Environmental input output analysis emissions embodied in trade and structural decomposition analysis

Issues on sector aggregation in EIOA are analyzed and next the possible effects of sector aggregation on estimating a country’s emissions embodied in trade and their practical implicatio

ENVIRONMENTAL INPUT-OUTPUT ANALYSIS

EMISSIONS EMBODIED IN TRADE AND STRUCTURAL DECOMPOSITION ANALYSIS

SU BIN

(M.Sc., Academy of Mathematics and Systems Science,

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF INDUSTRIAL & SYSTEMS ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2011

Acknowledgements

First and foremost, I would like to express my sincerest gratitude to my supervisor, Professor Ang Beng Wah, who has supported me with his patience and knowledge throughout my PhD research His amiable personality, systematic way of thinking and passion for research have made him a respected advisor who will always have a strong influence in my life Especially, his open-mind has given me the freedom to do the research topics that I am interested in My experience working with him is invaluable to my future research career

I would also like to thank Associate Professor Huang Huei Chuen of the National University of Singapore (NUS), Professor Zhou Peng of the Nanjing University of Aeronautics and Astronautics in China, and Professor Choi Ki-Hong of the Korea Polytechnic University, who gave helpful and constructive comments on Chapter 3 and Chapter 8 of this thesis Special thanks are given to Associate Professor Poh Kim Leng and Dr Ng Tsan Sheng, Adam of the NUS, who served as my oral examination committee and provided valuable comments on my research and thesis writing Many thanks are also given to my friends in the Chinese Academy of Sciences, who helped to collect some data for China which was used in the empirical studies in the research

I would like to express my gratitude to NUS for offering me a research scholarship, NUS libraries for providing abundant books and e-resources collections, and the Department of Industrial & Systems Engineering for the use of its facilities, without any of which I would not have been able to carry out my research I also wish

to thank Ow Lai Chun and Victor Cheo Peng Yim for their excellent administrative support during my PhD studies I also owe my thanks to my friends in laboratory and

classmates within the department Through working together with them, I have thoroughly enjoyed the time participating in coursework, seminars, tutoring, discussions and research work

Last but not least, I would like to thank my wife, Ji Yawen, my grandmother,

my aunts, my parents and parents-in-law for supporting my decision to pursue a PhD degree and for their continuous encouragement throughout the entire process

Summary

With the growing concern about climate change and related energy and environmental issues, the environmental input-output analysis (EIOA) has become an important tool in climate policy analysis Lately, two popular and important areas in EIOA are issues related to emissions embodied in trade (EET) and structural decomposition analysis (SDA) EET studies allow us to understand the embodied emissions flows through international trade and the resulting “carbon leakage” This information is useful for attributing a country’s responsibility to global emissions Through SDA studies, the driving forces of the historical changes of an aggregate indicator, such as carbon dioxide emissions, can be evaluated and quantified This thesis focuses on a number of methodological and application issues in EET and SDA under the EIOA framework

First, a literature survey is presented on studies published in the last one to two decades in each of these two areas Following that, the first part of the thesis focuses

on EET studies Issues on sector aggregation in EIOA are analyzed and next the possible effects of sector aggregation on estimating a country’s emissions embodied

in trade and their practical implications are analytically and empirically investigated The issues of spatial aggregation are then analyzed and a “hybrid emissions embodied

in trade” approach for regional emission studies is proposed The impacts of spatial aggregation on EET emission estimates are also evaluated With increasing interest in using multi-regional input-output (MRIO) models by researchers, a new method called the “stepwise distribution of emissions embodied in trade” is proposed to reveal the mechanism of feedback effects in multi-region EET studies The indirect trade

balance of emissions from bilateral trade between any two countries may be derived using the proposed method

The second part of the thesis deals with SDA studies applied to energy and emissions The recent methodological developments are first studied, which show a shift towards using SDA methods that are ideal in decomposition Arising from this development, four such methods are compared and guidelines on method selection are provided Following that, a comprehensive comparison between SDA and index decomposition analysis (IDA) based on the latest available information is conducted Since three aggregation issues are involved in SDA studies, the impacts on SDA results of the third aggregation, namely temporal aggregation (on top of sector aggregation and spatial aggregation) is examined The Fisher index and its extensions have been used by researchers in multiplicative SDA but the results derived are only given in the aggregate form An attribution analysis of the generalized Fisher index decomposition in SDA is proposed, where the contributions of the individual components at a finer level can be quantified

Finally, a summary of the main findings of our studies is given and suggestions are made for future research

Table of Contents

Acknowledgements………i

Summary……… iii

Table of Contents……… v

List of Tables………… vii

List of Figures………………x

List of Abbreviations………….xiii

Chapter 1 Introduction………………1

1.1 Background……… 1

1.2 Motivations of environmental input-output analysis……….……………3

1.3 Research scope and structure of the thesis…………5

Chapter 2 Literature Review………….…………….9

2.1 Review of EET studies and carbon leakage……… 9

2.2 Review of SDA studies applied to energy and emissions……… 20

2.3 Concluding comments………….29

Part I: Emissions Embodied in Trade (EET)……………31

Chapter 3 Sector Aggregation Issues in EET Analysis……… 32

3.1 Introduction………… 32

3.2 The I-O technique and data issues……… 34

3.3 Issues on sector aggregation……… 38

3.4 Effects of sector aggregation on emissions embodied in exports……… 40

3.5 An illustrative example……… 43

3.6 Emissions embodied in China and Singapore’s exports………… 45

3.7 Discussion and conclusions……… 57

Chapter 4 Spatial Aggregation Issues in EET Analysis……….61

4.1 Introduction……… 61

4.2 Issues on spatial aggregation……… 63

4.3 The I-O technique and model selection……… 64

4.4 Effects of spatial aggregation on emissions embodied in exports………… 68

4.5 Emissions embodied in China’s exports……….70

4.6 Discussion and conclusions………….81

Chapter 5 Feedback Effects in Multi-Region EET Analysis……….84

5.1 Introduction……….84

5.2 I-O models and feedback effects……….86

5.3 Two general approaches and their relationships……….88

5.4 Stepwise distribution of emissions embodied in trade………93

5.5 Measuring indirect absorption patterns and trade balance of emissions……96

5.6 Empirical study………… 100

5.7 Discussions and conclusions……….110

Part II: Structural Decomposition Analysis (SDA)……………113

Chapter 6 Methodological Developments in SDA studies……… 114

6.1 Introduction……… 114

6.2 Methodological developments and issues……….116

6.3 General additive decomposition framework……….119

6.4 Empirical study on China’s CO2 emissions……… 126

6.5 Guidelines on SDA method selection……… 136

6.6 Comparison between SDA and IDA……….138

6.7 Conclusion………….145

Chapter 7 Aggregation Issues in SDA studies…………………146

7.1 Introduction……… 146

7.2 SDA and temporal aggregation……….149

7.3 Temporal aggregation in additive SDA applied to CO2 decomposition… 151

7.4 Empirical study on China’s CO2 emissions……… 155

7.5 Aggregation issues in environmentally extended I-O analysis……….167

7.6 Conclusion……….169

Chapter 8 Attribution Analysis of Generalized Fisher Index……….171

8.1 Introduction……… 171

8.2 General Fisher index decomposition……….173

8.3 Attribution of the generalized Fisher index……… 174

8.4 Multiplicative SDA of emissions embodied in trade……….178

8.5 Empirical study in China……… 180

8.6 Conclusion……….186

Chapter 9 Conclusions………….189

9.1 Summary of the results……… 189

9.2 Future research areas……… 192

References……… 195

Appendix A-L…………209

List of Tables

Table 2.1 EET studies and their specific features, 1994-2010………15

Table 2.2 SDA studies dealing with energy and emissions and their specific features, 1999-2010……….25

Table 3.1 An example: the data……… 44

Table 3.2 An example: the main emission results……… 44

Table 3.3 An example: the effects of sector aggregation (Mt CO2)………45

Table 3.4 Basic emission and socio-economic indicators for China………………46

Table 3.5 Levels of sector aggregation and the number of sectors at each level…47 Table 3.6 Estimates of emissions for exports by sector at level L1 and the effects of sector disaggregation measured using L1 as base, China (2002)… 51

Table 3.7 Estimates of emissions for exports by sector at level L2 and the effects of sector disaggregation measured using L2 as base, China (2002)… 54

Table 3.8 Basic emission and socio-economic indicators for Singapore…………56

Table 4.1 The eight regions of China used in this study and their constituents and identification (ID) codes……… 71

Table 4.2 Basic emission and socio-economic indicators for the eight regions, 1997……….72

Table 4.3 The three regional groups and their identification (ID) codes…………72

Table 4.4 Emissions embodied in China’s exports (in aggregate) estimated at three different spatial aggregation levels using the HEET approach (Mt CO2)……….74

Table 4.5 Emissions embodied in China’s exports (by region) estimated at three

different spatial aggregation levels using the HEET approach (Mt

Table 5.1 Basic emission and socio-economic data by economy/region in the

empirical study, 2000………100 Table 5.2 Trade coverage by economy/region in the empirical study, 2000……101 Table 5.3 Comparisons of the results obtained using the EEBT and MRIO

approaches to measure the “consumption-based” emissions…………104 Table 6.1 Choices of the constant integral values for ideal decomposition by

method……… 123 Table 6.2 Basic emission and socio-economic data for China (GDP given in 2002

prices)………128 Table 6.3 Summary of the results in aggregate for China, 2002-2007 (Mt

CO2)……… 128 Table 6.4 Summary of the results at sub-category level for China, 2002-2007 (Mt

CO2)……… 129 Table 6.5 The outliers from the results at the sub-category level……….132 Table 6.6 An illustration of the “mean-rate-of-change” idea leading to two types of

outliers……… 134 Table 6.7 Further decomposition of the Leontief structure effect into two sub-

effects in aggregate for China, 2002-2007 (Mt CO2)………135 Table 6.8 Developments of IDA and SDA applied to energy and emissions and the

main features of the two techniques……… 140 Table 7.1 Summary of the results in aggregate obtained using non-chaining

analysis and chaining analysis at 38-sector level data for China,

1997-2007 (Mt CO2)……… 158

Table 7.2 Differences of the results in aggregate for non-chaining analysis

compared with for chaining analysis at 38-sector level data for China, 1997-2007 (Mt CO2)……….160 Table 7.3 Comparison of the results in aggregate obtained at different sectoral and

temporal aggregation levels for China, 1997-2007 (Mt CO2)……… 165 Table 8.1 Estimates of the CO2 emissions embodied in China’s exports, 2002-

2007……… 181 Table 8.2 Changes of the emissions embodied in China’s exports by industry

cluster, 2002-2007 (Mt CO2)……….181 Table 8.3 Summary of the decomposition results of embodied emission changes in

aggregate for China, 2002-2007………182 Table 8.4 Four types of sectoral exports amount and embodied emission changes

and the corresponding number of sectors……… 183 Table 8.5 Top 3 sectors with the largest relative exports amount and embodied

emission changes……… 183 Table 8.6 Summary of the attribution results at the sectoral level (given by industry

cluster) for China, 2002-2007……… 185 Table 8.7 Top 10 sectors with the largest sectoral export effects in the attribution

analysis……… 185 Table B.1 L1 sectors and the numbers of sectors within each of the L1 sectors for

L2, L3 and L4, China (2002)……….210 Table B.2 L2 sectors and the numbers of sectors within each of the L2 sectors for

L3 and L4, China (2002)……… 211 Table D.1 Industry sectors, industry clusters, and their identification (ID)

codes……… 213

List of Figures

Figure 1.1 EIOA and E3 systems modeling……… 2 Figure 1.2 Structure of the thesis………7 Figure 2.1 Illustration of embodied emissions in car manufacturing (Source: Ahmad

and Wyckoff, 2003)……….10 Figure 2.2 Illustration of emissions embodied in a country’s international trade…11 Figure 2.3 Summary of the specific features of EET studies shown in Table 2.1…17 Figure 2.4 Summary of the specific features of SDA studies shown in Table 2.2…28 Figure 3.1 Data treatment schemes……… 37 Figure 3.2 Estimates of the emissions embodied in exports for China in 2002 (Mt

Figure 3.3 Dotplot of sector emission intensities at various levels of aggregation,

China (2002)………48 Figure 3.4 Estimates of total emissions embodied in exports at different levels of

sector aggregation expressed as a percentage of the L1 estimates with L1

= 100% 49 Figure 3.5 Results for China analysed at level L1………52 Figure 3.6 Boxplot of differences in emissions embodied in exports for China

(2002) in absolute terms……… 55 Figure 3.7 Boxplot of variations in emissions embodied in exports for China (2002)

in percentage……… 55 Figure 4.1 Basic concepts of the EEBT and HEET approaches……… 66 Figure 4.2 The eight regions in China used in this study and their constituents… 73

Figure 4.3 Total emissions embodied in China’s exports estimated at three different

spatial aggregation levels using the HEET approach……… 74 Figure 4.4 Emissions embodied in China’s exports by region estimated using the

HEET approach……… 77 Figure 4.5 Aggregate emission intensities for the three regional groups as shown in

Table 4.2 by industry cluster estimated using the HEET approach……78 Figure 4.6 Boxplot of industry cluster emission intensities at spatial level L3

estimated using the HEET approach……… 79 Figure 4.7 Emissions embodied in exports at different spatial aggregation levels

based on the simulation results obtained using the HEET approach… 80 Figure 5.1 Relative changes of “consumption-based” emission estimates by step (as

indicated in the axis) in the SWD-EET analysis for each economy….106 Figure 5.2 The ratio of each economy’s indirect absorption to its direct absorption

of China’s emissions embodied in trade……… 108 Figure 5.3 Comparisons of China’s bilateral trade balances of emissions with other

economies by approach……….108 Figure 5.4 Bilateral trade balance of emissions estimated using the EEBT and

MRIO approaches……….110 Figure 6.1 Comparisons of the estimates of the intensity effect at sub-category level

for China………130 Figure 6.2 Comparisons of the estimates of the Leontief structure effect at sub-

category level for China………131 Figure 6.3 Comparisons of the estimates of the final demands effect at sub-category

level for China……… 131 Figure 6.4 Comparisons of the estimates of the input technology effects at sub-

category level for China………135

Figure 6.5 Guidelines on decomposition method selection in SDA……… 137 Figure 7.1 Aggregation issues in environmentally extended I-O analysis……….148 Figure 7.2 Comparisons of the decomposition results in aggregate using different

lengths of time piece for China, 1997-2007……… 159 Figure 7.3 Boxplots of the sectoral emission intensity effects at different lengths of

time piece (results compared to those with T={2,3}) for China, 2007……… 162 Figure 7.4 Comparisons of the sectoral emission intensity effects using different

1997-lengths of time piece for China, 1997-2007……… 163 Figure 7.5 Simulated decomposition results in aggregate for China, 1997-2007, at

different levels of sector aggregation………166 Figure 7.6 Interactions among three aggregation issues……….169 Figure 8.1 Dotplots of sectoral relative export amount (EX) and embodied emission

(EE) changes in China, 2002-2007 (excluding the 14 special sectors in Table 8.4 and the top 3 sectors in Table 8.5)………….184 Figure 8.2 Dotplots of sectoral export effects in the attribution analysis of China’s

embodied emission changes, 2002-2007 (excluding the top 10 sectors in Table 8.7)……… 186 Figure 8.3 Comparisons of the exact and approximate attribution results……….187

List of Abbreviations

CDM Clean Development Mechanism

CO2 Carbon Dioxide

E3 Economy-Energy-Environment

EET Emissions Embodied in Trade

EEBT Emissions Embodied in Bilateral Trade

EIOA Environmental Input-Output Analysis

ETS Emission Trading Scheme

EU ETS European Union Emission Trading Scheme

GDP Gross Domestic Product

GHG Greenhouse Gas

GTAP Global Trade Analysis Project

HEET Hybrid Emissions Embodied in Trade

IDA Index Decomposition Analysis

IDE-JETRO Institute of Developing Economies, Japan External Trade Organization IEA International Energy Agency

I-O Input-Output

IPCC Intergovernmental Panel on Climate Change

KLEM Capital, Labor, Energy and Material

LMDI Logarithmic Mean Divisia Index

MRCI Mean-Rate-of-Change Index

MRIO Multi-Region Input-Output

Mt Million tonne

NBS National Bureau of Statistics

NEI National Emission Inventory

SCP Sustainable Consumption and Production

SDA Structural Decomposition Analysis

SRIO Single-Region Input-Output

SWD-EET Stepwise Distribution of Emissions Embodied in Trade OECD Organisation for Economic Co-operation and Development UNFCCC United Nations Framework Convention on Climate Change WIOT World Input-Output Table

Chapter 1 Introduction

1.1 Background

Global warming and climate change, which refer to the increasing average temperature of the earth’s surface air and oceans and resulting changes in precipitation and increased frequency of extreme weather-related events, are serious global issues faced by humanity The International Panel on Climate Change (IPCC), formed in 1988, concludes that increasing greenhouse gas (GHG) concentrations resulting from human activity, such as fossil fuel burning and deforestation, are responsible for most of the observed temperature increases since the middle of the

20th century (IPCC, 2007) Due to the serious impact from global warming and climate change, how to control and reduce the GHG emissions has become an international problem The Kyoto Protocol, which is a protocol to the United Nations Framework Convention on Climate Change (UNFCCC), was established in December

1997 in Kyoto and went into effect in February 2005 It is the first attempt to control/reduce GHG emissions, especially carbon dioxide (CO2) emissions, at the national level by establishing legally binding commitments for all participating countries

As a major part of GHG emissions, CO2 emissions are mostly a result of the combustion of fossil fuels These emission estimators are often based on the IPCC (1996) or other guidelines using energy consumption data To reduce a country’s CO2emissions, it is therefore necessary to control its energy consumptions However, from

a pragmatic perspective, energy plays an important role in the economic development

of a country Large amounts of energy consumption are required to support the manufacturing processes and peoples’ lifestyles To achieve the reduction targets, it is

essential for the government to understand the mechanisms within the Energy-Environment (E3) systems

Economy-E3 systems modeling and analysis can provide insights into the interrelationships among economic development, energy consumption and resulting environmental pollutions Based on these interrelationships, the government can execute more efficient climate policies to achieve the goal of sustainable consumption and production In E3 systems modeling and analysis, many tools have been developed to support national policy makings (Finnveden and Moberg, 2005) One tool adopted by many countries for national analysis is the environmental input-output analysis (EIOA), based on the environmentally extended input-output (I-O) framework developed by Isard et al (1968) and Leontief (1970) The focus of our research is on using EIOA for E3 systems modeling and analysis as shown in Fig 1.1

Figure 1.1 EIOA and E3 systems modeling

EIOA has increasingly become an important tool in supporting the climate policy makings and economic decisions to achieve the sustainable development of society (Forsund, 1985; Forssell and Polenske, 1998) By introducing the environmental indicators, such as CO2 emissions, into the economic I-O framework,

we are able to understand the relationships between economic development and the

resulting environmental issues The insight into the responsibility of these GHG emissions can help us to understand the role of each country in fighting against climate change Moreover, by focusing on the environmental pollutions from energy consumption, more efficient ways of energy consumption relating to environmental issues and economic development can be pursued through comprehensive analysis

1.2 Motivations of environmental input-output analysis

There are several reasons why EIOA is chosen as the main objective for environmental analysis in this thesis First, I-O analysis is based on the I-O framework introduced by Professor Wassily Leontief in the late 1930s (Leontief, 1936,

1951, 1966) It has solid theoretical foundations, and has been widely used in various areas, such as general economic analysis, energy analysis, environmental/ecological analysis and structural decomposition analysis (Rose and Miernyk, 1989; United Nations, 1993; Miller and Blair, 2009) Recent review of the EIOA publications in the literature can be found in Hoekstra (2010)

Second, the I-O models can help us to understand the interdependence (both direct and indirect relationships) of industries in an economy or region or even in the whole world The literature shows that many I-O models have been developed, including single-region input-output (SRIO) and multi-region input-output (MRIO) models to cater to different situations of analysis (Miller and Blair, 1985) For single-country analysis, the SRIO model can be applied to analyze the economic, energy and emission flows within one country When multi-region data in a country or multi-country’s data are available, the MRIO model can be applied to further analyze the interregional or international flows among these regions or countries It is also

possible to extend the regional I-O studies to worldwide studies if sufficient amounts

of world economies’ I-O data are available

Third, the data to support I-O analysis are available in many countries with the development of national economic database The construction of national I-O tables has become an important and regular-basis task of national statistics departments National benchmark I-O tables are commonly constructed every five years through a comprehensive survey For the other years, the I-O tables can be updated from the latest benchmark I-O table through some adjusting techniques, such as the RAS techniques (Stone, 1961; Bacharach, 1970; United Nations, 1999) In addition to these, some multi-regional or international I-O tables have also been constructed to support regional and international analyses, such as the OECD I-O database (Yamano and Ahmad, 2006), the Asian international I-O tables (IDE, 2006), and the Global Trade Analysis Project (GTAP; Narayanan and Walmsley, 2008)

Finally, the I-O model can be used to evaluate the impacts from pricing the carbon to giving short-term predictions and scenario simulations To control and reduce the world’s CO2 emissions, two major economic strategies have been implemented in some countries: carbon tax (Baranzini et al., 2000) and Emissions Trading Scheme (ETS; Ellerman and Buchner, 2007) With the price models in I-O analysis (Oosterhaven, 1996), it is possible to evaluate the impacts of carbon tax on different sectors (Chung, 2005; Choi et al., 2010) Examples of short-term prediction

or scenario simulations using I-O techniques are shown in Proops et al (1993) and Liang et al (2007) Moreover, together with other economic models and dynamic programming techniques, more complicated simulation models can be constructed to support climate policy makings and economic decisions

The globalization and rapid growth of international trade in the last two decades have enhanced the relationships of the world’s economies How to fight against climate change has become a global issue that no country can avoid EIOA study has shown its advantages in supporting the climate policy makings at national, regional and worldwide levels Thus, it is worth investigating the methodological issues involved in recent EIOA studies and providing input on utilizing these techniques to give support to national, regional and international climate policy makings

1.3 Research scope and structure of the thesis

Although the Kyoto Protocol, adopted in 1997 by 84 countries, sets the target

of returning GHG emissions to the same level as in 1990 by the year 2000 and further reducing it at least 5% by 2008-2012, many of the participating countries are not able

to achieve these targets Some researchers argue that the developed countries can shift the production of energy-intensive goods from territorial areas to developing countries, which results in “carbon leakage” to achieve their reduction on domestic emissions However, carbon leakage does not help to reduce the global GHG emissions, especially CO2 emissions In order to understand the situations behind the international trade and resulting carbon leakage, many studies have been reported, and they are often referred to as “emissions embodied in trade (EET)” analysis using the EIOA framework (Wiedmann et al., 2007) The first research focus of this thesis is to investigate the data and methodological issues involved in EET studies

Although EET studies can analyze the embodied emission flows of a particular year, it is not able to find out the driving forces to the historical emission changes that happened Decomposition analysis has been widely used to evaluate the driving forces of the historical changes of an aggregate indicator in energy, economics,

environmental and other socio-economic areas In the literature, index decomposition analysis (IDA) and structural decomposition analysis (SDA) are two of the most popular decomposition techniques IDA is based on sector data, while SDA needs the I-O data Although developed independently, they are similar in a number of aspects (Hoekstra and van den Bergh, 2003) Due to the simpler data requirements in IDA, its theoretical developments are faster than those of SDA Some recent developments in IDA are reported in Ang (2004a, 2004b, 2005), Ang et al (2003, 2004, 2009, 2010) and Choi and Ang (2011) The second research focus of this thesis is to give a comprehensive comparison of these two techniques with developments which have happened in the last decade, and to implement some well-developed concepts in IDA into SDA studies

In summary, this thesis focuses on two areas in EIOA: EET and SDA studies EET analyzes the embodied emission flows among sectors and regions of a particular year, while SDA deals with the driving forces to the historical changes which have happened over time The structure of this thesis is shown in Fig 2.1

Chapter 2 gives the comprehensive survey of the studies in EET and SDA literature The survey on EET covers 45 empirical studies or publications reported in 1994-2010, while that on SDA includes 43 empirical studies or publications in 1998-

2010 Some research interests on these two areas are summarized in Section 2.3 Following that, the main body of the thesis is divided into two parts: EET and SDA studies, with a focus on issues of aggregation and methodology refinement

Figure 2.1 Structure of the thesis

Part I focuses on EET studies, including Chapters 3-5 Chapter 3 analyzes the sector aggregation issues in EIOA, analytically and empirically investigates the possible effects of sector aggregation on estimating one country’s emissions embodied in trade, and gives suggestions on data treatments and sector level selection Chapter 4 extends the study in Chapter 3 to analyze the second type of aggregation issues, i.e spatial aggregation, proposes the “hybrid emissions embodied in trade” approach for regional emission studies, and further evaluates the impacts of spatial aggregation on EET estimates Chapter 5 introduces a new method called the

“stepwise distribution of emissions embodied in trade” to reveal the mechanism of feedback effects in multi-region EET studies, and provides the indirect trade balance

of emissions from bilateral trade between any two countries using the MRIO approach

Part II deals with the SDA studies applied to energy and emissions Chapter 6 examines the new developments in SDA, compares four ideal decomposition methods

in the context of SDA and provides some guidelines on method selection, and finally discusses the similarities and differences between SDA and IDA based on the latest available information Chapter 7 extends the chaining and non-chaining concept in IDA to study the temporal aggregation issues in SDA, illustrates its impacts on SDA results using China’s data and discusses the possible interactions among three types of aggregation issues, i.e sector aggregation, spatial aggregation and spatial aggregation,

in EIOA study Chapter 8 extends the concept of “additive decomposition of the Fisher index” in the national accounts and proposes an “attribution analysis of the generalized Fisher index” in the context of SDA The proposed attribution analysis allows changes of an aggregate index, derived through the application of the generalized Fisher index decomposition in multiplicative SDA, to be attributed to give the contribution of the individual components at a finer level

In the conclusion, the findings from previous chapters are consolidated and some potential research topics for future study are discussed The contribution of the thesis includes systematic study of three aggregation issues in EIOA and their practical implications, comprehensive study of the methodological developments in SDA and the up-to-date comparison between SDA and IDA, and several new methods for EET and SDA studies The works presented in Chapter 3-6 have been published as four journal papers, while those in Chapter 7-8 have also been submitted to journals for possible publication

Chapter 2 Literature Review

In this chapter, reviews of two areas in EIOA (EET and SDA studies) are given Section 2.1 gives the literature review on EET studies, resulting “carbon leakage” from international trade, and responsibility issues to global emissions Section 2.2 reviews the SDA studies applied to energy and emissions in both earlier and recent studies

2.1 Review of EET studies and carbon leakage

With high energy prices and the potential risks of global warming, how to achieve the goal of Sustainable Consumption and Production (SCP) is a challenging task to many countries During the United Nations’ conference in 1997 in Kyoto, the Kyoto Protocol was proposed to come to an agreement on the policies that would return GHG emissions to the same level as in 1990 by the year 2000 At that time, the UNFCCC only established CO2 emission reduction targets for Annex B parties (or industrialized countries) From the recent environmental data, many of them failed to reach the reduction targets Some researchers explain that since the measuring mechanism of GHG emissions is only based on territorial production, many developed countries can substitute imports of certain goods from developing countries for domestic production in order to decrease the domestic emissions This would result in “carbon leakage” through the international trade (Peters and Hertwich, 2008a,b)

2.1.1 Emissions embodied in trade

Ahmad and Wyckoff (2003) use an example of the car manufacturing process in Fig 2.1 to explain the embodied emissions and resulting “carbon leakage” through

trade The direct and indirect emissions emitted in the process of manufacturing a car are denoted as A, B, and A1 to E1 Here the indirect emissions only consider the emissions related to the generation of electricity Then, the embodied emissions in a car can be calculated as T=A+B+A1+B1+C1+D1+E1 If the entire process takes place

in country “X”, the embodied emissions T are calculated as the domestic emissions for country “X”

Figure 2.1 Illustration of embodied emissions in car manufacturing

(Source: Ahmad and Wyckoff, 2003)

However, if country “X” shifts the first three steps of the manufacturing process

to country “Y”, the domestic emissions in country “X” can be reduced to T0=A+B+A1+B1+C1 from its total domestic emissions by changing its production patterns The required components to assemble the car can be imported from country

“Y” If the manufacturing technology in country “Y” is worse than that in country

“X”, the emissions related to the components production will then be more than T0 With the high growth in international trade in recent years, this phenomenon of

“carbon leakage” can easily happen by shifting the production of energy intensive goods from developed countries to developing countries

Figure 2.2 Illustration of emissions embodied in a country’s international trade

The flows of emissions embodied in international trade, including imports and exports, are illustrated in Fig 2.2 For simplicity, we do not differentiate the imports for manufacturing and domestic consumption, which is the approach of emissions embodied in bilateral trade (Peters, 2008) The emissions emitted during the territorial production are embodied in the domestic products, including the domestic products for domestic consumption C1 and for domestic exports C2 Supposing the emissions embodied in the imports for manufacturing are C3, and those for domestic

consumption are C4 The traditional “production-based” emissions or NEI (National Emission Inventory) is the summation of C1 and C2, while the “consumption-based” emissions equals to the “production-based” emissions minus the emissions embodied

in domestic exports C2 plus the emissions embodied in imports (C3+C4)

2.1.2 Review of recent studies

With the growing concern about climate change and the impacts of growth in international trade, many studies have been reported on energy-related CO2 emissions embodied in international trade The estimators of emissions embodied in trade are also used to calculate the trade balance of emissions and so-called “consumption-based” emissions From these studies, it can be concluded that the CO2 emissions embodied in a country’s international trade measured as a percentage of its total CO2emissions have been increasing over time For example, Ahmad and Wyckoff (2003) study 24 countries using data from 1995 and report that 14% of the CO2 emissions are embodied in trade Peters and Herwtich (2008b) investigate CO2 emissions in 87 countries using data from 2001 and estimate that the overall CO2 emissions embodied

in international trade is 21.5% Developed countries are generally net importers while developing countries are net exporters of CO2 emissions These results have implications on national climate policy and related international negations as have been pointed out in many previous studies

The study of emissions embodied in a country’s imports and exports is not something new There are publications that date back to the 1990s or even earlier, such as Wyckoff and Roop (1994) Although a good number of new studies have been reported in the last decade, there are variations in terms of the scope, technique, dataset and assumptions used among these studies Two important review papers on

EET studies are Wiedmann et al (2007) and Wiedmann (2009a) A literature survey

of EET studies has been conducted and a total of 45 publications from 1994 to 2010 have been collected These publications are listed in Table 2.1, which includes only application studies and those dealing with methodology are excluded All the publications are taken from journals except five working papers, i.e Harris (2001), Hayami and Nakamura (2002), Westin and Wadeskog (2002), Ahmad and Wyckoff (2003) and Nakano et al (2003)

For each study, the application area is shown, the time period of the results are reported and the indicator is analyzed Different sector levels used in the study are also listed The objective includes four categories: “EX”, “IM”, “TB” and “CF” The symbol “EX” indicates the estimates for emissions embodied in exports, while “IM” denotes the estimates for emissions embodied in imports The trade balance of emissions is denoted as “TB”, which is defined as the differences between emissions embodied in exports and imports And “CF” represents the so-called “consumption-based” emissions or “carbon footprint”, which are computed using “production-based” emissions minus the emissions embodied in exports, plus those embodied in imports For three categories under scope, “N” is the national scope, “R” is the regional scope (from two to ten regions), while “W” is the worldwide scope

The most essential part comes from the approach used in these empirical studies

We classified them into four categories: “SRIO”, “EEBT”, “MRIO-1” and “MRIO-n” Category “SRIO” represents using the SRIO model with domestic assumption to estimate a country’s emissions embodied in its exports If further estimates of the emissions embodied in its imports using single-country data is needed, the analyst can only make the assumption that the imports have the same technology as domestic

production With multi-country data, all remaining three approaches can be applied Category “EEBT” represents the approach of emissions embodied in bilateral trade (Peters, 2008) The “EEBT” approach also applies the SRIO model to estimate a country’s emissions embodied in its exports, while emissions embodied in imports are calculated as the sum of all other countries’ emissions embodied in their exports to the country.1 Category “MRIO-1” is the unidirectional trade approach, while category

“MRIO-n” is the full MRIO model approach (Lenzen et al., 2004; Peters, 2008).2

The key features and developments as revealed by the data in Table 2.1 have now been summarized Most of the studies reported in the literature focus on analyzing the developed countries Studies on China and its bilateral trade with some developed countries, such as the US and the UK, emerge as the major concerned topic after 2005 With the availability of multi-region data from the OECD I-O database and GTAP project, more worldwide studies have become available, such as Ahmad and Wyckoff (2003) and Peters and Hertwich (2008b) The sector level used in these empirical studies varies from 10 to around 500 sectors Most of them deal with emissions at the national level despite the size of the countries Therefore, uncertainties from both sector aggregation and spatial aggregation have been included

in the results of embodied emissions

Table 2.1 EET studies and their specific features, 1994-2010

level

Table 2.1 EET studies and their specific features, 1994-2010 (continued)

level

(a) Number of EET studies by scope and time period

(b) Number of EET studies by approach used and time period

Figure 2.3 Summary of the specific features of EET studies shown in Table 2.1

We split the period 1994-2010 into three sub-periods, i.e 1994-1999, 2000-2005 and 2006-2010, to reveal the changing profiles of EET studies Figure 2.3(a) shows an upward trend in the total number of publications per sub-period, and the variation of the scope in these studies As a whole, it is found that the focus has been expanded from national to regional and worldwide For example, the bilateral trade emission studies include Shui and Harriss on US-China trade, Ackerman et al (2007) on Japan-

US trade and Li and Hewitt (2008) on UK-China trade Figure 2.3(b) shows a strong shift from the simple SRIO approach using single-country data to the EEBT and MRIO approaches using the multi-region data Two important worldwide studies are Ahmad and Wyckoff (2003) using the MRIO approach and Peters and Hertwich (2008b) using the EEBT approach

2.1.3 Debates on two responsibilities

How to allocate the equitable responsibilities of GHG inventories between producers and consumers becomes a hot debate in the literature Traditional attitude emphasizes that the producer, which causes the direct environmental impacts in the production process, should take full responsibility for GHG emissions This results in the “production-based” measure to account for territorial GHG emissions without considering trade effect However, importing high energy-intensive goods and service from others countries can shift a country’s responsibility to the exporting country, and this calls for the consideration of GHG emissions embodied in the international trade and “corrections” in emission accounting

Although Wyckoff and Roop (1994) and Kondo et al (1998) comment on the problem of allocation GHG, Munksgaard and Pedersen (2001) are the first to formally identify the need to consider the consumer’s responsibility by providing the concept of “CO2 trade balance” into the negotiation framework when they analyze the different CO2 emissions in Denmark resulting from the import of electricity from Norway Some researchers (Bastianoni et al., 2004; Lenzen et al., 2007) claim that

“consumption is the sole end and purpose of all production”, so consumers should also share the burden of responsibility From then on, the emissions embodied in

trade have been examined in many studies, which show that import and export activities have much influence on the measurement of environmental pressure

The “production-based” NEI only considers the national production and export activities, which has the advantage of ease of computation In contrast, the

“consumption-based” NEI considers the import activities to account for international trade, i.e “Consumption = Production – Exports + Imports”, which has the advantage

of covering more of global emissions with limited participation, increasing mitigation options, naturally encouraging cleaner production, and making policies such as the Clean Development Mechanism (CDM; Peters and Hertwich, 2008a) Particularly, the problems of “carbon leakage” can be lightened by replacing “production-based” NEI with “consumption-based” NEI

It is also reasonable to combine these two measurements to share the responsibilities between the producer and consumer Kondo and Moriguchi (1998) introduce the shared responsibility concept by taking a simple average of

“production-based” NEI and “consumption-based” NEI A trade-off between the two responsibilities is taken by Ferng (2003) and Bastianoni et al (2004) Initial attempts

at sharing responsibilities lead to the problem of double counting, which has been resolved by Gallego and Lenzen (2005) Recent theoretical researches on deriving the meaningful indicators for consumption and production behaviours are conducted in Rodrigues et al (2006) and Lenzen et al (2007) The difference between these two indicators of environmental responsibility mainly focuses on the total or partial transfer of indirect effects (Rodrigues and Domingos, 2008) Recently, Andrew and Forgie (2008) apply the concept in Lenzen et al (2007) to analyze the shared responsibility for New Zealand

2.2 Review of SDA studies

In this section, we give an overview of SDA studies reported prior to 2000 and provide a more detailed analysis of the studies reported from 1999 onwards The objective is to identify trends and summarize key developments in the past three decades

2.2.1 Overview of earlier studies

With the introduction of the extended I-O framework (Leontief, 1970; Hudson and Jorgenson, 1974), applications of SDA have been extended to energy and emissions studies Earlier SDA studies in this application area include Leontief and Ford (1972), Pløger (1983, 1984), Gould and Kulshreshtha (1986), and Gowdy and Miller (1987) More than ten such studies were reported in the 1990s and a detailed review can be found in Hoekstra and van den Bergh (2002) From these early literatures, an important methodological development in SDA is the introduction of the two-tiered KLEM decomposition model in Rose and Chen (1991)3 In this model, changes in the input technology coefficients are decomposed into various KLEM effects based on the KLEM production function Casler and Rose (1998) further extend the concept from studying energy changes to emission changes Some other specific methodological developments are given in Rose and Casler (1996)

In the earlier years the use of traditional ad hoc decomposition methods in SDA

was the norm We define such methods as those based on changing a set of parameters at a time Decomposition methods using the Laspeyres and Paasche indices are such methods.The results given by these methods depend on the base year chosen which could be arbitrary Furthermore, they give non-exact (or imperfect)

3

The acronym “KLEM” stands for four-factor inputs, i.e capital (K), labor (L), energy (E) and material (M), in the neoclassical production function

decomposition and a residual term appears in the results.4 The residual term can be large and its existence often complicates the explanation of the result Since decomposition methods with exact properties help to avoid possible complications in result interpretation, Betts (1989) introduces two general exact decomposition forms.5Thereafter, other studies follow the same idea in selecting a decomposition method See, for example, Wier and Hasler (1999) and Jacobsen (2000)

Both traditional ad hoc decomposition methods and the two general methods

proposed by Betts (1989) are not ideal, i.e the decomposition results depend on the sequence of the factors in the product In index number theory, an index number passing the factor-reversal test is called ideal (Fisher, 1922; Balk, 2003) Ideal decomposition ensures exact decomposition of an aggregate and at the same time satisfies other conditions of the factor-reversal test, and is therefore a stronger condition for decomposition than exact property

Dietzenbacher and Los (1998) suggest using the average of all n! equivalent exact decomposition forms to achieve ideal decomposition However, the D&L method (henceforth will be referred to as D&L in short) is cumbersome to use when the number of main factors studied is large.6 To overcome this problem, some approximate D&L techniques have been proposed, such as by taking the average of the two polar decompositions (Dietzenbacher and Los, 1998) or of any pair of “mirror

4

Decomposition without a residual term is called exact decomposition It is also known as complete or perfect decomposition The term exact or complete decomposition is commonly used in SDA while perfect decomposition in IDA

For instance, when there are five factors (n=5) which is quite common in SDA studies, D&L will

have 5!=120 terms Although Seibel (2003) proves that D&L have only 2n-1 different coefficients for each factor, there are still a lot of calculations required when the number of factors is large

image” decompositions (De Haan, 2001) These approximate D&L techniques fail the factor-reversal test and are therefore not ideal Empirical studies in Dietzenbacher and Los (1998) show that the deviation of sectoral effect estimates, given in terms of the ratio of the average of two polar decomposition to that of full decomposition, ranges from 0.9% to 7.7%

In summary, a development in the early period, i.e till around 2000, was the adoption and slowly growing popularity of SDA in energy and emission studies Methodologically, two worth noting developments are the introduction of the two-tiered decomposition model and of models that are exact or ideal including their various approximate versions

2.2.2 Review of recent studies

In the literature there are other ideal decomposition methods given in the additive form besides the D&L method They include the logarithmic mean Divisia index methods, LMDI-I (Ang et al., 1998; Ang and Liu, 2001) and LMDI-II (Ang and Choi, 1997; Ang et al., 2003), the S/S method (Shapley, 1953; Sun, 1998) and the MRCI method (Chung and Rhee, 2001) These methods, especially the first three, have been widely used in IDA

LMDI-I is related to the Montgomery-Vartia index (Montgomery, 1937; Vartia, 1976) and is also called Montgomery-Vartia decomposition (Choi and Ang, 2003; de Boer, 2008) LMDI-II is related to the Sato-Vartia index (Sato, 1976; Vartia, 1976) and is also called Sato-Vartia decomposition (Choi and Ang, 2003; de Boer, 2009b) Although originally proposed for and used in IDA, LMDI-I has been adopted in several recent SDA studies Examples are de Boer (2008, 2009b), Wachsmann et al (2009), Weber (2009), and Wood (2009) The S/S method was proposed by Sun

(1998) Albrecht et al (2002) independently proposes using the Shapley value for decomposition and Ang et al (2003) shows that the method proposed by Sun (1998) and the Shapley value are identical Lenzen (2006) further shows the similarity between the S/S method in IDA and D&L in SDA in additive decomposition

The review by Hoekstra and van den Bergh (2003) includes SDA and IDA methods available up to 2001 They include the following three methods which give ideal decomposition: refined Divisia (presently known as LMDI-II) and Sun's method (presently known as the S/S method) in IDA, and D&L in SDA These methods were introduced only in the late 1990s so it was too early for Hoekstra and van den Bergh (2003) to report the extent of their application We conducted a literature survey of SDA studies and a total of 43 publications from 1999 to 2010 have been collected These publications are listed in Table 2.2, which includes only application studies and those that deal with methodology are excluded All the publications are journal papers except two working papers, i.e Hoen and Mulder (2003) and Seibel (2003)

All of the 43 publications in Table 2.2 employ additive decomposition In fact

no SDA study on energy and emissions using multiplicative formulation was reported

in the period For each study, we show the country and time period for which decomposition was performed The symbol “EGY” indicates that a study deals with energy consumption while “EMS” denotes an emission study Changes of a wide range of energy and emission aggregates are decomposed We divide the studies into two groups, depending on whether a one-stage decomposition model (referred to as

“DM-S1”) or two-stage decomposition model (referred to as “DM-S2”) is used stage decomposition models treat changes in the Leontief matrix as a single effect, i.e Leontief structure effect, while two-stage decomposition models treat these changes

One-as a combination of various sub-effects from input technology changes, such One-as the share of domestic inputs effect and total input technology effect in Jacobsen (2000) and various KLEM effects in Rose and Chen (1991).7 The symbol “#Factors” in Table 1 denotes the number of factors included in a SDA study The decomposition

methods used are grouped into four categories, i.e ad hoc, D&L, LMDI and Others

The methods in the “D&L” category include those using the full/equivalent D&L or

an approximate technique, while those in “Others” include the mean rate of change index of MRCI and the parametric Divisia methods (Ang, 1995)

2.2.3 Main features and findings from recent studies

We now summarize the key features and developments as revealed by the data

in Table 2.2 While many countries or economies are represented in the studies, around 40% of them (18 out of 43 studies) are on China and/or Japan Since the earliest energy SDA study on China by Lin and Polenske (1995), there is growing interest in analyzing China’s energy use and emission changes Not surprisingly, the majority of the studies deal with decomposition of changes in energy consumption or emissions over time in a country; only three deal with spatial decomposition between countries/regions (de Nooij et al., 2003; Alcántara and Duarte, 2004; and Hasegawa, 2006) There are more studies on emissions than energy consumption, and more than half of the studies deal with carbon emissions Due to the time needed to construct an I-O table, there is an obvious time lag between the publication date of a study and the data used in the study

7

If in a study the decomposition model only utilizes the I-O equality instead of the Leontief inverse matrix, we group this study under DM-S1 because there is no need to consider the inverse operation See Lee and Lin (2001) and Yabe (2004) for example In this sense, all decomposition models in IDA can be treated as equivalent to DM-S1 because they only consider the direct effects instead of the indirect effects captured by the Leontief inverse matrix The same grouping criterion is applied to the Ghosh inverse matrix (Ghosh, 1958), which is analogous to the Leontief inverse matrix from different model assumptions (Oosterhaven, 1996) See, for example, Zhang (2010)