.00 157.32 44.87 0.00 586.09 53.32 0.00 4,438.51 1,286.13 356.09 (66.83%) (19.37%) (5.36%)
NU
MA
D
TE
CE P
S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
AC
Sector ID
Estimates of Embodiment using Traditional I-O Model (Mt CO2) Domestic Normal Processing Final Exports Exports Demand 126.57 6.58 0.14 5.10 6.76 0.00 0.82 2.78 0.16 1.64 1.26 0.27 0.09 1.29 1.75 212.70 17.70 6.80 10.29 131.76 28.14 79.72 53.88 27.27 25.99 25.65 10.33 8.85 15.67 26.41 10.83 9.88 4.44 45.98 93.07 35.32 34.28 122.70 45.46 29.86 231.94 25.39 42.31 69.87 34.26 290.82 84.89 42.55 234.59 38.10 27.65 107.82 50.56 78.31 42.33 24.93 192.04 5.71 14.05 29.06 39.61 22.22 1.96 51.58 1.43 0.00 7.31 0.00 0.00 7.86 0.00 0.00 1,989.44 13.66 0.00 124.25 86.26 0.00 151.29 43.22 0.00 555.48 49.79 0.00 4,243.11 1,219.89 617.73 (63.89%) (18.37%) (9.30%)
— 26 —
ACCEPTED MANUSCRIPT
Table 5. China’s CO2 emissions embodied in sectoral normal exports using extended I-O model, 2006-2012 (Mt CO2)
Total
TE
MA
— 27 —
2011 5.83 2.28 1.34 0.12 0.68 18.03 116.33 59.41 29.83 18.98 5.38 104.12 134.59 150.68 60.54 117.64 49.33 86.81 52.51 25.03 22.41 1.93 0.00 0.00 24.82 81.54 40.69 43.30 1,254.15 (14.72%)
IP
T
2010 5.69 2.57 1.68 0.16 0.75 16.67 111.55 53.42 27.26 16.37 5.38 91.63 117.54 125.51 49.60 95.33 38.79 76.80 45.11 21.24 20.39 1.92 0.00 0.00 26.14 81.74 40.87 49.90 1,124.02 (14.52%)
SC R
2009 5.08 2.56 1.93 0.18 0.60 14.14 100.87 44.07 22.18 13.60 5.22 77.23 159.16 71.25 35.21 72.45 29.02 55.75 31.45 15.44 18.63 1.78 0.00 0.00 19.56 60.96 39.76 39.01 937.11 (12.90%)
NU
2008 5.15 7.27 3.53 1.26 1.31 16.65 119.79 50.69 27.18 16.15 11.84 96.91 116.87 232.95 77.55 100.90 44.55 65.84 36.67 19.57 22.07 1.70 0.00 0.00 21.88 96.01 40.56 46.21 1,281.06 (18.70%)
D
2007 6.75 6.91 2.89 1.31 1.34 18.55 140.85 58.87 27.63 16.66 10.20 97.11 127.61 235.74 72.68 90.01 40.78 56.26 30.78 17.07 23.81 1.48 0.00 0.00 14.08 88.55 44.87 53.32 1,286.13 (19.37%)
CE P
2006 7.32 8.74 3.34 1.75 1.68 19.53 133.92 57.16 26.09 15.26 12.43 92.81 136.54 196.96 61.77 72.61 32.83 47.77 24.92 13.90 24.78 1.42 0.00 0.00 8.99 67.30 49.52 46.55 1,165.90 (18.90%)
AC
Sector ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
2012 4.96 1.65 0.92 0.12 0.67 17.31 114.72 61.86 35.28 21.06 3.78 101.90 148.24 140.01 57.60 112.40 49.36 87.67 53.15 24.30 20.70 1.64 0.00 0.00 18.98 90.98 40.13 37.84 1,247.22 (14.02%)
ACCEPTED MANUSCRIPT
Table 6. China’s CO2 emissions embodied in sectoral processing exports using extended I-O model, 2006-2012 (Mt CO2)
Total
TE
MA
— 28 —
2011 0.05 0.00 0.03 0.00 0.00 2.91 10.62 9.86 3.23 12.80 1.20 13.52 14.79 12.08 12.77 17.62 35.11 32.54 126.62 20.00 0.97 0.00 0.00 0.00 0.00 0.00 0.00 0.00 326.73 (3.83%)
IP
T
2010 0.04 0.00 0.04 0.00 0.00 2.80 11.15 10.14 3.36 12.24 1.55 13.43 13.75 8.03 11.86 17.36 33.69 31.98 121.03 19.01 0.59 0.00 0.00 0.00 0.00 0.00 0.00 0.00 312.04 (4.03%)
SC R
2009 0.04 0.00 0.04 0.00 0.00 2.67 11.54 9.87 3.07 13.33 1.31 12.57 12.72 5.10 9.75 14.40 24.18 28.13 101.13 15.31 0.49 0.00 0.00 0.00 0.00 0.00 0.00 0.00 265.66 (3.66%)
NU
2008 0.06 0.00 0.02 0.01 0.16 3.20 14.54 12.66 4.30 18.93 1.11 18.41 17.25 9.03 14.07 17.53 18.92 33.92 121.50 18.86 0.64 0.00 0.00 0.00 0.00 0.00 0.00 0.00 325.14 (4.75%)
D
2007 0.08 0.00 0.03 0.09 1.12 4.05 17.99 14.67 5.15 19.30 1.63 20.05 20.60 13.05 18.13 19.55 14.85 37.04 128.18 19.72 0.82 0.00 0.00 0.00 0.00 0.00 0.00 0.00 356.09 (5.36%)
CE P
2006 0.08 0.00 0.03 0.10 1.44 4.59 18.56 15.84 5.78 19.73 1.41 19.60 21.69 12.34 20.01 20.00 12.13 34.57 115.87 20.48 1.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 345.28 (5.60%)
AC
Sector ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
2012 0.05 0.00 0.03 0.00 0.00 2.87 10.23 9.14 3.12 11.74 1.14 12.64 15.24 12.82 10.73 16.18 27.78 30.79 121.66 19.88 2.22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 308.26 (3.47%)
ACCEPTED MANUSCRIPT Highlights
Chinese time-series extended input-output dataset (2006-2012) was constructed for trade
The dataset was used to analyze China’s emissions embodied in both normal and processing
IP
T
analysis.
exports.
The structural decomposition analysis was applied to shed light on the driving forces behind the
SC R
changes.
Similar analysis can be applied to other indicators, such as energy, water, GHG, pollutants and
CE P
TE
D
MA
NU
materials.
AC
— 29 —
S0140-9883(16)30243-2 doi: 10.1016/j.eneco.2016.09.006 ENEECO 3436
To appear in:
Energy Economics
Received date: Revised date: Accepted date:
25 April 2016 24 August 2016 6 September 2016
Please cite this article as: Su, Bin, Thomson, Elspeth, China’s carbon emissions embodied in (normal and processing) exports and their driving forces, 2006–2012, Energy Economics (2016), doi: 10.1016/j.eneco.2016.09.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT China’s Carbon Emissions Embodied in (Normal and Processing) Exports and Their Driving Forces, 2006-2012
IP
T
Bin Su *, Elspeth Thomson
SC R
Energy Studies Institute, National University of Singapore, Singapore
NU
Abstract
This paper constructed a time-series extended input-output dataset (2006-2012) to
MA
analyze China’s carbon emissions embodied in both normal and processing exports at a detailed sectoral level. The structural decomposition analysis (SDA) was further applied to shed light on the driving forces behind the changes in their embodied emissions over the
D
entire time period. This empirical study confirms the importance of using the extended model
TE
for analyzing the trade-related embodiment, especially for processing exports. The embodied emissions in both normal and processing exports first increased from 2006 to 2008, then
CE P
dropped during the global financial crisis (2008-2009), and then rose again after 2009. The embodied emissions as a percentage of total CO2 emissions were quite stable before and after the global financial crisis, at around 24% over the 2006-2008 period, and 18% over the 2010-
AC
2012 period. From 2006 to 2012, emission intensity played the key role in reducing the embodied emissions (around 595 Mt CO2), while the total export effect contributed the most to the increase in embodied emissions (around 552 Mt CO2). Similar analysis can be applied to other indicators, such as energy, water, GHG emissions, pollutants and materials.
Keywords: Input-output analysis; Emissions embodied in trade; Structural decomposition analysis; Processing exports; China
*
Corresponding author. Tel.: +65-6601-2075; Fax: +65-6775-1831. E-mail address: [email protected] ; [email protected] (B. Su). —1—
ACCEPTED MANUSCRIPT 1. Introduction In November 2014, the General Office of the State Council (GOSC, 2014) announced ―China’s Energy Development Strategy Action Plan (2014-2020)‖, which provided the
T
guidelines for China’s energy developments during the 12th (2011-2015) and 13th (2016-
IP
2020) Five Year Plans. As the world’s largest energy consumer and CO2 emitter, China is doing all it can to conserve energy and reduce its CO2 emissions. At the COP21 conference
SC R
recently held in Paris (30 Nov – 11 Dec 2015), the countries of the world agreed to limit the temperature increase below 2°C. However, based on the national climate actions plans (intended nationally determined contributions - INDCs) submitted by the governments before
NU
COP21, these actions will not be sufficient to achieve this target.
Over the last decade, hundreds of studies have been carried out examining how
MA
international trade affects countries’ domestic emissions, or ―emissions embodied in trade‖ and result in ―carbon leakage‖ between developed and developing countries through trade. Since 1990, due to globalization, merchandise trade value (exports plus imports) increased
D
from USD 7,090 billion in 1990 to USD 32,732 billion in 2008 (WTO, 2015), while
TE
embodied emissions increased from 4.3 Gt CO2 to 7.8 Gt CO2 over the same period (Peters et al., 2011). China exported more than 20% of its total CO2 emissions after year 2000 (Su and
CE P
Ang, 2013; Sato, 2014). China’s energy efficiency improvements, especially those carried out in the 11th Five Year Plan (2006-2010) period greatly helped reduce the embodied emissions in China’s exports. However, the demand for China’s products continues to increase. Thus it
AC
is important for China to monitor this demand and understand how it affects China’s energy use and resulting CO2 emissions. To date there have been many studies on China’s embodied emissions at the national level, such as Weber et al. (2008), Su et al. (2010), Lin and Sun (2010), Chen and Chen (2010, 2011), Yan and Yang (2010), Xu et al. (2011) and Ren et al. (2014). Most of these were based on the environmental input-output (I-O) framework (Leontief, 1970; Miller and Blair, 2009). A review of them and their estimates for China can be found in Su and Ang (2013), Sato (2014) and Hawkins et al. (2015). Some recent studies look into China’s embodied emissions in both interregional and international trade, such as Su and Ang (2010), Feng et al. (2012), Guo et al. (2012), Meng et al. (2013), Chen et al. (2013), Su and Ang (2014a), Zhang et al. (2014), Liu et al. (2015) and Zhang and Tang (2015). In addition, there have been studies which use the structural decomposition analysis (SDA) to understand the driving forces behind the embodied emission changes. A review of SDA studies can be found at Su and Ang —2—
ACCEPTED MANUSCRIPT (2012a) and recent studies include Su and Ang (2012b, 2013, 2014b, 2015, 2016), Xu and Dietzenbacher (2014), Xia et al. (2015), and Zhang and Tang (2015). There is another interesting feature in China’s international exports. Figure 1 shows that
T
about half of China’s exports are processing exports, meaning exports of end products made
IP
from imported assembling/processing intermediate inputs, and which are exempted from Chinese tariffs. The emissions embodied per dollar of processing exports are found to be
SC R
much lower than emissions embodied per dollar of normal exports (Su et al., 2013). Differentiating the normal and processing exports requires the construction of the ―new‖ extended input-output tables and models. Such analysis is far more complicated than the
NU
national and multi-regional embodied emission analysis for China. Some studies that used an extended framework include Dietzenbacher et al. (2012), Su et al. (2013), Weitzel and Ma
MA
(2014), Xia et al. (2015) and Jiang et al. (2015). Their results are summarized in Table 1. Among them, only two papers (Su et al., 2013; Xia et al., 2015) further applied the SDA analysis to study the driving forces behind the embodied emission changes in normal and
D
processing exports.
TE
As revealed in the report by WTO and IDE-JETRO (2011), some forms of processing trade can be found in over 130 countries. Due to data constraints, the studies shown in Table 1,
CE P
generally use one or two years of data in the analysis. Currently, no study reports the timeseries estimate of the embodiment in a country’s normal and processing exports. From Table 1, the sector classification level is also found to vary from 28 to 104 sectors. Recent studies (e.g. Su et al., 2010) indicate that sector aggregation has significant impacts on the embodied
AC
emission estimates, especially at the sectoral level. Generally speaking, a higher sector disaggregation level is preferred in empirical studies (Su et al., 2010; Lenzen, 2013; Bouwmeester and Oosterhaven, 2013; De Koning et al., 2015). For SDA studies, it is also important to use the shorter time interval to reduce the potential impacts from temporal aggregation (Su and Ang, 2012b). Very few of the studies on China’s energy and emissions use the time-series dataset in SDA analysis. This paper is an attempt to construct time-series (2006-2012) estimates of China’s embodied emissions in normal and processing exports at the detailed 135-sector level. With these estimates, we can further investigate the driving forces behind the embodied emission changes using the SDA, and also discuss the contributions to the emission efficiency improvements by sector and by export types. Section 2 of the paper explains the estimation of emissions embodied in trade using the extended I-O framework, and the driving forces behind the embodied emission changes using the additive SDA framework. The numerical results of —3—
ACCEPTED MANUSCRIPT the empirical study on China’s embodied emissions from 2006 to 2012 are presented in Section 3. The final section summaries the paper’s main findings and conclusions.
T
2. Extended I-O Framework
IP
2.1 Emissions Embodied in Normal and Processing Exports
In order to account for processing trade in embodied emission studies, the traditional I-
SC R
O table compiled by the National Department of Statistics must be disaggregated into the extended I-O table. Su and Ang (2013) further indicate that non-competitive imports assumptions should be included to avoid overestimating the embodied emissions in trade.
NU
Table 2 shows the structure of the extended I-O table with processing trade and noncompetitive imports. This structure was first introduced by Chen et al. (2001) for the
MA
purposes of analyzing China’s domestic value added and employment induced by its exports. Recently, this extended I-O structure has been used in embodied emission studies (e.g. Dietzenbacher et al., 2012; Su et al., 2013; Weitzel and Ma, 2014; Xia et al., 2015; Jiang et
D
al., 2015).
TE
The economic extended I-O model with processing trade can be formulated as:
Z dp yd yne Add 1 y 0 pe 0
CE P
x y pe Z dd y pe 0
Adp x y pe yd yne 0 y pe y pe
(1)
where x is the vector of total outputs, Z dd is the matrix of domestic intermediate demands for domestic use and normal exports, Z dp is the matrix of domestic intermediate demands for
AC
processing exports, yd is the vector of domestic final demand, yne is the vector of normal exports, y pe is the vector of processing exports, Add Z dd diag ( x y pe ) domestic
production
coefficients
for
domestic
use
and
1
normal
is the matrix of exports,
and
Adp Z dp diag ( y pe ) is the matrix of domestic production coefficients for processing 1
exports. Rearranging Eq. (1) leads to the following equation for I-O analysis with processing trade: 1
x y pe I Add Adp yd yne I y pe y pe 0 I Add 1 I Add 1 Adp yd yne 0 I y pe Ldd Adp yd yne L dd I y pe 0 —4—
(2)
ACCEPTED MANUSCRIPT where Ldd I Add represents the domestic Leontief inverse matrix for domestic use and 1
normal exports. With the emission intensity vectors f d and f e representing the CO2 emissions per unit
T
of value of industry output for domestic use/normal exports and processing exports
IP
respectively, the total amount of CO2 emissions from production can be formulated as
L f p ' dd 0
fd '
L f p ' dd 0
Ldd Adp yd yne 0 I 0 0 y pe
NU
SC R
fd '
x y pe f p ' y pe Ldd Adp yd yne I y pe
Ctot Ctot ,d Ctot , p f d '
(3)
MA
f d ' Ldd yd f d ' Ldd yne f d ' Ldd Adp f p ' y pe Cd Cne C pe
D
where Ctot ,d f d ' x y pe is the total CO2 emissions for domestic use and normal exports,
TE
Ctot , p f p ' y pe is the total CO2 emissions for processing exports, Cd f d ' Ldd yd is the emissions embodied in domestic final demand, Cne f d ' Ldd yne is the emissions embodied in
CE P
normal exports, and C pe f d ' Ldd Adp f p ' y pe is the emissions embodied in processing exports.
AC
2.2 Structural Decomposition Analysis of Embodied Emission Changes With the extended I-O model in Table 2 and embodied emissions in Eq. (3), the changes of embodied emissions in normal exports and processing exports between t1 and t2 ( t1 t2 ) t2 t1 C pe can be formulated as Cne Cnet2 Cnet1 and C pe C pe . Using the simplest four-factor
decomposition identity, those changes of embodied emissions can be decomposed into four sub-effects, i.e.
Cne Cnet2 Cnet1 t2 t2 Ltdd2 Ltdd2 Adp yne t f ' f ' f d1 ' I 0 0 f t2 ' Ldt2 ynet2 f t1 ' Ldt1 ynet1 t2 d
t2 p
1 Ltdd f ' 0
t1 p
t1 t1 1 yne Ltdd Adp I 0
t2 t1 t1 t1 t1 f t2 ' Ldt2 snet2 ytot , ne f ' Ld sne ytot , ne
Ceint,ne (t1 , t2 ) Clstr,ne (t1 , t2 ) Cdstr,ne (t1 , t2 ) Cdtot,ne (t1 , t 2 ) —5—
(4)
ACCEPTED MANUSCRIPT C pe C tpe2 C tpe1 t2 Lt2 Ltdd2 Adp 0 t f pt2 ' dd y t2 f d 1 ' I pe 0 t2 t1 f t2 ' Ldt2 y pe f t1 ' Ldt1 y pe
t1 1 0 Ltdd Adp t I y pe1
Lt1 f pt1 ' dd 0
(5)
T
f dt2 '
IP
t2 t1 t1 t1 t1 f t2 ' Ldt2 s pet2 ytot , pe f ' Ld s pe ytot , pe
Ceint, pe (t1 , t2 ) Clstr,pe (t1 , t2 ) Cdstr,pe (t1 , t2 ) Cdtot,pe (t1 , t2 )
SC R
where n is the number of sectors in the traditional I-O table, f t fi t vector of emission intensity at time t , Ldt Ldt ,ij 2 n1
is the extended
is the extended domestic Leontief
and s pet s pet , j
NU
inverse matrix at time t , snet snet , j
2 n2 n
2 n1
2 n1
are the extended vectors of
t t normal and processing export structures at time t , ytot , ne and ytot , pe are the total amounts of
MA
normal exports and processing exports at time t ; Ceint, (t1 , t2 ) is the emission intensity effect,
Clstr, (t1 , t2 ) is the Leontief structure effect, Cdstr, (t1 , t2 ) is the final demand (export)
TE
between t1 and t2 ( t1 t2 ).
D
structure effect, and Cdtot, (t1 , t2 ) is the total final demand (export) effect in the period
Using the LMDI-I approach (Ang, 2005; Ang et al., 2010; Su and Ang, 2012, 2014), the
CE P
effects in Eqs. (4-5) can be calculated as
f i t2 Ceint, (t1 , t2 ) L(C , C ) ln t1 Ceint, ,i (t1 , t2 ) i, j fi i
(6)
Ldt2,ij Clstr, (t1 , t2 ) L(C , C ) ln t1 L i, j d ,ij
Clstr, ,ij (t1 , t2 ) i, j
(7)
s t,2j Cdstr, (t1 , t2 ) L(C , C ) ln t1 s i, j ,j
Cdstr, , j (t1 , t2 ) j
(8)
AC
t2 ,ij
t2 ,ij
t2 ,ij
t1 ,ij
t1 ,ij
t1 ,ij
t2 ytot Cdtot, (t1 , t2 ) L(C , C ) ln t1 , y i, j tot , t2 ,ij
t1 ,ij
(9)
t where L(a, b) (a b) (ln a ln b) is the logarithmic mean function, C t,ij fi t Ldt ,ij s t, j ytot , is
the embodied CO2 emissions from sub-category (i, j ) , Ceint, ,i (t1 , t2 ) is sector i ’s emission intensity effect, Clstr, ,ij (t1 , t2 ) is Leontief structure effect by sub-element Ld ,ij , Cdstr, , j (t1 , t2 ) is sector j ’s final demand structure (exports) effect. Detailed comparisons of different decomposition approaches in the context of additive SDA can be found in Su and Ang (2012). —6—
ACCEPTED MANUSCRIPT
3. Empirical Study 3.1 Data and Assumptions
T
The 2007 Chinese I-O tables (NBS, 2009), with 135 sector classifications, were used to
IP
construct the extended I-O model given in Table 2.1 Based on the IPCC (2006) guidelines2, the CO2 emissions from 44 sectors in 2007 were estimated from the data published in the
SC R
China Energy Statistical Yearbooks (NBS, 2007-2013a). The ―Data Treatment Scheme 2‖ proposed in Su et al. (2010) was used to disaggregate the energy data to match the detailed IO classifications in the I-O table to ensure comparability. Thus the traditional environmental
NU
I-O tables for China in 2007 have 135 sectors.
The first step was to estimate the Chinese extended I-O tables as outlined in Table 2.
MA
These are not officially published. We followed the procedures of minimizing the information loss function in the disaggregation proposed in Su et al. (2013) to construct the extended I-O tables.3 The datasets used in the disaggregation included the share vector of
D
processing trade (GAC, 2007-2013; Koopman et al., 2012) and economic indicators (such as
TE
output and value added) of China’s state-owned/state-holding industrial enterprises and foreign invested enterprises (NBS, 2007-2013a). The detailed formula used in the
CE P
disaggregation can be found in Appendix A in Su et al. (2013). The emission intensities for processing exports were assumed to be the same as the representative processing trade regions (such as Guangdong, Fujian and Jiangsu Provinces) 4 in China. The detailed steps for 1
AC
Some studies (such as Lau et al., 2010; Dietzenbacher et al., 2012; Xia et al., 2015) further differentiate the domestic use and normal exports into two parts. Unfortunately, their datasets and techniques used in the disaggregation are not clearly addressed. In this paper, we follow the same two-part structure used by Koopman et al. (2012), Su et al. (2013) and Jiang et al. (2015). 2
Alternatively, Chinese specific emission factors can be used in estimating China’s carbon emissions. See Liu et al. (2015) for example. 3
In the literature, there are some disaggregation methods such as Lenzen (2011) and Lindner et al. (2012). However, they are not designed specifically for estimating the extended I-O tables with processing trade. Other disaggregation methods for such extended I-O tables include Koopman et al. (2012) and Lau et al. (2010). 4
Based on regional statistical yearbooks, the processing trade from these three representative account for around 64.2% of China’s total processing exports and 68.3% of China’s total processing imports. In 2007, the aggregate CO2 intensity (CO2 per total output) for the manufacturing sectors in Guangdong, Fujian and Jiangsu Provinces were 4.82 kg CO2/1000 RMB, 9.18 kg CO2/1000 RMB and 7.36 kg CO2/1000 RMB, while the national average CO2 intensity for manufacturing was 11.36 kg CO2/1000 RMB. Following the same approach in Su et al. (2013), we assume the direct emission intensity of processing exports the same as the average emission intensity of major processing trade regions in China. Similar approach is used in Su et al. (2013) to estimate the emission intensity for processing exports in 1997 and 2002. Comparing with the 2007 estimates reported in Jiang et al. (2015), the total emission intensity of processing exports is 0.089 tone CO 2 per 1000 RMB (0.062-0.069 in Jiang et al., 2015), while that of normal exports is 0.232 tone CO2 per 1000 RMB (0.205-0.239 in Jiang et al., 2015). —7—
ACCEPTED MANUSCRIPT estimating the sectoral emissions at China’s regional level can be found in Appendix A in Su and Ang (2010). The final environmental extended I-O tables for China in 2007 have 270 sectors.
T
The second step was to estimate the time-series of emission intensity vector f t ,
IP
t domestic Leontief inverse matrix Ldt , and processing exports vectors ynet and y pe for the
SC R
period 2006-2012. The time-series of sectoral CO2 emissions for 2006-2012 were estimated on the basis of energy data published in the China Energy Statistical Yearbooks (NBS, 20072013a). The time-series of sectoral value added and total output values for 2006-2012 were obtained from the China Statistical Yearbooks (NBS, 2007-2013b) and China Industry
NU
Economy Statistical Yearbooks (NBS, 2007-2013c). The time-series for sectoral processing trade values for 2006-2012 were compiled from the China Customs Statistical Yearbooks
MA
(GAC, 2007-2013) and China Trade and External Economic Statistical Yearbooks (NBS, 2007-2013d). All the values were deflated to the constant 2007 price using the price index given in China Price Statistical Yearbooks (NBS, 2007-2013e).
D
The domestic production coefficient matrix in 2007 was updated to estimate the time-
TE
series of the domestic production coefficient matrix and domestic Leontief inverse matrix for 2006-2012. The following assumptions were used in estimating each sector’s secondary
CE P
domestic input coefficients: (a) the summation of domestic inputs coefficient equals to one minus value added coefficient and imports coefficient, (b) the major energy input coefficient (e.g. coal, oil, gas and electricity) is calculated using the physical energy consumption times
AC
the energy price indicator dividing by the sectoral output, and (c) the relative structure of the remaining inputs coefficients are the same as the structure in 2007.
3.2 China’s Emission Embodiment in 2007 We used the traditional and extended I-O models in Eq. (3) to estimate China’s carbon emission embodiments in 2007. The total amount of CO2 emissions was allocated to the emissions embodied in various types of final demand, including domestic final demand and exports. Domestic final demand includes five sub-categories: rural consumption, urban consumption, government consumption, gross fixed capital formation and inventory change; exports include normal exports and processing exports. The embodiment results are shown in Table 3. Both traditional and extended I-O models show that China’s carbon emissions were mainly driven by investment and exports. In 2007, around 39.79% (traditional I-O model) or —8—
ACCEPTED MANUSCRIPT 41.47% (extended I-O model) of total CO2 emissions were embodied in gross fixed capital formation, while around 27.67% (traditional I-O model) or 24.73% (extended I-O model) of total CO2 emissions were embodied in international exports. For three consumption
T
categories, urban consumption contributed more than 13%, and government consumption and
IP
rural consumption accounted for more than 5% and 4% of total CO2 emissions, respectively. The emissions embodied in rural and urban consumption are indirect carbon emissions. When
SC R
the CO2 emissions from direct energy use by rural and urban households shown in the last two rows of Table 3 are included, rural household consumption accounts for more than 9% while urban household consumption accounts for more than 16% of total CO2 emissions.
NU
The differences between the results using two I-O models are given in the last column of Table 3. The negative value means that when using the traditional I-O model, the
MA
embodiment estimates are under-estimated; the opposite means the over-estimation of the carbon embodiment. The finding is consistent with the finding in previous studies (e.g. Dietzenbacher et al., 2012 and Su et al., 2013) that the emissions embodied in processing
D
exports are over-estimated by more than 70% (617.73/356.09-1=73.48%) while the
TE
remaining embodiments are under-estimated. Compared with the results using detailed sector classifications for 1997 and 2002 discussed in Su et al. (2013), the differences between the
CE P
two sets of embodiment estimates become smaller. The main reason is that some processing services have already been deducted from the standard I-O tables in the compilation of the 2007 China I-O tables.5 Using the extended I-O model, only 24.73% of China’s total CO2 emissions are embodied in its international exports, and only 21.68% of the embodiments in
AC
exports come from processing exports. Detailed comparisons of the embodiment estimates using different models at the sectoral level are shown in Table 4.6 For ease of illustration, we aggregate the 135 sectoral results into 28 sectors. When measuring the relative changes from the estimates using the extended I-O model, the under-estimate of the embodiment in domestic final demand varies from 2% to 11%, the over-estimate of the embodiment in normal exports varies from 2% to 19%, while that in processing exports varies from 37% to 398%. The major absolute changes 5
Processing trade includes ―process and assembling‖ and ―process with imported materials‖ in China’s Customs Statistics. Both are included in the export column in the previous Chinese I-O tables (2002 and before). In 2009, the Department of Statistics adjusted the trade data by removing the ―process and assembling‖ part from exports and imports when publishing the Chinese I-O tables 2007. Detailed explanations can be found in the Appendix of NBS (2009). 6
The sectoral results shown in Table 4-6 and Figure 4 are available for 135 sectors. The relationship between the 135-sector and 28-sector level classification is shown in Table A.1 of Appendix A. Detailed results at the 135-sector level are given in the attached data file. —9—
ACCEPTED MANUSCRIPT are from the embodiment estimates for processing exports in the ―S18-Electric equipment and machinery‖ (41.27 Mt CO2) sector and the ―S19-Telecommunication equipment, computer and other electronic equipment‖ (63.86 Mt CO2) sector, and those for domestic final demand
IP
T
in the ―S25-Construction‖ (-61.70 Mt CO2) sector and ―S28-Services‖ (-30.61 Mt CO2). 3.3 China’s Emissions Embodied in Processing Trade, 2006-2012
SC R
With the estimated time-series data and extended I-O model, the embodied emissions in normal and processing exports for the 2006-2012 period can be obtained using Eq. (3). The results in aggregate are given in Figure 2 and those at the sectoral level are shown in Tables 5
NU
and 6. For ease of illustration, we aggregated the 135 sectoral results into a 28-sector level in Tables 5 and 6.
MA
From Figure 2, the overall trend shows that the embodied emissions in both normal and processing exports first increased (2006-2008), then dropped during the final crisis (20082009), and then rose again after 2009. The emissions embodied in normal exports amounted
D
to about 1,000-1,300 Mt CO2, while those embodied in processing exports were about 250-
TE
350 Mt CO2. When comparing these total embodied emissions in exports with national total CO2 emissions, the percentage was quite stable before and after the global financial crisis.
CE P
The embodied emissions in trade accounted for around 24% for the 2006-2008 period and some 18% for the 2010-2012 period. Some other interesting patterns emerge from the results at the sectoral level. For example, in Tables 5 and 6 it can be seen that the drops in embodied emissions in 2009 were
AC
mainly from the ―S14-Metal smelting and pressing‖ sector for normal exports (accounting for 47% of the total drop in 2008-2009) and the ―S19-Telecommunication equipment, computer and other electronic equipment‖ sector for processing exports (accounting for 34% of the total drop in 2008-2009). On the other hand, some sectors’ embodied emissions increased in 2009, such as for the ―S13-Non-metal minerals products‖ sector for normal exports and the ―S17-Transport equipment‖ sector for processing exports. These embodiment movements can be explained by various factors, including the emission intensity, production structure, exports structure and total exports. This was learned through the SDA analysis discussed in the next subsection.
3.4 Structural Decomposition Analysis of Embodied Emission Changes, 2006-2012 To better understand the driving forces behind the embodied emission changes, we applied the SDA analysis using the formula in Eqs. (6-9) for period 2006-2012. The chaining — 10 —
ACCEPTED MANUSCRIPT approach was adopted to minimize the temporal aggregation effect in the SDA results (Su and Ang, 2012b). The decomposition results for normal exports and processing exports are shown in Figure 3(a) and 3(b).
T
The overall patterns behind the driving factors throughout the period for embodiments
IP
in both normal and processing exports are very similar. Among the four driving factors discussed in Eqs. (4-5), the emission intensity effect played the key role in reducing the
SC R
embodied emissions, while the total export effect contributed the most to the increase in embodied emissions. Of particular note was the drop in embodied emissions in 2009, which was due mainly to the decline in total exports during the financial crisis. With the recovery in
NU
the global economy starting from 2010, China’s total exports steadily increased. After 2010, the emissions embodied in normal exports became stable but those from the processing
MA
exports started to decline in 2012 due to continuous emission efficiency improvements and a slight slowdown in the increase of processing exports. From Eq. (3), the emission intensity effect for embodiment changes in normal exports
D
was determined by the emission intensity improvements ( f d ) in the domestic use/normal
TE
exports production only. However, the emission intensity effect for embodiment changes in the processing exports came from both emission intensity improvements in the domestic
CE P
use/normal exports production ( f d ) and in the processing exports production ( f p ). Their contributions to total emission intensity effect for processing exports are shown in Figure 4. It was found that more than 70% of the contribution came from improvements in the domestic
AC
use/normal exports production.
The emission intensity effects by sector for normal exports and processing exports in each year are shown in Figure 5(a) and 5(b). From the 28-sector level, the largest three contributing sectors were ―S12-Chemicals‖, ―S13-Non-metal minerals products‖ and ―S14Metal smelting and processing‖. Their emission efficiency improvements accounted for more than 50% and 80% of the total improvements in the 2006-2011 and 2011-2012 periods, respectively. Indeed, the ―S14-Metal smelting and processing‖ sector alone contributed more than 70% and 60% of total emission efficiency improvements in the 2011-2012 period for embodiments in normal exports and processing exports, respectively. There was also some increase in emission intensity found in some sectors and time periods, such as the ―S3-Crude petroleum and natural gas products‖ sector, ―S21-Other manufacturing products‖ sector and ―S26-Transport and warehousing‖ sector in 2011 and 2012.
— 11 —
ACCEPTED MANUSCRIPT 4. Discussion and Conclusions This paper constructed time-series extended I-O datasets (2006-2012) to analyze China’s embodied emissions in both normal and processing exports at a detailed sectoral
T
level and applied SDA analysis to understand the driving forces behind the changes of their
IP
embodiment over the whole time period. The empirical results confirm the importance of using the extended I-O model for analyzing the trade-related embodiment, especially for
SC R
processing exports which account for about half of China’s total exports. The period under analysis covers all of the 11th Five Year Plan (2006-2010) period, the global financial crisis (2008-2009) and the first two years of the 12th Five Year Plan (2011-2012). Over the whole
NU
period, energy efficiency measures played the most important role in reducing the emissions and offsetting the increasing demand around the world for China’s exports.
MA
The climate negotiations in Paris in December 2015 highlighted that fact that countries have common but differentiated responsibilities when it comes to climate change. Globalization has made the relations among all countries closer than before through
D
international trade. The phenomenon of ―carbon leakage‖ through embodied emissions in
TE
trade will remain, and very likely increase in absolute volume. Thus it is very important to evaluate the benefits and losses from such leakages, especially when establishing the national
CE P
reduction targets in the INDCs. A key finding in this paper is that China contributed around 1,000 Mt of CO2 emissions each year from 2006 to 2012 through its exports (especially the normal exports) to numerous countries. The emission efficiency improvements in China from
AC
2006 to 2012 helped reduce emissions by about 2,469 Mt CO2 embodied emissions in exports. As developed countries are major importers of embodied emissions from developing countries, it is critical for developed countries to help reduce the carbon missions in developing countries through emission efficiency improvements. This can be achieved through technology transferring and paying the price of carbon when consuming the final products/services in developed countries. Emission efficiency improvements mainly occur through improvements in energy efficiency, optimizing the energy mix through using more renewable energy, and carbon capture and storage. China made significant progress in its 11th Five Year Plan (Price et al., 2011) and 12th Five Year Plan (Li and Wang, 2012). The energy efficiency efforts were focused primarily on the energy intensive sectors. The detailed sectoral direct and embodied emission results give an overview of the situation for the whole economy. By combining the efficiency potential studies and cost benefit analysis for individual sectors, it would be possible to design effective energy and climate policies for China. Our analysis also indicates — 12 —
ACCEPTED MANUSCRIPT that China’s demand is driven by investment and international exports. Although processing exports account for around half of China’s total exports, the energy/emissions caused by the processing exports are much less than those caused by the normal exports. The time series
T
estimates derived in this paper can be further used to analyze the contribution of processing
IP
trade to national aggregate emission intensity reduction (Su and Ang, 2015). In late 2015, after examining the achievements of seven pilot emission trading schemes,
SC R
the Chinese central government announced plans to establish a national emissions trading scheme in 2017. This market mechanism will reduce the mitigation costs for everyone. When it is launched, China will have the largest carbon market in the world, exceeding that of the
NU
EU and US-California. However, due to the large trade volume with other countries, carbon pricing in China will (directly or indirectly) influence the goods/service prices in other world
MA
countries through embodied emissions in exports. The possible impacts can be estimated by multiplying the embodied emissions by the carbon price rate (Su and Ang, 2014)7. Since China already plays an important role in the regional/global supply chain of various products,
D
it would be interesting to expand the existing national-level analysis on different trade types
TE
(normal and processing trade) to regional or world-level analysis for specific industries or even products.
CE P
Due to data constraints, there are very few studies focusing on embodied emissions and SDA studies using time-series data, and capturing both normal and processing exports. This paper uses only the Chinese Input-Output Tables 2007 and annual statistics data to estimate the time-series embodied emissions in 2006-2012. Although Chinese Input-Output Tables
AC
2012 has been published, the detailed custom data 2012 (HS-6 digit level) are not available to estimate the extended Input-Output Tables as shown in Table 2. If these datasets become available in future, the embodied emissions for 2008-2011 can also be estimated using the backward approach and Input-Output Tables 2012. This is an area that deserves further research, not only for China as a whole but also for specific regions in China and for other countries. A similar analysis could also be applied to other indicators, such as energy, water, GHG emissions, pollutant and materials, though there are uncertainties in data treatment and model assumptions that need special attention in empirical studies.
7
For a comprehensive analysis of carbon pricing, more complicated simulation models, such as the computable general equilibrium model discussed in Liang et al. (2016), are required. — 13 —
ACCEPTED MANUSCRIPT Acknowledgements This study is partially supported by the National Natural Science Foundation of China
IP
T
(nos. 71473010 and 71573186).
SC R
Appendix A. Sector Classification
Table A.1 Sectors at the 28-sector level and the number of sectors at the 135-sector level within each sector at the 28-sector level, China (2007)
NU
Sector Name
CE P
TE
D
MA
Agriculture Coal mining, washing and processing Crude petroleum and natural gas products Metal ore mining Non-ferrous mineral mining Manufacture of food products and tobacco processing Textile goods Wearing apparel, leather, furs and related products Sawmills and furniture Paper and paper products, printing and reproduction Petroleum processing, coking and nuclear fuel processing Chemicals Non-metal minerals products Metals smelting and pressing Metal products Common and special equipment Transport equipment Electric equipment and machinery Telecommunication equipment, computer and other electronic equipment Instruments, meters, cultural and office machinery Other manufacturing products Production and supply of electricity and heating power Gas production and supply Water production and supply Construction Transport and warehousing Wholesale and retail trade Services
AC
28-Sector Level ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
— 14 —
135-Sector Level 5 1 1 2 1 14 5 2 2 3 2 9 9 6 1 9 4 5 6 2 2 1 1 1 1 10 4 26
ACCEPTED MANUSCRIPT References
AC
CE P
TE
D
MA
NU
SC R
IP
T
Ang, B.W., 2005. The LMDI approach to decomposition analysis: A practical guide. Energy Policy 33 (7), 867-871. Ang, B.W., Mu, A.R., Zhou, P., 2010. Accounting frameworks for tracking energy efficiency trends. Energy Economics 32 (5), 1209-1219. Bouwmeester, M.C., Oosterhaven, J., 2013. Specification and aggregation errors in environmentally extended input-output models. Environmental and Resource Economics 56 (3), 307-335. Chen, G.Q., Chen, Z.M., 2010. Carbon emissions and resources use by Chinese economy 2007: A 135-sector inventory and input-output embodiment. Communications in Nonlinear Science and Numerical Simulation 15, 3647-3732. Chen, G.Q., Chen, Z.M., 2011. Greenhouse gas emissions and natural resources used by the world economy: Ecological input-output modelling. Ecological Modelling 222 (14), 2362-2376. Chen, G.Q., Guo, S., Shao, L., Li, J.S., Chen, Z.M., 2013. Three-scale input-output modelling for urban economy: Carbon emission by Beijing 2007. Communications in Nonlinear Science and Numerical Simulation 18 (9), 2493-2506. Chen, X., Cheng, L.K., Fung, K.C., Lau, L.J., 2001. The estimation of domestic value-added and employment induced by exports: An application to Chinese exports to the United States. Working Paper, Department of Economics, Stanford University, June 2001. De Koning, A., et al., 2015. Effect of aggregation and disaggregation on embodied material use of products in input-output analysis. Ecological Economics 116, 289-299. Dietzenbacher, E., Pei, J., Yang, C., 2012. Trade, production fragmentation, and China’s carbon dioxide emissions. Journal of Environmental Economics and Management 64, 88-101. Feng, K., et al., 2012. Outsourcing CO2 within China. Proceedings of the National Academy of Sciences of the United States of America 110 (28), 11654-11659. GAC, 2007-2013. China Customs Statistical Yearbook 2006-2012. General Administration of Customs of China, China Customs Press, Beijing. GOSC, 2014. Energy Development Strategy Action Plan (2014-2020). General Office of the State Council in China, 19 November 2014. Guo, J., Zhang, Z., Meng, L., 2012. China’s provincial CO2 emissions embodied in international and interprovincial trade. Energy Policy 42, 486-497. Hawkins, J., Ma, C., Schilizzi, S., Zhang, F., 2015. Promises and pitfalls in environmentally extended input-output analysis for China: A survey of the literature. Energy Economics 48, 81-88. IPCC, 2006. 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate Change. Jiang, X.M., Guan, D.B., Zhang, J., Zhu, K., Green, C., 2015. Firm ownership, China’s export related emissions, and the responsibility issue. Energy Economics 51, 466-474. Koopman, R., Wang, Z., Wei, S.J., 2012. Estimating domestic content in exports when processing trade is pervasice. Journal of Development Economics 99 (1), 178-189. — 15 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Lenzen, M., 2013. Aggregation versus disaggregation in input-output analysis of the environment. Economic Systems Research 23 (1), 73-89. Leontief, W.W., 1970. Environmental repercussions and the economic structure: An inputoutput approach. The Review of Economics and Statistics 52 (3), 262-271. Lau, L.J., 2010. Input-occupancy-output models of the non-competitive type and their application – An examination of the China-US trade surplus. Social Sciences in China, 31 (1), 35-54. Li, J., Wang, X., 2012. Energy and climate policy in China’s twelfth five-year plan: A paradigm shift. Energy Policy 41, 519-528. Liang, Q.M., Wang, T., Xue, M.M., 2016. Addressing the competitiveness effects of taxing carbon in China: domestic tax cuts versus border tax adjustments. Journal of Cleaner Production 112, 1568-1581. Lin, B., Sun, C., 2010. Evaluating carbon dioxide emissions in international trade of China. Energy Policy 38, 613-621. Lin, S., Legault, J., Guan, D., 2012. Disaggregating input-output models with incomplete information. Economic Systems Research 24 (4), 329-347. Liu H., Liu, W., Fan, X., Zhou, W., 2015. Caron emissions embodied in demand-supply chains in China. Energy Economics 50, 294-305. Liu, Z., et al., 2015. Reduced carbon emission estimates from fossil fuel combustion and cement production in China. Nature 524, 335-338. Meng, B., Xue, J., Feng, K., Guan, D., Fu, X., 2013. China’s inter-regional spillover of carbon emissions and domestic supply chains. Energy Policy 61, 1305-1321. Miller, R.E., Blair, P.D., 2009. Input-Output Analysis: Foundations and Extensions (2nd edition). New York, Cambridge University Press. NBS, 2009. 2007 Input-Output Table of China. China Statistics Press, Beijing. NBS, 2007-2013a. China Energy Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013b. China Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013c. China Industry Economy Statistical Yearbook 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013d. China Trade and External Economic Statistical Yearbooks 2006-2012. China Statistics Press, Beijing. NBS, 2007-2013e. China Price Statistical Yearbook 2006-2012. China Statistics Press, Beijing. Peters, G.P., Minx, J.C., Weber, C.L., Edenhofer, O., 2011. Growth in emission transfers via international trade from 1990 to 2008. Proceedings of the National Academy of Sciences of the United States of America 108 (21), 8903-8908. Price, L., et al., 2011. Assessment of China’s energy-saving and emission-reduction accomplishments and opportunities during the 11th Five Year Plan. Energy Policy 39, 2165-2178. Ren, S., Yuan, B., Ma, X., Chen, X., 2014. The impact of international trade on China’s industrial carbon emissions since its entry into WTO. Energy Policy 69, 624-634. Sato, M., 2014. Embodied carbon in trade: A survey of the empirical literature. Journal of Economic Survey 28 (5), 831-861. — 16 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Su, B., Ang, B.W., 2010. Input-output analysis of CO2 emissions embodied in trade: The effects of spatial aggregation. Ecological Economics 70 (1), 10-18. Su, B., Ang, B.W., 2011. Multi-region input-output analysis of CO2 emissions embodied in trade: The feedback effects. Ecological Economics 71, 42-53. Su, B., Ang, B.W., 2012a. Structural decomposition analysis applied to energy and emissions: Some recent developments. Energy Economics 34 (1), 177-188. Su, B., Ang, B.W., 2012b. Structural decomposition analysis applied to energy and emissions: Aggregation issues. Economic Systems Research 24 (3), 299-317. Su, B., Ang, B.W., 2013. Input-output analysis of CO2 emissions embodied in trade: Competitive versus non-competitive imports. Energy Policy 56, 83-87. Su, B., Ang, B.W., 2014a. Input-output analysis of CO2 emissions embodied in trade: A multi-region model for China. Applied Energy 114, 377-384. Su, B., Ang, B.W., 2014b. Attribution of changes in the generalized Fisher index with application to embodied emission studies. Energy 69, 778-786. Su, B., Ang, B.W., 2015. Multiplicative decomposition of aggregate carbon intensity change using input-output analysis. Applied Energy 154, 13-20. Su, B., Ang, B.W., 2016. Multi-region comparisons of emission performance: The structural decomposition analysis approach. Ecological Indicators 67, 78-87. Su, B., Ang, B.W., Low, M., 2013. Input-output analysis of CO2 emissions embodied in trade and the driving forces: Processing and normal exports. Ecological Economics 88, 119125. Su, B., Huang, H.C., Ang, B.W., Zhou, P., 2010. Input-output analysis of CO2 emissions embodied in trade: The effects of sector aggregation. Energy Economics 32 (1), 166-175. United Nations, 1993. Handbook of National Accounting: Integrated Environmental and Economic Account. Studies in Methods, United Nations, New York. Weber, C.L., Peters, G.P., Guan, D., Hubacek, K., 2008. The contribution of Chinese exports to Climate Change. Energy Policy 36, 3572-3577. Weitzel, M., Ma, T., 2014. Emissions embodied in Chinese exports taking into account of the special export structure of China. Energy Economics 45, 45-52. WTO, IDE-JETRO, 2011. Trade patterns and global value chains in East Asia: From trade in goods to trade in task. WTO and IDE-JETRO Publications. WTO, 2015. International Trade Statistics 2015. World Trade Organization, Switzerland. Xia, Y., Fan, Y., Yang, C.H., 2015. Assessing the impact of foreign content in China’s exports on the carbon outsourcing hypothesis. Applied Energy 150, 296-307. Xu, M., Li, R., Crittenden, J.C., Chen, Y., 2011. CO2 emissions embodied in China’s exports from 2002 to 2008: A structural decomposition analysis. Energy Policy 39 (11), 73817388. Xu, Y., Dietzenbacher, E., 2014. A structural decomposition analysis of the emissions embodied in trade. Ecological Economics 101, 10-20. Yan, Y., Yang, L., 2010. China’s foreign trade and climate change: A case study of CO 2 emissions. Energy Policy 38, 350-356. Zhang, Y., Tang, Z., 2015. Driving forces of carbon embodied in China’s provincial exports. Energy Economics 51, 445-454. — 17 —
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Zhang, Z., Guo, J., Hewings, G.J.D., 2014. The effects of direct trade within China on regional and national CO2 emissions. Energy Economics 46, 161-175.
— 18 —
ACCEPTED MANUSCRIPT
D
MA
NU
SC R
IP
T
List of Figures
AC
CE P
TE
Figure 1. China’s normal and processing exports, 1981-2012
Figure 2. China’s emissions embodied in normal and processing exports, 2006-2012
— 19 —
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
(a) Normal Exports
(b) Processing Exports Figure 3. Results of the additive SDA of China’s embodied emission changes, 2006-2012
— 20 —
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 4. Contribution of emission intensity effect for embodiment changes in processing
AC
CE P
TE
D
exports, 2006-2012
— 21 —
(a) Normal Exports
AC
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
— 22 —
(b) Processing Exports
AC
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 5. Sectoral emission intensity effects by sector, 2006-2012 (Mt CO2)
— 23 —
ACCEPTED MANUSCRIPT List of Tables Table 1. Summary of the estimates of China’s CO2 emissions embodied in processing trade
2002(b)
28
1997
104
2002
104
2007
42
2002 Xia et al. (2015)
Jiang et al. (2015) [2]
TE
D
2007
MA
Su et al. (2013)
Weitzel and Ma (2014)
2007
T
28
IP
2002(a) [1]
CO2 Embodied in Exports in Mt (% of Total CO2 Emissions) Total Normal Processing Exports Exports Exports 429 358 71 (12.6%) (10.5%) (2.1%) 464 368 96 (13.6%) (10.8%) (2.8%) 396.7 349.0 47.7 (12.6%) (11.1%) (1.5%) 563.6 495.0 68.6 (15.8%) (13.9%) (1.9%) 1,630 1,264 366 (28.3%) (21.9%) (6.4%) 403 329 74 (15.5%) (12.6%) (2.8%) 1,284 1,071 213 (27.2%) (22.7%) (4.5%) 1,596 1,309 287 (25.9%) (21.2%) (4.7%)
NU
Dietzenbacher et al. (2012)
Number of Sectors
SC R
Authors
Year of Studied
32 32 42
AC
CE P
Notes: [1] Dietzenbacher et al. (2012) utilize two different approaches, i.e. (a) separate coefficients and (b) identical coefficients, for estimating the emission coefficients of normal and processing exports productions. [2] Jiang et al. (2015) further differentiate the contributions from the Chinese owned enterprises (COEs) and foreign-invested enterprises (FIEs) for each exporting type.
— 24 —
ACCEPTED MANUSCRIPT
MA
NU
SC R
IP
T
Table 2. Structure of extended I-O table with processing trade and non-competitive imports
Table 3. Summary of the estimates of China’s emission embodiments in aggregate, 2007
Domestic Final Demands
Traditional I-O Model (1) 4,243.11 (63.89%) 270.71 (4.08%) 878.17 (13.22%) 348.75 (5.25%) 2,624.74 (39.79%) 102.74 (1.55%) 1,837.62 (27.67%) 1,219.89 (18.37%) 617.73 (9.30%)
D
CE P
Rural Consumption
TE
Unit: Mt CO2
Urban Consumption
AC
Government Consumption
Gross Fixed Capital Formation Inventory Change Exports Normal Exports Processing Exports Household (Direct Energy Use)
Extended I-O Model (2) 4,438.51 (66.83%) 284.22 (4.28%) 925.67 (13.94%) 366.88 (5.52%) 2,754.31 (41.47%) 107.44 (1.62%) 1,642.21 (24.73%) 1,286.13 (19.37%) 356.09 (5.36%)
560.66 (8.44%)
Rural Household
350.74 (5.28%)
Urban Household
209.92 (3.16%)
— 25 —
Difference (3) = (1)-(2) -195.40 (-2.94%) -13.51 (-0.20%) -47.50 (-0.72%) -18.12 (-0.27%) -111.57 (-1.68%) -4.70 (-0.07%) 195.40 (2.94%) -66.24 (-1.00%) 261.64 (3.94%)
ACCEPTED MANUSCRIPT
Table 4. Comparison of the estimates of China’s sectoral emission embodiments, 2007
Total
SC R
IP
T
Estimates of Embodiment using Extended I-O Model (Mt CO2) Domestic Normal Processing Final Exports Exports Demand 129.90 6.75 0.08 5.21 6.91 0.00 0.86 2.89 0.03 1.69 1.31 0.09 0.09 1.34 1.12 222.57 18.55 4.05 10.97 140.85 17.99 87.11 58.87 14.67 28.12 27.63 5.15 9.37 16.66 19.30 11.32 10.20 1.63 48.22 97.11 20.05 36.13 127.61 20.60 30.33 235.74 13.05 44.01 72.68 18.13 309.09 90.01 19.55 250.55 40.78 14.85 119.83 56.26 37.04 52.58 30.78 128.18 6.41 17.07 19.72 42.45 23.81 0.82 53.44 1.48 0.00 7.58 0.00 0.00 8.04 0.00 0.00 2,051.13 14.08 0.00 128.08 88.55 0.00 157.32 44.87 0.00 586.09 53.32 0.00 4,438.51 1,286.13 356.09 (66.83%) (19.37%) (5.36%)
NU
MA
D
TE
CE P
S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
AC
Sector ID
Estimates of Embodiment using Traditional I-O Model (Mt CO2) Domestic Normal Processing Final Exports Exports Demand 126.57 6.58 0.14 5.10 6.76 0.00 0.82 2.78 0.16 1.64 1.26 0.27 0.09 1.29 1.75 212.70 17.70 6.80 10.29 131.76 28.14 79.72 53.88 27.27 25.99 25.65 10.33 8.85 15.67 26.41 10.83 9.88 4.44 45.98 93.07 35.32 34.28 122.70 45.46 29.86 231.94 25.39 42.31 69.87 34.26 290.82 84.89 42.55 234.59 38.10 27.65 107.82 50.56 78.31 42.33 24.93 192.04 5.71 14.05 29.06 39.61 22.22 1.96 51.58 1.43 0.00 7.31 0.00 0.00 7.86 0.00 0.00 1,989.44 13.66 0.00 124.25 86.26 0.00 151.29 43.22 0.00 555.48 49.79 0.00 4,243.11 1,219.89 617.73 (63.89%) (18.37%) (9.30%)
— 26 —
ACCEPTED MANUSCRIPT
Table 5. China’s CO2 emissions embodied in sectoral normal exports using extended I-O model, 2006-2012 (Mt CO2)
Total
TE
MA
— 27 —
2011 5.83 2.28 1.34 0.12 0.68 18.03 116.33 59.41 29.83 18.98 5.38 104.12 134.59 150.68 60.54 117.64 49.33 86.81 52.51 25.03 22.41 1.93 0.00 0.00 24.82 81.54 40.69 43.30 1,254.15 (14.72%)
IP
T
2010 5.69 2.57 1.68 0.16 0.75 16.67 111.55 53.42 27.26 16.37 5.38 91.63 117.54 125.51 49.60 95.33 38.79 76.80 45.11 21.24 20.39 1.92 0.00 0.00 26.14 81.74 40.87 49.90 1,124.02 (14.52%)
SC R
2009 5.08 2.56 1.93 0.18 0.60 14.14 100.87 44.07 22.18 13.60 5.22 77.23 159.16 71.25 35.21 72.45 29.02 55.75 31.45 15.44 18.63 1.78 0.00 0.00 19.56 60.96 39.76 39.01 937.11 (12.90%)
NU
2008 5.15 7.27 3.53 1.26 1.31 16.65 119.79 50.69 27.18 16.15 11.84 96.91 116.87 232.95 77.55 100.90 44.55 65.84 36.67 19.57 22.07 1.70 0.00 0.00 21.88 96.01 40.56 46.21 1,281.06 (18.70%)
D
2007 6.75 6.91 2.89 1.31 1.34 18.55 140.85 58.87 27.63 16.66 10.20 97.11 127.61 235.74 72.68 90.01 40.78 56.26 30.78 17.07 23.81 1.48 0.00 0.00 14.08 88.55 44.87 53.32 1,286.13 (19.37%)
CE P
2006 7.32 8.74 3.34 1.75 1.68 19.53 133.92 57.16 26.09 15.26 12.43 92.81 136.54 196.96 61.77 72.61 32.83 47.77 24.92 13.90 24.78 1.42 0.00 0.00 8.99 67.30 49.52 46.55 1,165.90 (18.90%)
AC
Sector ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
2012 4.96 1.65 0.92 0.12 0.67 17.31 114.72 61.86 35.28 21.06 3.78 101.90 148.24 140.01 57.60 112.40 49.36 87.67 53.15 24.30 20.70 1.64 0.00 0.00 18.98 90.98 40.13 37.84 1,247.22 (14.02%)
ACCEPTED MANUSCRIPT
Table 6. China’s CO2 emissions embodied in sectoral processing exports using extended I-O model, 2006-2012 (Mt CO2)
Total
TE
MA
— 28 —
2011 0.05 0.00 0.03 0.00 0.00 2.91 10.62 9.86 3.23 12.80 1.20 13.52 14.79 12.08 12.77 17.62 35.11 32.54 126.62 20.00 0.97 0.00 0.00 0.00 0.00 0.00 0.00 0.00 326.73 (3.83%)
IP
T
2010 0.04 0.00 0.04 0.00 0.00 2.80 11.15 10.14 3.36 12.24 1.55 13.43 13.75 8.03 11.86 17.36 33.69 31.98 121.03 19.01 0.59 0.00 0.00 0.00 0.00 0.00 0.00 0.00 312.04 (4.03%)
SC R
2009 0.04 0.00 0.04 0.00 0.00 2.67 11.54 9.87 3.07 13.33 1.31 12.57 12.72 5.10 9.75 14.40 24.18 28.13 101.13 15.31 0.49 0.00 0.00 0.00 0.00 0.00 0.00 0.00 265.66 (3.66%)
NU
2008 0.06 0.00 0.02 0.01 0.16 3.20 14.54 12.66 4.30 18.93 1.11 18.41 17.25 9.03 14.07 17.53 18.92 33.92 121.50 18.86 0.64 0.00 0.00 0.00 0.00 0.00 0.00 0.00 325.14 (4.75%)
D
2007 0.08 0.00 0.03 0.09 1.12 4.05 17.99 14.67 5.15 19.30 1.63 20.05 20.60 13.05 18.13 19.55 14.85 37.04 128.18 19.72 0.82 0.00 0.00 0.00 0.00 0.00 0.00 0.00 356.09 (5.36%)
CE P
2006 0.08 0.00 0.03 0.10 1.44 4.59 18.56 15.84 5.78 19.73 1.41 19.60 21.69 12.34 20.01 20.00 12.13 34.57 115.87 20.48 1.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 345.28 (5.60%)
AC
Sector ID S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28
2012 0.05 0.00 0.03 0.00 0.00 2.87 10.23 9.14 3.12 11.74 1.14 12.64 15.24 12.82 10.73 16.18 27.78 30.79 121.66 19.88 2.22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 308.26 (3.47%)
ACCEPTED MANUSCRIPT Highlights
Chinese time-series extended input-output dataset (2006-2012) was constructed for trade
The dataset was used to analyze China’s emissions embodied in both normal and processing
IP
T
analysis.
exports.
The structural decomposition analysis was applied to shed light on the driving forces behind the
SC R
changes.
Similar analysis can be applied to other indicators, such as energy, water, GHG, pollutants and
CE P
TE
D
MA
NU
materials.
AC
— 29 —