44452 Discussion Paper Opportunities for Mitigating the Environmental Impact of Energy Use in the Middle East and North Africa Region By Yabei Zhang March 2008 This is not a formal publication of the World Bank. It represents preliminary analysis and research. Circulation is intended to encourage discussion and comments; citation and the use of the paper should take account of its provisional character. The findings and conclusions of the paper are entirely those of the author and should not be attributes to the World Bank, its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. TABLE OF CONTENTS 1. Introduction............................................................................................................... 7 1.1 Objectives ........................................................................................................... 7 1.2 Methodologies..................................................................................................... 7 1.3 Structure.............................................................................................................. 8 2. Overview of GHG Emissions and Fossil Fuel Use in MENA................................ 9 2.1 Total GHG Emissions......................................................................................... 9 2.2 GHG (CO2) Emissions from Fossil Fuel Combustion by Fuel Type ............... 11 2.3 Fossil Fuel Use by Country/Sector ................................................................... 13 2.4 GHG (CO2) Emissions by Sector...................................................................... 16 3. Decomposition of Changes in CO2 Emissions from Fossil Fuel for MENA Countries.................................................................................................................. 19 3.1 Theoretical Model............................................................................................. 19 3.2 Data and Application ........................................................................................ 20 3.3 Results............................................................................................................... 22 3.4 Alternative Decomposition Method.................................................................. 24 4. GHG Emissions and Energy Efficiency Analysis by Sector................................ 28 4.1 Estimating Changes in CO2 Emissions by Scenario of Improved Energy Efficiency by Sector.......................................................................................... 28 4.2 Benchmarking Energy Consumption and CO2 Emissions in the Manufacturing Sector ................................................................................................................ 30 4.3 Benchmarking Energy Consumption and CO2 Emissions in the Residential Sector ................................................................................................................ 32 5. PM Concentrations in MENA................................................................................ 35 5.1 Cost of Environmental Degradation Studies in MENA.................................... 35 5.2 The 2006 Little Green Data Book and the GMAPS ......................................... 36 5.3 Magnitude of PM Concentrations in MENA.................................................... 37 5.4 Damage Costs of PM Emissions in MENA...................................................... 38 5.5 Estimating Changes in PM Concentrations by Scenario of Improved Energy Efficiency.......................................................................................................... 41 6. Cross-Country Regression Analysis...................................................................... 44 Reference ......................................................................................................................... 49 1 FIGURES Figure 1 GHG Emissions by Region, 2005..................................................................... 9 Figure 2 Total GHG Emissions by MENA Country, 2005........................................... 10 Figure 3 Share of CO2 Emissions by Fossil Fuel Type by Region, 2005 ..................... 11 Figure 4 CO2 Emissions by Fossil Fuel Type by MENA Country, 2005 ..................... 12 Figure 5 Total Primary Energy Supply from Fossil Fuel by MENA Country, 2005.... 13 Figure 6 Share of Total Primary Energy Supply by Fossil Fuel by Region, 2005........ 14 Figure 7 Total Final Consumption by Sector by MENA Country, 2005 ...................... 15 Figure 8 Share of Total Final Consumption by Sector by Region, 2005...............Error! Bookmark not defined. Figure 9 CO2 Emissions by Sector by MENA Country, 2005...................................... 17 Figure 10 Share of CO2 Emissions by Sector by MENA Country, 2005........................ 17 Figure 11 Share of CO2 Emissions by Sector by Region, 2005...................................... 18 Figure 12 Decomposition of the Changes in CO2 Emissions between 1995 and 2005 (Mt of CO2)............................................................................................................ 23 Figure 13 UAE Gas Flaring Estimated from Defense Meteorological Satellite Program Data................................................................................................................. 24 Figure 14 Decomposition of the Change in CO2 Emissions between 1995 and 2005 (Mt of CO2)-Alternative Method ........................................................................... 27 Figure 15 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by MENA Countries, 2005 ......................................... 30 Figure 16 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by Region/Group, 2005............................................... 32 Figure 17 Energy Consumption and CO2 Emission per unit of GDP per Capita in the Residential Sector by MENA Countries, 2005............................................... 33 Figure 18 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by Region/Group, 2005 .................................................... 34 Figure 19 Particular Matter Concentrations by Region, 2006......................................... 37 Figure 20 Particular Matter Concentrations by MENA Country, 2006 .......................... 38 Figure 21 Particulate Matter Emission Damage (% of GNI in 2004) by Region ........... 40 Figure 22 Particulate Matter Emission Damage (% of GNI in 2004) by MENA Country ......................................................................................................................... 40 Figure 23 The Approach Used to Estimate Change in PM Concentrations by Scenario of Improved Energy Efficiency........................................................................... 42 Figure 24 PM Concentrations by Scenario of Improved Energy Efficiency .................. 43 2 TABLES Table 1 Total GHG Emissions by MENA Country and Region, 2005........................ 10 Table 2 CO2 Emissions from Fossil Fuel Combustion by Fuel, 2005......................... 12 Table 3 Total Primary Energy Supply from Fossil Fuel by Fuel, 2005....................... 14 Table 4 Total Final Consumption by Sector, 2005...................................................... 16 Table 5 CO2 Emissions by Sector, 2005...................................................................... 18 Table 6 Decomposition of the Changes in CO2 Emissions between 1995 and 2005 (Mt of CO2)............................................................................................................ 22 Table 7 Decomposition of the Change in CO2 Emissions between 1995 and 2005 (Mt of CO2)-Alternative Method ........................................................................... 25 Table 8 Comparison of the 1994-2004 CO2 Decomposition and the 1995-2005 CO2 Decomposition................................................................................................ 26 Table 9 Comparison of Results of Using Two Decomposition Methods .................... 26 Table 10 CO2 Emissions Reduction by Scenario of Energy Efficiency Improvements by Sector .............................................................................................................. 29 Table 11 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by MENA Countries, 2005 ......................................... 31 Table 12 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by Region/Group, 2005............................................... 31 Table 13 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by MENA Countries, 2005............................................... 33 Table 14 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by Region/Group, 2005 .................................................... 34 Table 15 Cost of Environmental Degradation Studies: Damage Costs of Air Pollution by MENA Country.......................................................................................... 35 Table 16 PM Concentrations and Damages by MENA Country and by Region........... 39 Table 17 The GMAPS Estimated PM Emission Coefficients for Variables on Energy Consumption................................................................................................... 42 Table 18 PM Concentrations (g/m3) by Scenario of Improved Energy Efficiency .... 43 Table 19 OLS Regression on Natural Log of Energy Consumption per Capita............ 46 Table 20 Correlations of Key Variables ........................................................................ 47 Table 21 OLS Regression on Natural Log of CO2 Emissions per Capita...................... 47 Table 22 Summary Statistics of Key Variables ............................................................. 48 Table 23 OLS Regression including Electricity Price ................................................... 48 3 Executive Summary Energy efficiency reduces the fiscal cost of energy subsidies and energy sector investments, improves economic competitiveness, raises household welfare, lowers local and global pollution and reduces the perception of national energy insecurity. The Middle East and North Africa (MENA) Region has a low level of energy efficiency and energy efficiency has become a high perceived priority for most MENA governments. The World Bank is conducting a study on Energy Efficiency in the Middle East and North Africa (MENA) to create a platform for policy dialogue on energy efficiency based on lessons of international experience and the specific needs of MENA countries. The main motivation for MENA governments to improve energy efficiency is to address pressing domestic concerns, such as urban air pollution, energy security, economic competitiveness, the fiscal cost of energy subsidies and the balance of payments. However, improved energy efficiency is also seen as the most promising route for reducing the region's green house gas (GHG) emissions. Therefore, as part of the study on Energy Efficiency in MENA, this report aims to describe and quantify the links between energy efficiency and GHG emissions as well as urban air pollution in MENA. More specifically, the objectives of the report is · To create scenarios of emissions of GHG and selected urban air pollutants (specifically, particulate matter (PM)) under different assumptions of energy efficiency improvements; · To identify the country/sectors where energy efficiency improvements are likely to have the greatest impact upon GHG and PM emissions; · To benchmark MENA against other regions in terms of energy intensity and GHG emissions. This report uses International Energy Agency (IEA) databases, the World Bank 2006 Little Green Data Book, and the World Bank Development Data Platform (DDP) as the primary data source for the analysis. In particular, a decomposition analysis focusing on sector structure of the economy is developed to estimate CO2 changes between 1995 and 2005 for MENA countries and to build scenarios showing how given percentage energy efficiency improvements by sector would reduce CO2 emissions. The Global Model of Ambient Particulates (GMAPS) developed by the World Bank (Cohen et al, 2004) is adopted to estimate the changes in PM concentrations by scenario of improved energy. A cross-country regression analysis is also used to find key factors that determine energy consumption and CO2 emissions per capita and to identify whether MENA is a relatively intensive energy user or CO2 producer. The primary findings of the report are: · The MENA region contributes to 5% of the total GHG emissions in the world in 2005. Iran has the highest GHG emissions in MENA and contributes to 27% of the total GHG emissions in the region. 4 · The total primary energy supply in the MENA region is about 7% of the world total in 2005. In terms of fossil fuel types, compared to other regions, the MENA region has a bigger share of natural gas and smaller share of coal. Except for Israel and Morocco, few MENA countries use coal and coal products. Iran leads the region in terms of the total primary energy supply from fossil fuel. · The energy sector is the leading sector with the highest CO2 emissions for most MENA countries, contributing to 44% of CO2 emissions for the MENA region. The transportation sector contributes 22% of CO2 emissions in the MENA region, similar to the world average and higher than Asia excluding China and Latin America excluding Mexico. The industry sector contributes another 20% of CO2 emissions in the MENA region. · Among the five MENA countries (Egypt, Iran, Jordan, Morocco, and the United Arab Emirates) that have available data for the decomposition analysis of changes in CO2 emissions between 1995 and 2005, Iran has the highest increase in CO2 emissions. It also has the highest increase in CO2 emission per capita and GDP per capita. The major factor that contributes to the increase in CO2 emissions in Iran is scale effect of increased GDP per capita. Among the five countries, only the United Arab Emirates (UAE) has a negative pollution intensity effect, which is probably due to the fact that the country has significantly reduced gas flaring over the period. · Using 2005 as the baseline, the decomposition analysis shows that energy efficiency improvements in energy and other sectors have the highest potential to reduce CO2 emissions. If energy efficiency is improved by 10% in energy and other sectors, the total CO2 emissions reduction is in the range of 4-8% and Israel has the highest reduction of 8.05%. If energy efficiency is improved by 10% in the transportation sector, MENA countries can achieve a total CO2 emissions reduction of 1-3.5% and Iraq has the highest reduction of 3.5%. If the manufacturing sector has an energy efficiency improvement of 10%, the total CO2 emissions reduction in MENA countries is in the range of 1-3%, with the highest reduction of 2.86% in the United Arab Emirates. · Among its comparators-East Asia (EAP), Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), upper middle income group, and lower middle income group-MENA ranks the first in terms of CO2 emission per dollar of output in the manufacturing sector and ranks second in terms of energy consumption per dollar of output in the manufacturing sector. This shows that MENA has very high pollution intensity and energy intensity in the manufacturing sector and has great potential for energy efficiency improvement in this sector. Within the MENA region, the UAE has the highest energy consumption and CO2 emissions per dollar of manufacturing output, which means that energy efficiency improvements in the manufacturing sector in the UAE will achieve the highest CO2 reduction among MENA countries. 5 · In terms of energy consumption and CO2 emissions per unit of GDP per capita in the residential sector, MENA has relatively low rank among its comparators. Within the MENA region, the UAE is well above other MENA countries in terms of both energy consumption and CO2 emissions per unit of GDP per capita in the residential sector, which means that there is a great potential for energy efficiency improvements in the residential sector in the UAE. · Among the World Bank's analytical regions (EAP, ECA, LAC, MENA, South Asia, and Sub-Saharan Africa), MENA is the second most polluting region in terms of PM concentrations, only behind South Asia. The estimated PM concentration is 90 g/m3, which is 50% higher than the world average. Within the MENA region, Iraq has the highest PM concentration (167 g/m3), followed by Egypt, Kuwait, Oman, Libya, United Arab Emirates, and Saudi Arabia, which are all above the region's average. · In terms of magnitude of current damage costs of PM emissions, using percentage of 2004 GNI as the indicator, MENA ranks the second highest among the World Bank's analytical regions, following the EAP region. Damage cost due to PM emissions is equivalent to 0.9% of the 2004 GNI in the MENA region (about US$5.3 billion), well above the world average of 0.5%. Within the MENA region, Iraq and Kuwait have the highest damage cost of PM emissions--2.7% of GNI, more than five times the world average. Most MENA countries have higher than the world average damage costs of PM emissions, except for Yemen, Tunisia, and Morocco. · The results from the Global Model of Ambient Particulates show that energy efficiency improvements that will lead to decreased per capita consumption in oil, diesel and coal will achieve higher PM concentration reductions. If there are 10% reduction in per capita oil, diesel and coal consumption through energy efficiency improvements, the MENA region as a whole is able to reduce PM concentrations to 83 g/m3, a 7.8% reduction from the baseline (90 g/m3). Within MENA countries, since Iraq has the highest PM concentrations in the baseline, it can achieve the highest reduction in PM concentrations in absolute terms through energy efficiency improvements. · The cross-country regression analysis shows that GDP per capita is the most significant factor in determining energy consumption per capita and CO2 emissions per capita. Other factors such as the sectoral composition of the economy, energy prices, and hydrocarbon production also potentially play important roles. Due to the data limitations, the cross-country regression analysis is not able to provide evidences to confirm the hypothesis that MENA is a relatively intensive energy user and CO2 producer when controlling all the other factors. 6 1. Introduction Energy efficiency reduces the fiscal cost of energy subsidies and energy sector investments, improves economic competitiveness, raises household welfare, lowers local and global pollution and reduces the perception of national energy insecurity. The Middle East and North Africa (MENA) Region has a low level of energy efficiency and energy efficiency has become a high perceived priority for most MENA governments. The World Bank is conducting a study on Energy Efficiency in MENA to: (1) engage a dialogue on energy efficiency with MENA governments by bringing together the lessons of international experience on options, constraint, and opportunities for energy efficiency and their applicability in the context of MENA countries' political economy; and (2) develop the analytical foundation and the paradigm for the Bank's energy efficiency lending and advisory work in MENA. 1.1 Objectives The main motivation for MENA governments to improve energy efficiency is to address pressing domestic concerns, such as urban air pollution, energy security, economic competitiveness, the fiscal cost of energy subsidies and the balance of payments. However, improved energy efficiency is also seen as the most promising route for reducing the region's green house gas (GHG) emissions. Therefore, as part of the study on Energy Efficiency in MENA, this report aims to describe and quantify the links between energy efficiency and GHG emissions as well as urban air pollution in MENA. More specifically, the objectives of the report are to: · Create scenarios of emissions of GHG and selected urban air pollutants (specifically, particulate matter (PM)) under different assumptions of energy efficiency improvements; · Identify the country/sectors where energy efficiency improvements are likely to have the greatest impact upon GHG and PM emissions; · Benchmark MENA against other regions in terms of energy intensity and GHG emissions. 1.2 Methodologies We use International Energy Agency (IEA) databases, the World Bank 2006 Little Green Data Book, and the World Bank Development Data Platform (DDP) as the primary data source for the analysis in this report. To estimate the changes in CO2 emissions by scenario of improved energy efficiency, we first conduct a decomposition analysis to identify how key factors such as population growth, GDP growth, energy intensity, pollution intensity, and sector structure of the 7 economy contribute to changes in CO2 emissions between 1995 and 2005 for MENA countries. Since a recent World Bank report (World Bank, 2007a) uses a different decomposition method which does not consider sector structure, we also compare our results with the ones using the alternative decomposition method. Based on the decomposition analysis, we build scenarios to show how given percentage energy efficiency improvements by sector would reduce CO2 emissions and then identify the country/sectors where energy efficiency improvements are likely to have the greatest impact upon GHG emissions. To estimate the changes in PM concentrations by scenario of improved energy efficiency, we adopt the Global Model of Ambient Particulates (GMAPS) developed by the World Bank (Cohen et al, 2004). Then we are able to identify the country/sectors where energy efficiency improvements are likely to have the greatest impact upon PM concentration reductions. Finally, we adopt a cross-country regression analysis to explain energy consumption and GHG emissions per capita in terms of GDP per capita, the sectoral composition of the economy, hydrocarbon production, and energy prices, and to identify whether MENA is a relatively intensive energy user or CO2 producer. 1.3 Structure This report is structured as follows. Chapter 2 provides an overview of GHG emissions and fossil fuel use in MENA. Chapter 3 details the decomposition analysis of changes in CO2 emissions from fossil fuel for MENA countries and compares the results with the ones using the alternative decomposition method. Based on the decomposition analysis, changes in CO2 emissions by scenario of improved energy efficiency are estimated in Chapter 4. In addition, Chapter 4 presents the benchmarking analysis of MENA's manufacturing and residential sectors against each other and comparators, country by county, in terms of energy consumption and CO2 emissions per unit of GDP per capita. Chapter 5 provides an overview of PM emissions in MENA and presents the estimations of changes in PM concentrations by scenario of improved energy efficiency. The cross- country regression analysis is developed in Chapter 6. 8 2. Overview of GHG Emissions and Fossil Fuel Use in MENA 2.1 Total GHG Emissions According to the IEA Database, as shown in Figure 1 and Table 1, the green house gas (GHG) emissions in the MENA region in 2005 is 2207.65 Mt CO2 equivalent, which is about 5% of the total GHG emissions in the World, and half of the GHG emissions in Asia (excluding China). Within the MENA region, as shown in Figure 2 and Table 1, the Islamic Republic of Iran has the highest GHG emissions and contributes to 27% of the total GHG emissions in the region. Saudi Arabia and Egypt are the next two leading GHG emitters. Together with Iran, the three countries contribute to 56% of the total GHG emissions in the region. Figure 1GHG Emissions by Region, 2005 GHG Emissions by Region, 2005 50000 45000 40000 t en 35000 valiuqE 30000 25000 2OCt 20000 15000 M 10000 5000 0 World Non-OECD Asia (excl. Latin America MENA Total China) (excl. Mexico) Source: the IEA database 9 Figure 2 Total GHG Emissions by MENA Country, 2005 Total GHG Emissions by MENA Country, 2005 700 600 t enl 500 vaiuqE 400 2OC 300 200 Mt 100 0 Arabia EgEmirates ypt Morocco it tar n Islamic Republic oSa f Iran Iraq wa hrain anon udi Algeria Ku Syria Israel LibyaQa OmaYemenTunisiaJordanBa Leb ited Arab Un Source: the IEA database Table 1 Total GHG Emissions by MENA Country and Region, 2005 MENA Country GHG Emissions (Mt CO2 eq.) % of the Region Islamic Republic of Iran 604.33 27% Saudi Arabia 406.06 18% Egypt 231.60 10% United Arab Emirates 152.94 7% Algeria 132.27 6% Iraq 103.86 5% Kuwait 90.17 4% Morocco 75.97 3% Syria 68.53 3% Israel 66.44 3% Libya 60.17 3% Qatar 43.04 2% Oman 36.92 2% Yemen 36.18 2% Tunisia 35.97 2% Jordan 22.97 1% Bahrain 20.65 1% Lebanon 19.60 1% Region GHG Emissions (Mt CO2 eq.) % of the World World 43291.91 100% Non-OECD Total 25659.71 59% Asia (excl. China) 5696.55 13% Latin America (excl. Mexico) 3627.14 8% MENA 2207.65 5% Source: the IEA database 10 2.2 GHG (CO2) Emissions from Fossil Fuel Combustion by Fuel Type Because 74% of GHG are CO2 and the IEA only has detailed data on CO2 emissions, we present CO2 emissions as an indicator of GHG emissions. Fuel combustion is a major source of CO2 emissions. The IEA published CO2 Emissions from Fuel Combustion 1971-2005 (2007) including global annual statistics on CO2 emissions from fossil fuel combustion and sector-level evolution in energy use and associated CO2 emissions. The fuel type classification is coal and coal products, natural gas, oil, and other. The following information is based on this dataset. As shown in Figure 3 and Table 2, compared to other regions in terms of share of CO2 emissions from coal and coal products, natural gas, and oil, coal and coal products contribute the least to CO2 emissions in the MENA region. This is primarily driven by the fact that coal and coal products only represents 2% of primary fuel supplies in this region and most MENA countries are abundant in oil and natural gas resources. As shown in Figure 4 and Table 2, CO2 emissions from oil contribute to more than 50% for most MENA countries and in the case of Yemen, all CO2 emissions are from oil combustion. Figure 3 Share of CO2 Emissions by Fossil Fuel Type by Region, 2005 Share of CO2 Emissions by Fossil Fuel Type by Region, 2005 100% 90% Oil sno 80% sisi 70% mE 60% Natural 2OCfo Gas 50% 40% e 30% Coal and arhS Coal 20% Products 10% 0% MENA World Asia (excl. Latin America Non-OECD China) (excl. Mexico) Total Source: the IEA database 11 Figure 4 CO2 Emissions by Fossil Fuel Type by MENA Country, 2005 CO2 Emissions by Fossil Fuel Type by MENA Country, 2005 450 400 Oil Natural Gas 350 Coal and Coal Products 2OCfo 300 250 200 Mt 150 100 50 0 ic of IranArabi EgEmirates a ypt Iraq ia ait JordaLe n n Alger Kuw Israel Syria LibyaoroccoQatarOmanTuni sia men ain Ye Bahr bano publ Saudi Islamic Re ited Arab M Un Source: the IEA database Table 2 CO2 Emissions from Fossil Fuel Combustion by Fuel, 2005 Coal and Coal Country/Region Products Natural Gas Oil Total Mt of Mt of Mt of Country CO2 % Mt of CO2 % CO2 % CO2 Islamic Republic of Iran 3.9 1% 186.3 46% 216.9 53% 407.1 Saudi Arabia 0.0 0% 109.8 34% 209.9 66% 319.7 Egypt 3.5 2% 63.1 43% 81.0 55% 147.6 United Arab Emirates 0.0 0% 79.1 72% 31.3 28% 110.4 Iraq 0.0 0% 5.1 6% 79.6 94% 84.6 Algeria 1.0 1% 52.6 62% 30.7 36% 84.3 Kuwait 0.0 0% 22.0 30% 52.6 70% 74.6 Israel 29.6 50% 3.0 5% 27.3 46% 59.9 Syria 0.0 0% 12.7 27% 35.1 73% 47.8 Libya 0.0 0% 10.8 24% 34.7 76% 45.4 Morocco 15.5 37% 0.9 2% 25.0 60% 41.3 Qatar 0.0 0% 29.7 82% 6.7 18% 36.4 Oman 0.0 0% 15.0 56% 12.0 44% 27.0 Tunisia 0.0 0% 7.2 37% 12.1 63% 19.3 Yemen 0.0 0% 0.0 0% 18.7 100% 18.7 Bahrain 0.0 0% 14.6 80% 3.7 20% 18.3 Jordan 0.0 0% 3.2 18% 14.7 82% 17.9 Lebanon 0.5 3% 0.0 0% 15.3 97% 15.8 Region MENA 54.1 3% 614.9 39% 907.0 58% 1576.0 World 10980.1 41% 5346.8 20% 10716.7 40% 27043.6 Asia (excl. China) 1242.4 48% 366.8 14% 979.7 38% 2588.9 Latin America (excl. Mexico) 87.7 9% 230.5 25% 619.5 66% 937.7 Non-OECD Total 6644.8 50% 2544.7 19% 4059.0 31% 13248.5 Source: the IEA database 12 2.3 Fossil Fuel Use by Country/Sector As shown in Figure 5 and Table 3, nearly all primary energy supply comes from fossil fuels in the MENA region. Again, the Islamic Republic of Iran leads the region in terms of the total primary energy supply from fossil fuel. Except for Israel and Morocco, few MENA countries use coal and coal products. The total primary energy supply in the MENA region is about 7% of the world total. Compared to other regions, the MENA region has a bigger share of natural gas and smaller share of coal, as illustrated in Figure 6. In terms of total final consumption (TFC) by sector (as shown in Table 3), the transportation sector consumes 31% of the TFC in the MENA region, above the world average 28% and well above Asia excluding China and Non-OECD total, which is 17% and 18%, respectively. The industry sector consumes 25% of the TFC in the MENA region, comparable to the world average but a bit lower than other regions. Figure 5 Total Primary Energy Supply from Fossil Fuel by MENA Country, 2005 ) Total Primary Energy Supply from Fossil Fuel by MENA Country, 2005 oetk(l 180000 Oil and Oil Products 160000 Natural Gas Fuelis 140000 Coal and Coal Products osF 120000 omrf 100000 y 80000 ppluS 60000 gyrenE 40000 20000 yra 0 mirPla Iran ypt IraqwaitIsrael Libya Sy ria tar anroccBahrainTunisiaJordanYe o of ArabiaEgEmir atesAlgeria Ku Qa Om Mo men banon Le Tot public Sa udi Re itedArab Islamic Un Source: the IEA database 13 Figure 6 Share of Total Primary Energy Supply by Fossil Fuel by Region, 2005 Share of Total Primary Energy Supply by Fossil Fuel by Region, 2005 100% 80% 60% eo kt 40% 20% 0% MENA Asia Excluding Latin America Non-OECD Total World China Coal and Coal Products Natural Gas Oil and Oil Products Source: the IEA database Table 3 Total Primary Energy Supply from Fossil Fuel by Fuel, 2005 Coal and Coal Oil and Oil Country/Region Products Natural Gas Products Total MENA Country ktoe % ktoe % ktoe % ktoe Islamic Republic of Iran 1113 1% 82055 51% 77224 48% 160392 Saudi Arabia 0 0% 51077 36% 89196 64% 140273 Egypt 895 2% 27765 47% 30136 51% 58796 United Arab Emirates 0 0% 33839 72% 13081 28% 46920 Algeria 685 2% 22931 66% 11021 32% 34637 Iraq 0 0% 2164 7% 28405 93% 30570 Kuwait 0 0% 9426 33% 18717 67% 28143 Israel 7633 40% 1289 7% 9980 53% 18902 Libya 0 0% 5145 27% 13748 73% 18892 Syria 3 0% 5914 34% 11687 66% 17603 Qatar 0 0% 13347 84% 2478 16% 15825 Oman 0 0% 9318 67% 4646 33% 13964 Morocco 4455 34% 386 3% 8313 63% 13153 Bahrain 0 0% 6245 77% 1883 23% 8128 Tunisia 0 0% 3091 42% 4223 58% 7314 Jordan 0 0% 1384 20% 5570 80% 6954 Yemen 0 0% 0 0% 6650 100% 6650 Lebanon 132 2% 0 0% 5181 98% 5313 Region MENA 14916 2% 275375 44% 342138 54% 632429 Asia Excluding China 329167 36% 173239 19% 401151 44% 903557 Latin America 23090 7% 100764 29% 226959 65% 350813 Non-OECD Total 1761841 39% 1150097 26% 1582319 35% 4494256 World 2892114 31% 2361537 26% 4002077 43% 9255729 Source: the IEA database 14 Figure 7 Total Final Consumption by Sector by MENA Country, 2005 Total Final Consumption by Sector by MENA Country, 2005 140000 Non-Energy Use 120000 Other Sectors 100000 Transport Industry e 80000 kto 60000 40000 20000 0 ria siamenJord an an Islamic Republic oSa f IranArabia Iraq iaIsraelKu waitSy hrain udiited ArabEm Egypt irates Alger Libyaorocco QatarTuni M Ye Om banonBa Le Un Source: the IEA database Figure 8 Share of Total Final Consumption by Sector by Region, 2005 Share of Total Final Consumption by Sector by Region, 2005 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% MENA Asia Latin America Non-OECD World Excluding Total China Industry Transport Other Sectors Non-Energy Use Source: the IEA database 15 Table 4 Total Final Consumption by Sector, 2005 Non-Energy Country/Region Industry Transport Other Sectors Use TFC MENA Country ktoe % ktoe % ktoe % ktoe % Islamic Republic of Iran 29313 23% 34924 27% 55805 44% 7305 6% 127348 Saudi Arabia 12587 15% 27315 33% 11680 14% 31206 38% 82787 Egypt 14276 34% 11304 27% 11659 28% 4351 10% 41590 United Arab Emirates 13617 45% 11157 37% 5519 18% 30 0% 30322 Iraq 5976 25% 10841 46% 5899 25% 790 3% 23507 Algeria 3862 18% 6716 31% 9152 42% 2143 10% 21872 Israel 1383 11% 3976 31% 6012 47% 1335 11% 12707 Kuwait 4165 34% 3377 27% 3888 31% 948 8% 12378 Syria 3138 27% 4029 34% 2573 22% 1939 17% 11679 Libya 1871 17% 4174 37% 2346 21% 2844 25% 11235 Morocco 2874 28% 1057 10% 6157 59% 263 3% 10351 Qatar 2781 36% 2067 27% 855 11% 1983 26% 7686 Tunisia 1548 24% 1751 27% 2848 44% 288 4% 6434 Yemen 595 12% 2231 45% 2021 41% 109 2% 4956 Jordan 1129 23% 1934 40% 1686 35% 110 2% 4859 Oman 1726 38% 1658 37% 1098 24% 39 1% 4522 Lebanon 1135 28% 1537 38% 1327 33% 57 1% 4057 Bahrain 1854 46% 1480 37% 638 16% 43 1% 4016 Region MENA 103829 25% 131526 31% 131164 31% 55786 13% 422305 Asia Excluding China 249078 28% 156420 17% 412247 46% 79765 9% 897509 Latin America 135980 35% 122550 31% 105679 27% 29198 7% 393408 Non-OECD Total 1232284 32% 713348 18% 1626101 42% 316424 8% 3888156 World 2092799 26% 2182915 28% 2933084 37% 702902 9% 7911699 Source: the IEA database 2.4 GHG (CO2) Emissions by Sector In terms of CO2 emissions by sector, IEA's classification of sectors includes the energy sector, manufacturing industries and construction, transport, and other sectors. Manufacturing industries and construction can be treated as the industrial sector. The energy sector is the leading sector with the highest CO2 emissions for most MENA countries, contributing to 44% of CO2 emissions for the MENA region. The transportation sector contributes 22% of CO2 emissions in the MENA region (as compared to 31% of TFC), similar to the world average and higher than Asia excluding China and Latin America excluding Mexico. The industry sector contributes another 20% of CO2 emissions in the MENA region. Among MENA countries, Kuwait and Israel have the highest share of CO2 emissions in the energy sector, Iraq has the highest share in the transportation sector, and the UAE has the highest share in the industry sector. 16 Figure 9 CO2 Emissions by Sector by MENA Country, 2005 CO2 Emissions by Sector by MENA Country, 2005 450 400 350 2OC 300 250 of 200 Mt 150 100 50 0 Iran a ypt Iraq eria waitIsrael SyriaLibya rocco Qa tar an ainJordaLe n public ofSaudi Arabi Islamic Re itedArabEmirates Eg Alg nisia men Ku Om banon Mo Tu Ye Bahr Un Energy Sector Manufacturing Industries and Construction Transport Other Sectors Source: the IEA database Figure 10 Share of CO2 Emissions by Sector by MENA Country, 2005 Share of CO2 Emissions by Sector by MENA Country, 2005 100% 80% 2OCfo 60% 40% Mt 20% 0% Israel Sy ria Libya rocco Qa tar an n ic Republic oSa f Iran Arabia ypt Iraq eria wait men udi Eg irates Alg Ku Om TunisiaYe hrainJordan ano b Em Mo Ba Le itedArab Islam Un Energy Sector Manufacturing Industries and Construction Transport Other Sectors Source: the IEA database 17 Figure 11 Share of CO2 Emissions by Sector by Region, 2005 Share of CO2 Emissions by Sector by Region, 2005 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% MENA Asia (excl. China) Latin America Non-OECD Total World (excl. Mexico) Energy Sector Manufacturing Industries and Construction Transport Other Sectors Source: the IEA database Table 5 CO2 Emissions by Sector, 2005 Manufacturing Industries and Country/Region Energy Sector Construction Transport Other Sectors Total Mt of Mt of Mt of MENA Country Mt of CO2 % CO2 % CO2 % CO2 % Islamic Republic of Iran 108.14 27% 76.44 19% 100.31 25% 122.21 30% 407.10 Saudi Arabia 171.98 54% 71.44 22% 72.52 23% 3.74 1% 319.68 Egypt 61.56 42% 37.41 25% 31.59 21% 17.03 12% 147.59 United Arab Emirates 53.08 48% 31.52 29% 22.06 20% 3.72 3% 110.38 Iraq 29.08 34% 17.54 21% 29.63 35% 8.39 10% 84.64 Algeria 35.59 42% 11.18 13% 17.78 21% 19.75 23% 84.30 Kuwait 50.14 67% 12.42 17% 8.04 11% 4.02 5% 74.62 Israel 40.27 67% 1.81 3% 9.86 16% 7.91 13% 59.85 Syria 22.30 47% 9.78 20% 11.74 25% 3.94 8% 47.76 Libya 23.60 52% 7.31 16% 11.72 26% 2.77 6% 45.40 Morocco 18.03 44% 7.16 17% 1.82 4% 14.34 35% 41.35 Qatar 21.05 58% 9.04 25% 6.14 17% 0.14 0% 36.37 Oman 17.08 63% 5.10 19% 3.61 13% 1.19 4% 26.98 Tunisia 6.78 35% 3.85 20% 4.54 24% 4.11 21% 19.28 Yemen 5.76 31% 1.83 10% 6.18 33% 4.92 26% 18.69 Bahrain 11.39 62% 4.05 22% 2.65 14% 0.23 1% 18.32 Jordan 7.02 39% 2.91 16% 4.78 27% 3.20 18% 17.91 Lebanon 6.76 43% 3.19 20% 3.97 25% 1.89 12% 15.81 Region MENA 689.61 44% 313.98 20% 348.94 22% 223.50 14% 1576.03 Asia (excl. China) 1303.11 50% 634.89 25% 407.53 16% 245.46 9% 2590.99 Latin America (excl. Mexico) 262.94 28% 238.83 25% 324.37 35% 111.54 12% 937.68 Non-OECD Total 6636.35 50% 3294.63 25% 1856.69 14% 1479.87 11% 13267.54 World 12307.24 45% 5184.04 19% 6337.02 23% 3308.06 12% 27136.36 Source: the IEA database 18 3. Decomposition of Changes in CO2 Emissions from Fossil Fuel for MENA Countries In order to understand the factors that contribute to the changes in CO2 emissions for MENA countries, we conduct a decomposition analysis in this chapter. 3.1 Theoretical Model Fossil fuel related CO2 emissions in a given year decomposed into the individual components, can be written as: (1) P = E Pj E j Yj Y B = C I S jGB j j Yj Y B j j j P = the amount of CO2 emissions from the consumption of fossil fuels E = the amount of fossil fuel consumption Y = GDP B = population j = sector C j Pj shows pollution intensity in sector j. E j I E j shows energy intensity in sector j. j Yj S j Yj shows sector structure of the economy. Y G Y shows GDP per capita B The change in a country's emission (P) from period 0 to period T can be decomposed into (1) pollution intensity effect (CC ) , (2) energy intensity effect (CI ) , (3) sector structure effect(CS ) , (4) scale effect(CG ) , and (5) population effect(CB). (2) P = PT - P 0 = CC + CM + CI + CS + CG + C B Using the logarithmic mean Divisia Index (LMDI) formulae developed by Ang (2005), these effects can be calculated as follows. (3) CC = ln Pj - Pj T 0 j Pj - ln Pj T 0lnCTj C0j 19 T 0 (4) CI = ln Pj - Pj j Pj - ln Pj T 0ln ITj I 0j (5) CS = ln Pj - Pj T 0 j Pj - ln Pj T 0 ln STj S0j (6) CG = ln Pj - Pj T 0 j Pj - ln Pj T 0lnGT G0 (7) CB = ln Pj - Pj T 0 j Pj - ln Pj T 0 ln BT B0 The change in emissions will reflect changes in the five effects because of the nature of the identity in equation (2). The pollution intensity effect reflects energy mix in terms of polluting levels. If more clean fuels are replacing dirty fuels or dirty fuels are less dirty due to strengthened environment policies, the pollution intensity will decrease and negatively contribute to the increase of CO2 emissions from period 0 to period T. The energy intensity effect reflects several factors including energy efficiency of the economy. If industries adopt higher energy efficiency standards, which means that less energy is required to produce the same output, the energy intensity will fall. Therefore, with proper policies, this effect could also reduce CO2 emissions. The sector structure effect reflects the composition of the economy. Some sectors are more energy intensive than others, e.g. manufacturing sector is more energy intensive than most service sectors. If the sector structure of GDP changes towards sectors that are less energy intensive, the average use of energy in total GDP would fall, so would the total CO2 emissions. The scale effect and the population effect show how GDP per capita and population change over the period. We generally expect the population to grow over time and people become wealthier, so the two effects generally contribute to the increase of CO2 emissions, unless there are price or regulatory incentives to change behaviors towards less energy intensive consumption patterns. 3.2 Data and Application We used the data from period of 1995 to 2005 to conduct the decomposition analysis. The major challenge of the estimation is that there are no consistent data sources for the classification of economic sectors. Due to the time and data constraints, we split the economy into three sector categories: transport sector, manufacturing sector, and energy and other sectors. The data sources and data adjustment are explained as follows. (1) The emission of CO2 by sector Pj IEA has CO2 emission data measured in million tons (Mt) of CO2 by fossil fuel type and by sector. Fossil fuel types include coal and coal products, oil, and natural gas. The sectors are classified as: 20 Energy Sector --Main Activity Producer Electricity and Heat --Unallocated Autoproducers --Other Energy industries Manufacturing Industries and Construction Transport Other Sectors In order to match the three sectors we proposed, we aggregated the energy sector with other sectors. (2) Fossil fuel consumption by sector Ej IEA has energy balance data which include energy consumption measured in thousand tonnes of oil equivalent (ktoe) by fuel type and by sector. The total final consumption by sector includes industry sector, transportation sector, other sectors, and non-energy use. We aggregated other sectors and non-energy use and treated the industry sector as the manufacturing sector in our classification. (3) The level of GDP Y The GDP data in 1995 and 2005 are taken from the IEA database and measured in constant 2000 US$ and valued according to the purchasing power parity (PPP). We then extrapolated 2000 US$ into 2005 US$ by using the GDP deflator published by the US government. Thus, all GDP data are measured in billions of 2005 US$ PPP. We initially attempted to use the GDP data in 2005 taken from the 2005 International Comparison Program (ICP) Preliminary Results report published in December 2007 which used an updated method to estimate PPPs. However, we found the new GDP 2005 data represent a new benchmark and are not comparable with previous results, which gave us unexpected growth rates for certain countries. Therefore, we use the GDP data from the IEA database that uses the previous version of PPP conversion factors. (4) The level of GDP by sector Yj The UN national accounts database in the Development Data Platform (DDP) has categories of manufacturing (% of GDP) and transport and communications (current US$). Although the database has a separate row for transport value-added alone, unfortunately the data are missing for nearly all MENA countries. Thus, we used the transport and communications category as a proxy for the transport sector. We did notice the limitation of using such proxy because this category includes postal services and telecommunications which is a low-carbon intensity sector. Therefore, we could not estimate the effect of changes in the carbon-intensity of the transport sector over time, because the rapid growth of the low-carbon telecoms sector would have biased downwards the carbon intensity growth of transport and communications sectors. However, this proxy should not affect the validity of the decomposition analysis regarding the sector share effect because the growth decomposition is an identity. In addition, our assumption that communication is zero carbon is necessary to allow us to map IEA energy data onto DDP value-added data. Furthermore, we noticed that the GDP from the transport sector did not include private passenger transport. Thus, to ensure the 21 consistency of the energy intensity effect, we further assumed that the private and commercial transport sub-sectors have grown at the same rates. Finally, the level of GDP by sector was adjusted to constant 2005 US$ PPP in billions. (5) The population of the country B Data on population is taken from the IEA database and measured in millions. 3.3 Results Since complete GDP data by sector for the year 1995 and the year 2005 is only available for Egypt, Iran, Jordan, Morocco, and the United Arab Emirates, we present the decomposition results for these five countries only. As shown in Table 6 and Figure 12, among these five MENA countries, Iran has the highest increase in CO2 emissions between 1995 and 2005. In addition, Iran has the highest increase in CO2 emission per capita and GDP per capita. This is why although Iran has significant energy efficiency improvements by sector, due to the big scale effect it still has the highest CO2 emissions increase. Egypt and Morocco have positive energy intensity effect, which means that energy intensity is getting worse in the two countries and contributes to the increase of CO2 emissions. Egypt has a negative sector structure, which means that the economy is moving towards less carbon intensive sectors. Among the five countries, only the UAE has a negative pollution intensity effect. It is probably due to the fact that the country has significantly reduced gas flaring over the period, as shown in Figure 13. Table 6 Decomposition of the Changes in CO2 Emissions between 1995 and 2005 (Mt of CO2) CO2 GDP per Pollution Energy Sector Change Scale Population of CO2 Emission per capita Country Intensity Intensity Structure Emissions Capita (2005 US$ (t/capita) PPP) CC CI CS CG CB P 2005 2005- 2005 2005- 1995 1995 Iran 6.4 -9.5 10.4 103.3 47.1 157.8 6.0 1.7 7949 2184 Egypt 8.9 10.7 -5.6 28.2 21.4 63.6 2.0 0.6 4327 959 United Arab Emirates -8.4 -15.9 11.8 0.9 54.1 42.6 24.4 -3.8 25472 272 Morocco 1.6 1.9 0.1 7.9 4.5 15.9 1.4 0.4 4544 978 Jordan 0.4 -1.7 0.3 2.9 3.9 5.8 3.3 0.4 5521 971 22 Figure 12 Decomposition of the Changes in CO2 Emissions between 1995 and 2005 (Mt of CO2) Decomposition of Changes in CO2 Emissions, 1995-2005 Iran Egypt United Arab Emirates Morocco Jordan -20 -10 0 10 20 30 40 50 60 70 80 90 100 110 Mt of CO2 Pollution Intensity Energy Intensity Sector Structure Scale Population 23 Figure 13 UAE Gas Flaring Estimated from Defense Meteorological Satellite Program Data UAE Gas Flaring Estimated From Defense Meteorological Satellite Program Data 2.5 s tere 2.0 m ci cub lion 1.5 bil 1.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year Source: Elvidge et al (2007) 3.4 Alternative Decomposition Method For comparison, we applied the decomposition method used in the recent World Bank report Growth and CO2 Emissions: How do Different Countries Fare (World Bank, 2007a) for MENA countries to analyze the CO2 emissions change between 1995 and 2005. The major difference between this decomposition method and the method we used above is that this method does not consider sector structures but includes the share of fossil fuels in total energy (the substitution effect). For ease of comparison, we use the same notations here as in the report. Results are presented in Table 7 and Figure 14. Table 8 shows comparison of the 1994-2004 CO2 decomposition from Bacon et al (2007) and the 1995-2005 CO2 decomposition using the same method. We see that since the two periods have significant overlaps, the results are relatively comparable, especially for GDP per capita and population effects. Table 9 shows comparison of the decomposition results using two different methods for the five MENA countries. We denote the one includes sector structure as decomposition method 1 and the one includes the share of fossil fuels in total energy as decomposition method 2. As we expected, both methods have obtained the same scale effects and population effects. In terms of energy intensity, method 1 defines it as the fossil fuel consumption per unit of GDP while method 2 defines it as the total energy consumption 24 per unit of GDP. Taking Iran as an example, energy intensity is negative using method 1 which means that less fossil fuel is required per unit of GDP so it negatively contributes to the CO2 increase. However, sector structure effect is positive which means that the economy switches to more CO2 intensive sectors and such switch contributes to the CO2 increase. Using method 2, Iran has a positive energy intensity effect and a negative substitution effect which mean that total energy use per unit of GDP increases and such effect contributes to the CO2 increase; but among the total energy use, the share of fossil fuel use decreases and such effect negatively contribute to the CO2 increase. Therefore, using different decomposition methods give us different perspectives on what factors contribute to the changes of CO2 emissions. Table 7 Decomposition of the Change in CO2 Emissions between 1995 and 2005 (Mt of CO2)- Alternative Method Coefficient Substitution Energy GDP Change of CO2 Emission Intensity per Population CO2 per Capita (t of GDP per capita Country Effect Effect Effect Capita Emissions CO2 per capita) (2005 US$ PPP) Ceff Seff Ieff Geff Peff E 2005 2005- 2005 2005- 1995 1995 Iran 3.8 -5.3 8.7 103.4 47.1 157.8 6.0 1.7 7949 2184 Saudi 13.8 2.9 Arabia -24.7 -0.2 72.7 12.4 57.1 117.2 15673 736 Egypt 9.8 -2.4 6.5 28.3 21.4 63.6 2.0 0.6 4327 959 United Arab 24.4 -3.8 Emirates 28.4 -7.4 -34.6 0.9 55.1 42.6 25472 272 Kuwait 9.3 -1.6 1.8 4.2 19.4 33.2 29.4 6.4 26211 1877 Algeria -5.2 -1.2 3.4 17.4 10.7 25.0 2.6 0.5 7046 1529 Qatar 1.1 -0.9 -5.4 11.2 11.4 17.3 44.7 8.5 43235 14756 Morocco 0.6 0.4 2.5 7.9 4.5 15.9 1.4 0.4 4544 978 Israel 9.2 -5.6 -6.7 5.1 11.7 13.5 8.7 0.3 25816 2380 Iraq -1.7 0.1 -17.9 10.0 22.4 12.9 2.9 -0.4 1016 121 Oman 2.5 0.0 3.9 2.9 3.3 12.6 10.5 3.9 15503 2093 Libya -0.4 -2.2 1.2 3.9 7.8 10.3 7.8 0.5 7884 736 Yemen 0.9 0.0 2.6 1.5 4.3 9.4 0.9 0.3 928 98 Syria 8.7 -4.2 -8.5 2.2 11.0 9.1 2.5 -0.1 3799 194 Bahrain 2.0 -0.6 -2.0 4.0 3.4 6.7 25.2 5.3 21337 5000 Jordan -0.3 -0.3 -0.4 2.9 3.9 5.8 3.3 0.4 5521 971 Tunisia 0.5 -0.8 -2.9 6.3 1.9 5.0 1.9 0.3 8350 2635 Lebanon 3.4 -1.4 -3.1 2.6 1.7 3.2 4.4 0.5 5566 950 25 Table 8 Comparison of the 1994-2004 CO2 Decomposition and the 1995-2005 CO2 Decomposition Coefficient Substitution Energy GDP per Change of Effect Effect Intensity Population CO2 Effect Capita Emissions Country Period Ceff Seff Ieff Geff Peff E Algeria 1994-2004 -6.7 -0.3 -30.4 18.2 12.4 -6.9 Algeria 1995-2005 -5.2 -1.2 3.4 17.4 10.7 25.0 Bahrain 1994-2004 -0.3 0.0 -1.7 4.4 4.5 7.0 Bahrain 1995-2005 2.0 -0.6 -2.0 4.0 3.4 6.7 Egypt 1994-2004 -11.4 2.6 5.8 29.8 22.9 49.7 Egypt 1995-2005 9.8 -2.4 6.5 28.3 21.4 63.6 Iran 1994-2004 -27.3 1.3 35.2 107.5 36.6 153.1 Iran 1995-2005 3.8 -5.3 8.7 103.4 47.1 157.8 Israel 1994-2004 -1.6 0.2 0.7 5.3 13.5 18.1 Israel 1995-2005 9.2 -5.6 -6.7 5.1 11.7 13.5 Morocco 1994-2004 -0.9 -0.9 -4.6 3.9 4.4 1.9 Morocco 1995-2005 0.6 0.4 2.5 7.9 4.5 15.9 Oman 1994-2004 -0.3 0.0 2.0 3.1 3.4 8.3 Oman 1995-2005 2.5 0.0 3.9 2.9 3.3 12.6 Saudi Arabia 1994-2004 -6.6 0.0 70.9 -18.7 81.2 126.8 Saudi Arabia 1995-2005 -24.7 -0.2 72.7 12.4 57.1 117.2 Syria 1994-2004 -1.2 0.7 -2.3 2.6 12.0 11.9 Syria 1995-2005 8.7 -4.2 -8.5 2.2 11.0 9.1 Tunisia 1994-2004 -1.3 -0.1 -2.5 6.4 2.2 4.9 Tunisia 1995-2005 0.5 -0.8 -2.9 6.3 1.9 5.0 United Arab Emirates 1994-2004 -4.6 0.0 -21.6 2.0 71.7 47.4 United Arab Emirates 1995-2005 28.4 -7.4 -34.6 0.9 55.1 42.6 Table 9 Comparison of Results of Using Two Decomposition Methods Pollution Sector Intensity/ Structure/ Change of Decomposition Coefficient Energy Substitution CO2 Method Effect Intensity Effect Scale Population Emissions Iran 1 6.4 -9.5 10.4 103.3 47.1 157.8 Iran 2 3.8 8.7 -5.3 103.4 47.1 157.8 Egypt 1 8.9 10.7 -5.6 28.2 21.4 63.6 Egypt 2 9.8 6.5 -2.4 28.3 21.4 63.6 United Arab Emirates 1 -8.4 -15.9 11.8 0.9 54.1 42.6 United Arab Emirates 2 28.4 -34.6 -7.4 0.9 55.1 42.6 Morocco 1 1.6 1.9 0.1 7.9 4.5 15.9 Morocco 2 0.6 2.5 0.4 7.9 4.5 15.9 Jordan 1 0.4 -1.7 0.3 2.9 3.9 5.8 Jordan 2 -0.3 -0.4 -0.3 2.9 3.9 5.8 26 Figure 14 Decomposition of the Change in CO2 Emissions between 1995 and 2005 (Mt of CO2)- Alternative Method Decomposition of Changes in CO2 Emissions for MENA Countries, 1995-2005 Lebanon Tunisia Jordan Bahrain Syria Yemen Libya Oman Iraq Israel Morocco Qatar Algeria Kuwait United Arab Emirates Egypt Saudi Arabia Iran -60 -40 -20 0 20 40 60 80 100 120 Mt of CO2 Emissions Intensity of Fossil Fuels Share of Fossil Fuels Energy Intensity GDP per Capita Population 27 4. GHG Emissions and Energy Efficiency Analysis by Sector 4.1 Estimating Changes in CO2 Emissions by Scenario of Improved Energy Efficiency by Sector As analyzed in Chapter 3, CO2 emissions can be decomposed as the following. (8) P = E Pj E j YjY B = C I S jGB j j Yj Y B j j j j=1, transportation sector j=2, manufacturing sector j=3, energy and other sectors. In this chapter, we are interested in how energy intensity (energy efficiency) Ij change would affect the total CO2 emissions P, assuming other factors do not change. Equation (8) can be rewritten as (9) P = A I , j j j where Aj = C jS jGB , a combined factor. We use the year 2005 as the baseline and consider three scenarios: energy efficiency improvement (i.e. energy intensity reduction) by 10% in transportation sector, manufacturing sector, and energy and other sectors. The results are presented in Table 10. In Scenario 1 if energy efficiency is improved by 10% in the transportation sector, MENA countries can achieve a total CO2 emissions reduction of 1-3.5% and Iraq has the highest reduction of 3.5%. In Scenario 2, if the manufacturing sector has an energy efficiency improvement of 10%, the total CO2 emissions reduction in MENA countries is in the range of 1-3%, with the highest reduction of 2.86% in the United Arab Emirates. In Scenario 3, if energy efficiency is improved by 10% in energy and other sectors, the total CO2 emissions reduction is in the range of 4-8% and Israel has the highest reduction of 8.05%. 28 Table 10 CO2 Emissions Reduction by Scenario of Energy Efficiency Improvements by Sector Scenario 1: 10% EE Scenario 2: 10% EE Scenario 3: 10% EE Baseline: 2005 improvement in improvement in improvement in other Transportation sector Manufacturing sector sectors Country A1 A2 A3 % % % P I1 I2 I3 I1_new P_new change I2_new P_new change I3_new P_new change of P of P of P Algeria 0.02 0.04 1.00 84.29 746.35 279.11 55.10 671.72 82.51 2.11% 251.20 83.17 1.33% 49.59 78.76 6.56% Bahrain 0.00 0.01 1.40 18.31 1553.27 428.43 8.28 1397.94 18.05 1.45% 385.58 17.91 2.21% 7.45 17.15 6.34% Egypt 0.06 0.21 2.01 147.60 498.57 180.16 39.05 448.71 144.44 2.14% 162.14 143.86 2.53% 35.15 139.74 5.33% Islamic Republic 0.19 0.26 2.44 407.08 521.83 290.13 94.39 469.65 397.05 2.46% 261.12 399.44 1.88% 84.96 384.05 5.66% of Iran Iraq 0.02 0.00 0.80 84.64 1752.77 3709.65 46.68 1577.49 81.68 3.50% 3338.69 82.89 2.07% 42.01 80.89 4.43% Israel 0.02 0.16 1.43 59.86 537.82 11.67 33.66 484.04 58.87 1.65% 10.51 59.68 0.30% 30.30 55.04 8.05% Jordan 0.01 0.01 0.13 17.90 577.23 205.26 77.11 519.51 17.42 2.67% 184.73 17.61 1.63% 69.40 16.88 5.70% Kuwait 0.02 0.01 2.21 74.63 531.45 1507.67 24.49 478.30 73.83 1.08% 1356.90 73.39 1.66% 22.04 69.21 7.26% Lebanon 0.01 0.02 0.38 15.80 621.02 178.80 22.77 558.92 15.40 2.51% 160.92 15.48 2.02% 20.49 14.94 5.47% Libya 0.01 0.06 0.32 45.40 1463.20 127.03 81.56 1316.88 44.23 2.58% 114.33 44.67 1.61% 73.40 42.76 5.81% Morocco 0.02 0.06 0.50 41.34 110.79 124.74 64.33 99.71 41.16 0.44% 112.26 40.62 1.73% 57.90 38.10 7.83% Oman 0.00 0.01 1.81 26.98 930.64 380.55 10.10 837.58 26.62 1.34% 342.49 26.47 1.89% 9.09 25.15 6.77% Qatar 0.01 0.06 0.38 36.37 971.87 151.33 55.07 874.68 35.76 1.69% 136.20 35.47 2.49% 49.56 34.25 5.83% Saudi Arabia 0.05 0.34 2.26 319.68 1554.84 208.33 77.74 1399.36 312.43 2.27% 187.50 312.54 2.23% 69.97 302.11 5.50% Syria 0.03 0.03 0.46 47.76 436.03 379.47 57.06 392.43 46.59 2.46% 341.52 46.78 2.05% 51.36 45.14 5.49% Tunisia 0.01 0.04 0.31 19.29 461.98 96.62 35.13 415.78 18.84 2.35% 86.96 18.90 2.00% 31.62 18.20 5.65% United Arab 0.02 0.04 4.02 110.38 1136.47 707.10 14.14 1022.82 108.17 2.00% 636.39 107.23 2.86% 12.73 104.70 5.15% Emirates Yemen 0.01 0.01 0.23 18.69 442.93 257.63 46.57 398.64 18.07 3.31% 231.87 18.51 0.98% 41.91 17.62 5.71% 29 4.2 Benchmarking Energy Consumption and CO2 Emissions in the Manufacturing Sector To compare energy consumption and CO2 emissions per dollar of output in the manufacturing sector in MENA countries, we used energy consumption and CO2 emissions data from IEA and the manufacturing output data from DDP. Since manufacturing output data are only available for eight MENA countries, we report the results for these eight countries in Figure 15 and Table 11. The UAE has the highest energy consumption and CO2 emissions per dollar of manufacturing output among the eight countries. It means that the UAE has the worst energy intensity and pollution intensity in the manufacturing sector. This result is also consistent with the result in Chapter 3.1 that energy efficiency improvements in the manufacturing sector in the UAE will achieve the highest CO2 reduction among MENA countries. In terms of region/group comparison, we select EAP, ECA, LAC, upper middle income group, and lower middle income group as comparators. As shown in Table 12 and Figure 16, the MENA region ranks the first among the comparators in terms of CO2 emissions per dollar of output in the manufacturing sector and ranks second in terms of energy consumption per dollar of output in the manufacturing sector. This result shows that the MENA region has very high pollution intensity and energy intensity in the manufacturing sector and has great potential for energy efficiency improvement in this sector. Figure 15 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by MENA Countries, 2005 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by MENACountries, 2005 0.0018 0.0016 0.0014 PPP PPP 0.0012 $SU $SU 0.0010 2005 0.0008 2005 e/ot 2/OCfot0.0006 0.0004 0.0002 0.0000 United Algeria Islamic Jordan Egypt Lebanon Morocco Tunisia Arab Republic Emirates of Iran Energy Consumption/Dollar of Manufacturing Output (toe/2005 US$ PPP) CO2 Emission/Dollar of Manufacturing Output (t of CO2/2005 US$ PPP) 30 Table 11 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by MENA Countries, 2005 Energy CO2 Consumption/D Emission/Doll CO2 Emission ollar of ar of Manufacturing in Total Energy Manufacturing Manufacturing Output (2005 Manufacturing Consumption in Output Output (t of US$ PPP in Sector (Mt of Manufacturing (toe/2005 US$ CO2/2005 Country billions) CO2) Sector (ktoe) PPP) US$ PPP) United Arab Emirates 18.472468 31.52 13616.588 0.00074 0.00171 Algeria 11.106486 11.18 3861.516 0.00035 0.00101 Islamic Republic of Iran 86.400948 76.43 29313.325 0.00034 0.00088 Jordan 4.5125295 2.91 1128.67 0.00025 0.00064 Egypt 59.556768 37.41 14275.592 0.00024 0.00063 Lebanon 5.2133339 3.19 1135.166 0.00022 0.00061 Morocco 17.737661 7.16 2873.856 0.00016 0.00040 Tunisia 11.362116 3.86 1547.73 0.00014 0.00034 Table 12 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by Region/Group, 2005 Energy CO2 CO2 Emission in Total Energy Consumption/ Emission/Doll Dollar of ar of Region/Group a Manufacturing Manufacture Consumption in Manufacturing Manufacturing Output (2005 Sector (Mt of Manufacturing Output Output (t of US$ PPP in CO2) Sector (ktoe) (toe/2005 CO2/2005 billions) US$ PPP)b US$ PPP) MENA 315.87 313.98 103828.66 0.00033 0.00099 ECA 489.49 433.68 208514.68 0.00043 0.00089 Lower Middle Income 3666.91 2073.37 671729.83 0.00018 0.00057 Upper Middle Income 1057.97 584.84 311001.08 0.00029 0.00055 EAP 3619.26 1860.59 579775.12 0.00016 0.00051 LAC 654.93 237.07 134855.84 0.00021 0.00036 Notes: a. Regions and income groups are based on the World Bank's classification and only cover countries that have available data for this analysis. b. GDP data used in this table is from IEA and measured in 2000 US$ PPP and then converted to 2005 US$ PPP using the GDP deflator. 31 Figure 16 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by Region/Group, 2005 Energy Consumption and CO2 Emissions per Dollar of Output in the Manufacturing Sector by Region/Group, 2005 0.0012 0.0010 PPP PPP $SU $SU 0.0008 5 0.0006 200 2005/eot 2/OCfot0.0004 0.0002 0.0000 MENA ECA Lower Middle Upper Middle EAP LAC Income Income Energy Consumption/Dollar of Manufacturing Output (toe/2005 US$ PPP) CO2 Emission/Dollar of Manufacturing Output (t of CO2/2005 US$ PPP) 4.3 Benchmarking Energy Consumption and CO2 Emissions in the Residential Sector To compare energy consumption and CO2 emissions per unit of GDP per capita in residential sector for MENA countries, we used 2005 energy consumption and CO2 emissions data in the residential sector from IEA. GDP data for 2005 are from the latest ICP report and the population data are from the IEA database. The results are presented in Figure 17 and Table 13, which show significant variations among MENA countries in terms of the two indicators. The UAE is well above other MENA countries in term of both energy consumption and CO2 emissions per unit of GDP per capita in the residential sector. Libya has the lowest energy intensity and pollution intensity in terms of the two indicators. In terms of region/group comparison, as shown in Table 14 and Figure 18, the MENA region has the relative low energy consumption and CO2 emissions per unit of GDP per capita in the residential sector. The LAC region ranks the lowest among the comparators. 32 Figure 17 Energy Consumption and CO2 Emission per unit of GDP per Capita in the Residential Sector by MENA Countries, 2005 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the r Residential Sector by MENA Countries, 2005 pe PDG 4000 9000 8000 )PPP 3500 atipacr oft 3000 7000 pe 6000 PPP) unir $SU 2500 5000 PDG 5 2000 pe 4000 oft $SU onitp 200 1500 3000 e/ot( 1000 0052 2000 unir pe um ati 500 1000 nsoC apc 0 0 onissi 2/OCfot( ygr sia Iraq ain ria ypt n Sy Iran bya mE Eg wait Li neE Arabia Oman Le T i Bahr Jorda Ku lic of Saud 2OC United ArabEmiratesAlgeriMo arocco bano IsraelYemen Qat uni n ar Islamic Repub CO2 Emission per unit of GDP per capita (t of CO2/2005 US$ PPP) Energy Consumption per unit of GDP per capita (toe/2005 US$ PPP) Table 13 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by MENA Countries, 2005 Total CO2 Energy Total Energy Emissions in GDP per Consumption per CO2 Emission per Consumption Residential Capita unit of GDP per unit of GDP per in Residential Sector (Mt of (2005 US$ capita (toe/2005 capita (t of CO2/2005 Country Sector (ktoe) CO2) PPP) US$ PPP) US$ PPP) United Arab Emirates 40399.17 91.11 10762 3753.905 8465.973 Algeria 9151.90 19.75 5985 1529.139 3299.916 Morocco 3023.90 8.39 3104 974.067 2702.611 Lebanon 8154.42 11.79 4775 1707.699 2469.063 Israel 2409.72 4.05 3550 678.816 1140.882 Yemen 918.71 1.99 2202 417.195 903.684 Qatar 1877.64 2.12 3944 476.035 537.481 Tunisia 1063.44 2.13 4296 247.533 495.791 Saudi Arabia 1872.90 1.94 6461 289.894 300.281 Iraq 1750.68 2.77 10727 163.203 258.227 Bahrain 945.73 1.89 10726 88.169 176.203 Oman 8138.84 3.74 21237 383.238 176.108 Syria 3045.59 3.72 28930 105.274 128.586 Egypt 2876.51 2.61 22645 127.029 115.260 Jordan 3184.42 4.02 43504 73.198 92.405 Kuwait 515.49 0.28 19883 25.926 14.082 Islamic Republic of Iran 453.22 0.23 27671 16.379 8.312 Libya 320.10 0.14 69012 4.638 2.029 33 Table 14 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by Region/Group, 2005 Total CO2 GDP per Energy Total Energy CO2 Emission Emissions in Capita Consumption per unit of GDP Region/Group a Consumption in per unit of GDP Residential Residential (2005 per capita per capita (t of Sector (ktoe) Sector (Mt of US$ CO2/2005 US$ CO2) PPP ) b (toe/2005 US$ PPP) PPP) Lower Middle Income 543764.31 489.42 6120.79 88838.85 79960.22 EAP 447632.24 287.50 6052.67 73956.16 47499.70 ECA 181619.92 218.68 8326.02 21813.54 26264.66 MENA 90102.36 162.67 6893.67 13070.30 23597.00 Upper Middle Income 194814.68 196.29 9816.75 19845.12 19995.41 LAC 66714.40 57.82 7944.24 8397.83 7278.23 Notes: a. Regions and income groups are based on the World Bank's classification and only cover countries that have available data for this analysis. b. GDP and population data used in this table is from IEA and measured in 2000$ PPP and then converted to 2005 US$ PPP using the GDP deflator. Figure 18 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by Region/Group, 2005 Energy Consumption and CO2 Emissions per unit of GDP per Capita in the Residential Sector by Region/Group, 2005 r 100000 90000 pe PDG 90000 80000 80000 70000 ofti aitpacrep PPP) 70000 60000 )PPP unrep S$U 60000 PDGfo $SU 5002/eot( 50000 50000 oni 40000 pt 40000 itnurep um 30000 n nsoC atipac 30000 ios 20000 20000 is 2005/2OCfot( gyrenE 10000 10000 mE 0 0 2OC Lower Middle EAP ECA MENA Upper Middle LAC Income Income CO2 Emission per unit of GDP per capita (t of CO2/2005 US$ PPP) Energy Consumption per unit of GDP per capita (toe/2005 US$ PPP) 34 5. PM Concentrations in MENA Combustion of fossil fuels for power generation, transportation, industry, and other human activities produce a complex mixture of particular matter (PM) and gaseous pollutants, with serous health consequences for the exposed population. Fine particles less than 10 microns (m) (PM10) and especially less than 2.5 m (PM2.5) in diameter are considered the most harmful to health because they are small enough to be inhaled and transported deep into the lungs. The concentration of PM10 (in micrograms per cubic meter, or g/m3) is used in this study as the indicator of air pollution. 5.1 Cost of Environmental Degradation Studies in MENA During the period 2001-2005, Cost of Environmental Degradation Studies were prepared for eight MENA countries by the World Bank to provide a first order of magnitude of the cost of environmental degradation as a percentage of GDP as regards to the health impacts of air pollution and waterborne illnesses, the economic cost of water resources and soil/land degradation, impacts related to waste management, and the cost of coastal zone degradation (Sarraf et al, 2004). The studies rely on existing data and analysis of environmental issues, and apply commonly used methodologies of valuation and quantitative impact assessments to country specific issues in order to arrive at estimates of the cost of degradation. Urban air pollution dose response coefficients from international studies reflecting acute mortality of particulates (PM) have been applied in all the eight country reports. The summary of damage costs of air pollution by the eight MENA countries is presented in Table 15. Among the eight studied countries, Egypt has the highest damage cost of air pollution as the percentage of GDP and air pollution is the leading category that has the most significant impacts on health and quality of life in Egypt. Table 15 Cost of Environmental Degradation Studies: Damage Costs of Air Pollution by MENA Country Number of estimated % of overall annual Total estimated premature Disability cost of death due to Adjusted Life environmental Reference US$/Year % of urban air Years degradation Country Year (millions) GDP pollution (DALYs) to the country Algeria 1999 446 0.9% N/A 157,000 26% Egypt 2001 1,890 2.1% 20,000 450,000 44% Iran 2002 1,810 1.6% 13,200 191,000 21% Jordan 2000 64 0.8% 600 15,300 28% Lebanon 2000 170 1.0% 350 9,000 30% Morocco 2000 340 1.0% 2,300 73,000 28% Syria 2001 218 1.2% 3,400 95,000 35% Tunisia 1999 121 0.6% 590 15,000 28% Note: damage cost due to air pollution includes both urban air pollution and indoor air pollution (IAP). However, IAP is minimal in Syria and Jordan since there is nearly universal access to commercial fuels. Source: compiled by the author from Cost of Environmental Degradation Studies in MENA. 35 5.2 The 2006 Little Green Data Book and the GMAPS Because Cost of Environmental Degradation Studies are not available for all MENA countries and their reference years are not all the same, we use the 2006 Little Green Data Book by the World Bank as the primary source for the following analysis of urban air pollution. The PM data in the Little Green Data Book were estimated using the Global Model of Ambient Particulates (GMAPS) (Cohen et al, 2004) developed at the World Bank. The GMAPS model is used to generate estimates of concentrations of PM10 in all world cities with a population of >100,000. The GMAPS model econometrically estimates a fixed-effect model of the concentrations of urban ambient PM using the latest available data from a sample of cities from the World Health Organization (WHO) and other sources and then uses regression estimates to predict PM concentrations worldwide. The primary determinants of the GMAPS include the scale and composition of economic activity, the energy mix, the strength of local regulation of pollution, and geographic and atmospheric conditions that affect the transport of pollutants. The GMAPS is a log-log linear model. The dependent variable is log [PM10] and the independent variables are the log form of the following variables plus time-tend variables: -Energy consumption, -Meteorological and geographic factors, -City and national population and national population density, -Local population density, -Local intensity of economic activity, and -National income per capita. The GMAPS model is designed to obtain the best city-level prediction of concentrations of PM for a wide range of cities on the basis of the limited amount of data. Subregion PM concentrations are computed from estimates of concentrations of PM10 in the cities, using the populations of each city as weights. In terms of the cost of PM emissions, the 2006 Little Green Data Book uses a percentage of 2004 gross national income (GNI) to provide an order of magnitude. It is also useful to compare the cost of air pollution to GNI to assess the relative magnitude over time. The process of estimating the cost of air pollution involves placing a monetary value on the consequences of air pollution. A three-step process is usually used. Step (1) is to translate PM levels to health effects -- mortality and morbidity by age/sex for each country. The concentration response relationships are the same for all countries. These define the percentage increase in health effects for each unit increase in PM levels. The concentration-response functions vary by the age of a person. As a result, the health effects for a country depend on population characteristics (age structure) and baseline health characteristics. This step is outlined in Cohen et al (2004). Step (2) is to translate 36 health effects to economic costs. The lost lives and morbidity effects are translated to Disability Adjusted Life Years (DALYs). Since DALYS are age/sex dependent, demographic characteristics are also important in this step. Finally, Step (3) is to translate DALYS to monetary costs. This requires valuation of a statistical life. This is assumed to vary in proportion to GNI. 5.3 Magnitude of PM Concentrations in MENA According to the 2006 Little Green Data Book, as shown in Figure 19, the MENA region is the second most polluting region in terms of PM concentrations, only behind South Asia. The estimated PM concentration weighted by the share of the urban population [weighted average concentration] is 90 g/m3, which is 50% higher than the world average. Within the MENA region, as shown in Figure 20, Iraq has the highest PM concentration (167 g/m3), followed by Egypt, Kuwait, Oman, Libya, United Arab Emirates, and Saudi Arabia, which are all above the region's average. Figure 19 Particular Matter Concentrations by Region, 2006 Particular Matter Concentrations by Region, 2006 120 100 80 3 m 60 g/ 40 20 0 South MENA EAP SSA World LAC ECA Asia Source: World Bank, 2006 Little Green Data Book 37 Figure 20 Particular Matter Concentrations by MENA Country, 2006 Particulate Matter Concentrations by MENA Country, 2006 180 160 140 120 3 m 100 g/ 80 60 40 20 0 Iraq EgyptKuwait an menJordanDjibolic uti Qatar IsraelTunisLe iabanMorocco on Om Arabia ria Sy of IranAlgeriaBa hrain Ye itedArabEmSaud Libya iratesi Un Islamic Repub Source: World Bank, 2006 Little Green Data Book 5.4 Damage Costs of PM Emissions in MENA In terms of magnitude of current damage costs of PM emissions, as we mentioned earlier, the 2006 Little Green Data Book provides particulate emission damage as a percentage of 2004 GNI as the indicator. As shown in Table 16 and Figure 21, the MENA region ranks the second highest, following the EAP region. Damage cost due to PM emissions is equivalent to 0.9% of the 2004 GNI in the MENA region (about $5.3 billion), well above the world average of 0.5%. Within the MENA region, as shown in Table 16 and Figure 22, Iraq and Kuwait have the highest damage cost of PM emission--2.7% of GNI, more than five times the world average. Most MENA countries have higher than the world average damage costs of PM emissions, except for Yemen, Tunisia, and Morocco. 38 Table 16 PM Concentrations and Damages by MENA Country and by Region Particulate Particulate matter (urban- emission pop.-weighted damage (% of Country/Region avg., g/m3) GNI) MENA Country Iraq 167 2.7 Egypt 136 1.7 Kuwait 129 2.7 Oman 124 1.1 Libya 121 .. United Arab Emirates 109 2.5 Saudi Arabia 91 1 Syria 89 1 Yemen 82 0.5 Jordan 69 0.9 Djibouti 68 .. Islamic Republic of Iran 68 0.9 Algeria 65 0.7 Bahrain 65 0.6 Qatar 57 1.3 Israel 53 1.2 Tunisia 46 0.4 Lebanon 43 0.9 Morocco 27 0.3 Region EAP 80 1.2 MENA 90 0.9 South Asia 99 0.8 ECA 35 0.7 LAC 43 0.6 SSA 73 0.5 World 60 0.5 Source: World Bank, 2006 Little Green Data Book 39 Figure 21 Particulate Matter Emission Damage (% of GNI in 2004) by Region Particulate Matter Emission Damage (% of 2004 GNI) by Region 1.4 1.2 1 INGfo 0.8 0.6 % 0.4 0.2 0 EAP MENA South Asia ECA LAC SSA World Source: World Bank, 2006 Little Green Data Book Figure 22 Particulate Matter Emission Damage (% of GNI in 2004) by MENA Country Particulate Matter Emission Damage (% of 2004 GNI) by MENA Country 3 2.5 INGfo 2 1.5 % 1 0.5 0 Iraq wait ates ypt ar an n n en Ku ir Eg Qat IsraelOm Syria TunisiaMo rocco udiArabia Jordaic of IraLebanonAlgeriaBa Sa ubl hrainYem itedArabEm Un Islamic Rep Source: 2006 Little Green Data Book 40 5.5 Estimating Changes in PM Concentrations by Scenario of Improved Energy Efficiency We used results from the GMAPS to estimate change in PM concentrations by scenario of improved energy efficiency. Among the GMAPS independent variables, energy consumption variables are of our primary interest. The model includes six separate per capita energy consumption categories-coal, oil, natural gas, nuclear, hydro-electric, and combustible renewables and wastes. In addition, it includes per capita consumption of petrol and diesel used in the transportation sector. All the energy consumption data are from IEA. The method for estimating change in PM concentrations by scenario of improved energy efficiency can be illustrated in Figure 23. We assume energy consumption is inversely proportional to energy efficiency, given all other factors do not change. The GMAPS estimated coefficients for energy consumption variables are summarized in Table 1. Note that the model is designed to provide stable predictions under a wide domain space and may not be able to provide all cause-effect relationships of individual factors. Although the model specification assumes independent relationships among fuel types, some fuel types are substitutable (e.g. gasoline and diesel in the transportation sector). Thus, the coefficients which can be interpreted as elasticities capture both a scale effect and a substitution effect. This is why we see a negative sign for the gasoline variable coefficient. Therefore, the following results should be interpreted based on this limitation of the model. As shown in Table 17, per capita oil consumption has the highest coefficient, followed by per capita diesel and coal consumption. This implies that energy efficiency improvements that will lead to decreased per capita consumption in oil, diesel and coal will achieve higher PM concentration reductions. For example, if a country is able to reduce per capita oil consumption by 10% through an energy efficiency program, such program will result in 5.9% of reduction in PM concentrations. Similarly, if an energy efficiency program is targeted to reduce per capita diesel consumption in the transportation sector by 10%, PM concentrations are expected to be reduced by 1.1%. We assume three scenarios of reduced energy consumption from improved energy efficiency: per capita coal, oil, and diesel consumption are reduced by 10%, 20%, and 30%. Table 18 and Figure 24 present the results for each MENA country under the three scenarios. Since Iraq has the highest PM concentrations in the baseline, it achieves the highest reduction in PM concentrations in absolute terms, from 167 g/m3 to 129 g/m3 in Scenario 3. Also under Scenario3, the MENA region as a whole is able to reduce PM concentrations to 70 g/m3, a 22% reduction from the baseline. 41 Table 17 The GMAPS Estimated PM Emission Coefficients for Variables on Energy Consumption Variable on Estimated PM Energy Emission Consumption Description Coefficients lnymotorgaspc ln country per capita motor gasoline consumption -0.1408 lnygasdiespc ln country per capita diesel consumption 0.1123 lnycoalpc ln country per capita coal consumption 0.0557 lnyhydpc ln country per capita hydro consumption -0.0850 lnynatgaspc ln country per capita natural gas consumption -0.0489 lnynucpc ln country per capita nuclear consumption -0.0036 ln country per capita combustible and renewable lnycmrepc consumption -0.0654 lnyoilallpc ln country per capita oil consumption 0.5867 Source: the GMAPS model result. Figure 23 The Approach Used to Estimate Change in PM Concentrations by Scenario of Improved Energy Efficiency Change in PM Concentrations Scenario 1 Emission Change in GMAPS Coefficients x Energy Scenario 2 Consumption Scenario 3 ... 42 Table 18 PM Concentrations (g/m3) by Scenario of Improved Energy Efficiency Scenario 1 Scenario 2 Scenario 3 10% reduction in 20% reduction in 30% reduction in per capita coal, oil, per capita coal, oil, per capita coal, and diesel and diesel oil, and diesel MENA Country Baseline consumption consumption consumption Iraq 167 154 142 129 Egypt 136 126 115 105 Kuwait 129 119 110 100 Oman 124 115 105 96 Libya 121 112 103 94 United Arab Emirates 109 101 93 84 Saudi Arabia 91 84 77 70 Syria 89 82 76 69 Yemen 82 76 70 63 Jordan 69 64 59 53 Djibouti 68 63 58 53 Islamic Republic of Iran 68 63 58 53 Algeria 65 60 55 50 Bahrain 65 60 55 50 Qatar 57 53 48 44 Israel 53 49 45 41 Tunisia 46 43 39 36 Lebanon 43 40 37 33 Morocco 27 25 23 21 MENA Region 90 83 76 70 Figure 24 PM Concentrations by Scenario of Improved Energy Efficiency PM Concentration by Scenario of Improved Energy Efficiency 180 Baseline )3 160 Scenario 1 m g/( 140 Scenario 2 120 Scenario 3 onsitar 100 nte 80 ncoC 60 40 MP 20 0 Iraq ypt an Djibouic of IranAlgeriaBahrain ti ar n Eg wait men rdan nisia ano roccoRegion Ku Om ArabiaSyriaYe Jo Qat IsraelTu ArabEmSaud Libya iratesi publ Leb MoMENA United Islamic Re 43 6. Cross-Country Regression Analysis In order to determine whether MENA is a relatively intensive user of energy and CO2 producer, we conducted a cross-country regression analysis to explain energy consumption and CO2 emissions per capita. The factors we considered included GDP per capita, the sectoral composition of the economy, hydrocarbon production, and energy prices. Again, the main challenge is to find comparable data for this analysis and we detail the data source for each variable as follows. All data are in year 2005. · The two dependent variables energy consumption per capita measured in toe/capita and CO2 emissions per capita measured in ton of CO2 per capita are calculated using the data from IEA. · GDP per capita measured in 2005 US$ PPP is from the latest ICP report (World Bank, 2007b). · In term of sectoral composition of the economy, we use two sets of sectoral compositions. The first set is what we used for the decomposition analysis, by manufacturing sector, transportation sector, and energy and other sectors. The second set is the standard classification of GDP by sector: industry, agriculture, and service. We are more interested in the first set of sectoral composition; however, due to missing data, analysis using such sectoral composition yields a very small sample size and significantly limits the interpretation of the results. Nevertheless, we report both results. Data for the first set of sectoral composition are from the DDP database and for the second set of sectoral composition from the 2005 CIA World Factbook. · In terms of hydrocarbon production, we use two indicators. One is fuel exports as a percentage of GDP in order to capture the importance of fuel exports in the economy. The hypothesis that we are trying to test is that a high ratio of fuel exports to GDP encourages investment in energy intensive activities. The advantage of using fuel exports data is that the cost of energy for a net exporting country would be the export parity price. So it is exports that would explain substitution towards energy-intensive activities on the basis of economic cost. The other indicator we use is energy production measured in ktoe. The advantage of using energy production data is that it would capture non-tradable energy production such as natural gas. Both data sets are from the DDP database. · In terms of energy prices, there is no consistent energy price index available for all countries. We used the diesel price measured in egg index from the GTZ report International Fuel Prices 2005 (GTZ, 2005) as an indicator for energy prices. However, we notice that although the price of diesel price is well correlated with gasoline prices, it is not well correlated with electricity prices. Thus diesel price is unable to capture variations of electricity price across countries. IEA has electricity prices for both industry and household uses 44 measured in current US$ (IEA, 2007) and we converted the data into 2005 US$ PPP using the PPP ratios in the latest CIP report. However, electricity price data are not available for MENA countries. Therefore, we run separate regressions for countries with available data to see how important the electricity price is in determining energy consumption and CO2 emissions. Since electricity prices for industry and household uses are highly correlated, we only included electricity price for industry. We used a log-log linear model to estimate elastisities of different factors on energy consumption per capita and CO2 emissions per capita. The Ordinary Least Square (OLS) regression results on energy consumption per capita are presented in Table 19. We used different specifications to show how inclusion of different factors would change the results. Since Columns (1) and (2) use the first set of sectoral composition and yield very small sample sizes, we focus on our discussion on Column (3)-(5) which use the second set of sectoral composition (industrial, agriculture, and service). Column (3) include share of fuel exports, Column (4) include energy production, and Column (5) include both of indicators. GDP per capita has the positive effect on energy consumption per capita and it is statistically significant at 1% level for all three specifications. It also has the highest coefficient, which means that GDP per capita is the most significant factor on energy consumption per capita. Share of fuel exports has a negative effect on energy consumption per capita although it is not significant. So our results do not support the hypothesis that a high ratio of fuel exports to GDP encourages investment in energy intensive activities. Energy production shows a positive effect on energy consumption per capita, but again it is not statistically significant. So we do not find evidences that being a hydrocarbon producer will increase energy consumption per capita. Diesel price has negative effect on energy consumption per capita as expected and it is statistically significant at 5% level in specification (4). In addition, increase in the share of industry sector production has a positive sign and being a MENA country has an ambiguous sign, but since both of them are not statistically significant, we cannot give definitive conclusions on their effects. The reason that few variables show significant effects may be that in addition to the small sample size, some key independent variables have relatively high correlations as shown in Table 20, which can cause high multicollinearity and lead to high standard errors. The regression results on CO2 emissions per capita are presented in Table 21. Again, we focus on the results in Columns (3)-(5). GDP per capita has a positive effect on CO2 emissions per capita and it is statistically significant at 1% level. Increase in the share of industry sector production has a positive effect and it is statistically significant in Columns (3) and (4) but not in Column (5). All other factors are not statistically significant. Since the specification in Column (4) yields the highest sample size and gives more statistical power, it is our preferred specification. Summary statistics of variables in Column (4) are presented in Table 22. 45 For reference, the regression results including electricity price are presented in Table 23. Only around 20 countries have all the available data. Electricity prices shows a negative effect on both energy consumption and CO2 emissions per capita and it is statistically significant in Columns (1)-(3). The results show that omitting electricity price in our earlier regression analysis may result in biased estimates. Therefore, our cross-country regression analysis above needs to be interpreted with caution. Table 19 OLS Regression on Natural Log of Energy Consumption per Capita (1) (2) (3) (4) (5) Ln GDP per capita 0.610** 0.492** 0.636** 0.560** 0.635** [3.97] [5.06] [5.20] [6.64] [5.14] Ln Share of manufacturing output 0.372 0.207 [0.95] [1.10] Ln Share of transportation sector 0.209 -0.066 [0.39] [0.24] Ln Fuel exports as % of GDP -0.073 -0.017 -0.025 [0.58] [0.27] [0.37] MENA region -0.241 -0.082 0.064 -0.13 0.09 [0.18] [0.20] [0.20] [0.80] [0.28] Ln Diesel price (egg index) -0.411 -0.22 -0.04 -0.147* -0.045 [1.07] [1.17] [0.29] [2.11] [0.32] Ln Energy production 0.027 0.049 0.034 [0.66] [1.65] [0.73] Ln Share of industry 0.506 0.102 0.363 [1.66] [0.46] [1.00] Ln Share of agriculture -0.054 -0.045 -0.054 [0.53] [0.61] [0.53] Constant -6.367* -4.902** -7.239** -5.545** -7.058** [2.97] [4.45] [5.40] [5.78] [5.14] Observations 17 26 36 65 36 R-squared 0.807 0.753 0.777 0.719 0.781 Absolute value of t statistics in brackets + significant at 10%; * significant at 5%; ** significant at 1% 46 Table 20 Correlations of Key Variables GDP per capita Share of Share Diesel Energy (2005 industry Share of of fuel Energy price Consumption US$ sector agriculture exports Production MENA (egg per Capita PPP) (%) (%) (%) (ktoe) region index) Energy Consumption per Capita 1 GDP per capita (2005 US$ PPP) 0.90 1 Share of industry sector (%) 0.50 0.43 1 Share of agriculture (%) -0.54 -0.64 -0.51 1 Share of fuel exports (%) 0.69 0.55 0.58 -0.35 1 Energy Production (ktoe) 0.24 0.11 0.45 -0.17 0.12 1 MENA region 0.49 0.39 0.39 -0.30 0.67 0.04 1 Diesel price (egg index) -0.36 -0.27 -0.39 0.23 -0.55 0.06 -0.49 1 Note: Number of Observations=36 Table 21 OLS Regression on Natural Log of CO2 Emissions per Capita (1) (2) (3) (4) (5) Ln GDP per capita 1.071** 1.028** 1.222** 0.971** 1.214** [6.44] [6.00] [8.17] [7.53] [8.08] Ln Share of manufacturing output 0.388 0.908* [0.88] [2.69] Ln Share of transportation sector -0.627 0.062 [1.02] [0.13] Ln Fuel exports as % of GDP 0.057 0.04 0.03 [0.40] [0.50] [0.38] MENA region -1.058 0.129 -0.161 0.041 -0.128 [0.73] [0.17] [0.42] [0.16] [0.33] Ln Diesel price (egg index) -0.494 -0.363 -0.081 -0.168 -0.095 [1.14] [1.06] [0.49] [1.57] [0.57] Ln Energy production -0.018 0.017 0.047 [0.25] [0.39] [0.88] Ln Share of industry 0.796* 1.099** 0.614 [2.15] [3.26] [1.44] Ln Share of agriculture 0.075 -0.022 0.081 [0.63] [0.20] [0.68] Constant -7.678** -10.304** -12.645** -11.403** -12.416** [3.13] [5.40] [7.66] [7.63] [7.40] Observations 18 27 39 69 39 R-squared 0.91 0.813 0.866 0.800 0.869 Absolute value of t statistics in brackets + significant at 10%; * significant at 5%; ** significant at 1% 47 Table 22 Summary Statistics of Key Variables Variable Obs Mean Std. Dev. Min Max Energy Consumption per Capita 65 0.93 0.90 0.14 5.50 CO2 Emissions per Capita 69 3.30 4.28 0.04 25.19 GDP per capita (2005 US$ PPP) 69 7,238 8,535 591 47,465 Share of industry sector (%) 69 32.09 10.79 11.00 67.20 Share of agriculture (%) 69 17.17 12.26 0.10 55.00 Energy Production (ktoe) 69 98,745 243,486 230 1,536,782 Share of fuel exports (%) 42 8.28 13.82 0.00 59.85 MENA region 69 0.16 0.37 0 1.00 Diesel price (egg index) 69 5.11 3.07 0.10 15.50 Table 23 OLS Regression including Electricity Price (1) (2) (3) (4) Energy Consumption per CO2 Emissions per Capita Capita Ln GDP per capita 1.201** 0.751** 1.287** 0.917** [7.75] [4.52] [4.28] [4.33] Ln Share of industry 1.511* -0.008 0.998 0.456 [3.35] [0.02] [1.23] [0.82] Ln Share of agriculture 0.191 0.144 0.167 0.076 [1.65] [0.94] [0.92] [0.47] Ln Fuel exports as % of GDP 0.232* 0.26 [2.85] [1.70] Ln Diesel price (egg index) 0.079 -0.243 0.063 -0.13 [0.39] [1.17] [0.20] [0.62] Ln Electricity price for industry -0.530** -0.338+ -0.635* -0.348 [3.48] [1.90] [2.23] [1.58] Ln Energy production 0.038 0.132+ [0.64] [2.09] Constant -18.112** -7.748** -16.114** -10.895** [7.47] [4.25] [3.63] [4.67] Observations 15 22 18 26 R-squared 0.936 0.835 0.809 0.859 Absolute value of t statistics in brackets + significant at 10%; * significant at 5%; ** significant at 1% 48 Reference Ang, B.W. (2005). "The LMDI Approach to Decomposition Analysis: A Practical Guide." Energy Policy 33:867-71. Bruvoll and Medin (2003) "Factors Behind the Environmental Kuznet Curve" Environmental and Resource Economics 24: 27-48. Cohen et al (2004) "Urban Air Pollution", Chapter 17 in Comparative Quantification of Health Risks-Global and Regional Burden of Disease Attribution to Selected MajorRisk Factors. The World Health Organization. http://www.who.int/publications/cra/chapters/volume2/1353-1434.pdf Elvidge et al (2007) A Twelve Year Record of National and Global Gas Flaring Volumes Estimated Using Satellite Data, Final Report to the World Bank - May 30, 2007. http://www.ngdc.noaa.gov/dmsp/interest/DMSP_flares_20070530_b.pdf GTZ (2005) International Fuel Prices 2005. http://www.gtz.de/de/dokumente/en_International_Fuel_Prices_2005.pdf IEA(2007) Energy Prices & Taxes 2007. Sarraf, M., B. Larsen, and M. Owaygen (2004) Cost of Environmental Degradation -- The Case of Lebanon and Tunisia. Environmental Economics Series Paper No.97. World Bank (2001) Cost Assessment of Environmental Degradation Algeria. http://lnweb18.worldbank.org/mna/mena.nsf/METAP+Documents/61BE59E441984230 85256C860074E8C6?OpenDocument World Bank (2002) Arab Republic of Egypt Cost Assessment of Environmental Degradation. Report No. 25175 ­ EGT. World Bank (2003) Kingdom of Morocco Cost Assessment of Environmental Degradation. Report No 25992-MOR. World Bank (2003) Syrian Arab Republic Cost Assessment of Environmental Degradation. World Bank (2004) Cost Assessment of Environmental Degradation Jordan. World Bank (2005) Islamic Republic of Iran Cost Assessment of Environmental Degradation. Report No. 32043-IR. 49 World Bank (2007a) "Growth and CO2 Emissions: How to Different Countries Fare?" World Bank (2007b) 2005 International Comparison Program (ICP) Preliminary Results. 50 Appendix Data for Table 21 Column (4) GDP per Share CO2 capita of Diesel Emissions (2005 industry Share of Energy price per US$ sector agriculture Production MENA (egg Country Capita PPP) (%) (%) (ktoe) region index) Algeria 2.57 5985 57.4 10.3 165728 1 1.5 Argentina 3.64 11076 35.9 10.6 85446 0 6.1 Armenia 1.37 3903 36.1 22.9 746 0 4.7 Azerbaijan 3.73 4648 45.7 14.1 20053 0 1.8 Bahrain 25.19 27257 41 0.7 15842 1 1.7 Bangladesh 0.26 1268 27.1 21.2 18390 0 4.9 Benin 0.3 1390 14.3 36.3 1623 0 6 Bolivia 1.29 3623 28 13 11818 0 6.7 Botswana 2.44 12057 44 4 1008 0 5.1 Brazil 1.77 8606 38.6 10.1 176312 0 9.8 Brunei Darussalam 13.62 47465 45 5 20768 0 1.4 Bulgaria 5.96 9353 30.1 11.5 10270 0 8.9 Cameroon 0.18 1995 20.1 43.7 12476 0 5.5 Chile 3.6 12277 38.2 6.3 8390 0 5.3 China 3.88 4091 52.9 13.8 1536782 0 7.2 Taiwan 11.41 26068 30.9 1.7 12760 0 4.6 Colombia 1.31 6314 32.1 13.4 76233 0 2.8 Congo, Dem. Rep 0.04 3621 11 55 17002 0 6.2 Ecuador 1.77 6541 30.5 8.7 29295 0 3.4 Egypt 1.99 5051 33 17.2 64662 1 2 El Salvador 0.86 5212 31.1 9.2 2441 0 5.8 Ethiopia 0.07 591 12.4 47 19370 0 8.4 FYR of Macedonia 4.07 7393 26 11.2 1536 0 7.7 Georgia 0.84 3505 22.6 20.5 1287 0 5.2 Germany 9.86 30496 31 1 136009 0 12.9 Ghana 0.32 1225 24.2 34.3 6230 0 2.7 Guatemala 0.83 4902 19.5 22.7 5331 0 4.9 Haiti 0.2 1242 20 30 1654 0 4.3 Honduras 0.89 3048 32.1 12.7 1747 0 9.4 India 1.05 2126 28.4 23.6 466873 0 15.5 Indonesia 1.55 3234 45 14.6 258009 0 3 Iraq 2.94 3202 58.6 13.6 103419 1 0.1 Iran, Islamic Rep. 5.96 10692 40.9 11.2 277992 1 0.3 Jordan 3.27 4297 26 2.4 292 1 2.1 Kazakhstan 10.22 8699 37.8 7.4 118597 0 3.8 51 GDP per Share CO2 capita of Diesel Emissions (2005 industry Share of Energy price per US$ sector agriculture Production MENA (egg Country Capita PPP) (%) (%) (ktoe) region index) Kenya 0.29 1359 18.5 19.3 13675 0 10.9 Kyrgyzstan 1.06 1728 22.8 38.5 1482 0 3.9 Lebanon 4.42 10220 21 12 230 1 3.9 Mexico 3.7 11317 27.2 4 253859 0 5 Mozambique 0.08 743 32.1 21.1 8236 0 5.3 Namibia 1.36 4547 30.8 11.3 321 0 5 Nepal 0.11 1081 20 40 8066 0 8.2 Nicaragua 0.8 2725 24.7 20.7 1930 0 6.4 Nigeria 0.42 1892 30.5 36.3 229440 0 3 Pakistan 0.76 2396 24.1 22.6 58993 0 6.8 Paraguay 0.58 3905 24.9 25.3 6628 0 8.5 Peru 1.02 6474 27 8 9474 0 10.9 Philippines 0.92 2932 31.9 14.8 23391 0 3.8 Romania 4.2 9374 33.7 13.1 28110 0 4.8 Russia 10.79 11861 33.9 4.9 1158465 0 3.8 Saudi Arabia 13.83 21236 67.2 4.2 556212 1 1.7 Senegal 0.4 1676 21.4 15.9 1106 0 5.6 South Africa 7.05 8477 31.2 3.6 155998 0 4.4 Sri Lanka 0.63 3481 26.2 19.1 5161 0 5.9 Switzerland 6 35520 34 1.5 11822 0 4.9 Syria 2.51 4062 31 25 29516 1 1.6 Tajikistan 0.87 1413 24.3 23.7 1517 0 4.5 Thailand 3.34 6869 44.3 9 50103 0 4.1 Togo 0.16 888 20.4 39.5 1910 0 5.5 Tunisia 1.92 6461 31.8 13.8 6805 1 4.9 Turkey 3.04 7786 29.8 11.7 24111 0 11.2 Turkmenistan 8.59 4211 42.7 28.5 58151 0 0.1 Ukraine 6.31 5583 45.1 18 76287 0 4.9 Uruguay 1.52 9277 27.4 7.9 850 0 7.1 Uzbekistan 4.21 1970 26.3 38 56867 0 2.3 Venezuela 5.35 9888 46.5 0.1 196064 0 0.3 Vietnam 0.97 2142 40.1 21.8 65271 0 3.2 Yemen 0.89 2278 44.7 15.5 20609 1 1.3 Zambia 0.18 1175 28.9 14.9 6360 0 7.5 52