Toward integrating climate change externalities in Bank sector work

E N V I R O N ME N T D E P A R T M E N T _ *^t P A P E R S -PAPER NO. 060 TOWARD ENVIRONMENTALLY AND SOCIALLY SUSTAINABLE DEVELOPMENT Climate Change Series Toward Integrating Climate Change Externalities in Bank Sector Work Jens Rosebrock October 1997 -\ - - -Environmentally and Socially Sustainable Development The World Bank ESSD  i u Global Environment Division Toward Integrating Climate Change Externalities in Bank Sector Work Jens Rosebrock October 1997 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the World Bank's Environment Department, Global Environment Division, Room S-2145.  Acknowledgement This Paper was written while the author was a consultant to the Climate Change Unit of the World Bank's Global Environment Division. It has since been revised and updated to reflect later developments. The author is grateful to Charles Feinstein for valuable advice and guidance during the course of writing this paper. Particular thanks go to Samuel Fankhauser, Annika Haksar, Todd Johnson, Louise Scura and Noreen Beg for commenting on earlier drafts and for suitable revisions and extensions Climate Change Series  Contents Introduction 1 1. Policy Framework 3 FCCC Requirements for GHG Mitigation Studies 3 Previous GHG Mitigation Work 3 Global Overlays 4 2. Typology of GHG Abatement Cost Studies 7 Top-down vs. Bottom-up 7 A Typology of GHG Abatement Cost Studies 8 3. Review of Previous GHG Mitigation Studies in Developing Countries 11 The UNEP Greenhouse Gas Abatement Costing Study Project 11 The World Bank/UNDP Study on China 17 4. Toward Best Practice 23 Four Conclusions from Previous Studies 23 Building Blocks for Global Overlays 24 Next Steps 25 Acronyms 27 References 29  Introduction Global climate change is becoming increasingly Integration Scheme has been completed, which important as a policy issue for World Bank may develop into a full Global Overlay. A client countries. While the Framework Conven- Phase 1 Global Overlay for the Russia Oil Sector tion on Climate Change (FCCC) is unlikely to has also been completed. The U.K., in co- impose legally binding limits on developing operation with the Bank, is funding a study on countries in the near future, it does require Greenhouse Gas Emissions from the India countries to "formulate, implement, publish Power sector in two provinces, which is nearing and regularly update national and, where completion. A study on Greenhouse Gas appropriate, regional programs containing emissions in the urban transport sectors of measures to mitigate climate change." These Mexico City and Santiago, examining how communication requirements will in turn lead control cost curves for local pollution reduction to a large number of climate-change related are altered by the presence of a globally moti- studies, including greenhouse gas (GHG) vated agent (e.g., the Global Carbon Initiative mitigation studies, which analyze and evaluate [GCII), is also nearing completion. abatement costs of measures and policies that reduce GHG emissions. This paper examines the methodological groundwork for Global Overlays in the climate The World Bank hopes to play a significant role change area. Chapter 1 describes the communi- in this effort through an analytical tool, the cation requirements of the FCCC for Bank client Global Overlay, which extends economic and countries and the implications these may have sector work (ESW) in relevant sectors such as for Bank Country Assistance Strategies (CAS). energy, transport, and forestry to include It also describes the Global Overlays concept climate change externalities, i.e. GHG emis- and outlines recent changes in the Bank's sions. This extension program is being spear- Operational Policy regarding the specification headed by the Climate Change Unit of the of global externalities. Chapter 2 lays out some World Bank's Global Environment Division of the conceptual and methodological problems (ENVGC). Global Overlays are undertaken as a behind GHG abatement costing. It describes modest extension of regular Bank ESW and do objectives, scope and limitations of the two not require any changes in the operational main conceptual approaches and the method- responsibilities for the sector work. ological components of previous mitigation studies. Chapter 3 reviews the two most Three Global Overlays have already been comprehensive mitigation studies completed so completed by the regions in collaboration with far and analyzes their strengths and limita- Bank clients (Argentina Carbon Sequestration, tions. Chapter 4 summarizes the findings from Ukraine Energy Options Global Environmental previous chapters and suggests building blocks Analysis, and Mexico Greenhouse Gas Assess- and a four-step analytical structure for Global ment). An initial study on the Mekong Power Overlays. Climate Change Series  Toward Integrating Climate Change Externalities in Bank Sector Work 2 Environment Department Papers  1 Policy Framework FCCC Requirements for GHG cations. These national communications will Mitigation Studies likely be informed by preceding GHG country studies. It should also be noted that, under Art. Since the mid 1980s, climate change has 4.3 of the FCCC, developed countries committed evolved from an issue of scientific inquiry to a themselves to meeting the full costs of such global policy concern. The Framework Conven- studies. tion on Climate Change (FCCC) that was signed at the Rio Summit in June 1992 entered into Previous GHG Mitigation Work force in March 1994, and in the meantime has been ratified by more than 150 countries. While GHG mitigation studies are a subset of abate- the FCCC does not impose any legally binding ment cost studies for greenhouse gas reduction targets on emissions of greenhouse gases which identify and evaluate different techno- (GHG) on any of its contracting parties and any logical measures and policy options to reduce emission constraints on most developing GHG emnissions in the context of the FCCC (see countries are unlikely in the near future, the Chapter 2) 3. Such studies are carried out under Convention does have immediate effects different auspices. One of the first and largest through its communication requirements. GHG mitigation studies was coordinated by the These will contribute to a large number of United Nations Environment Programme climate-change related studies, including GHG (UNEP) Collaborating Centre on Energy and the inventories, vulnerability assessments and Environment in Denmark (see Chapter 3.1) and GHG mitigation studies. completed in May 1994. The Global Environ- ment Facility (GEF) financed an extensive study In its Art. 4 (b), the FCCC requires countries to on "Issues and Options in Greenhouse Gas "formulate, implement, publish and regularly Emissions Control in China" that was jointly update national and, where appropriate, implemented by World Bank staff and two regional programs containing measures to high-ranking Chinese ministries, under the mitigate climate change". According to Art. administrative guidance of the United Nations 12.1., all ratifiers shall communicate these DevelopmentProgramme (UNDP) (see Chapter programs, together with an inventory of green- 3.2). This project was completed in December house gas sources and sinks, to the interim 1994. An analysis of energy mitigation options secretariat. Developed countries and others in Ukraine prepared for the Kiev resident listed in Annex I, which includes most econo- mission of the World Bank by RCG/Hagler mies in transition', also have to present a Bailly was completed in December 1994 (RCG/ detailed description of the policies adopted to Hagler Bailly 1994). The GEF financed a study stabilize emission by 2000 at 1990 levels. Their on Greenhouse Gas Emissions and Mitigation first communication was due in September Strategies in Mexico in collaboration with the 19952. Universidad Nacional Autonoma de Mexico. A regional study called "Climate Change in Countries not listed in Annex I, i.e. most Asia", written by the Climate Change Institute developing nations, have not been given a and financed by the Asian Development Bank, deadline for submission of national communi- Clinate Change Series 3  Toward Integrating Climate Change Externalities in Bank Sector Work also contains some estimates of abatement * it uses an existing sector development costs. An earlier intercountry comparison was strategy for sectors with significant undertaken by the Lawrence Berkeley Laborato- greenhouse gas (GHG) emissions such as ries in 1991. energy, transport, forestry and agriculture and calculates the emissions associated More extensive attempts are currently under- with this strategy, and way, financed by the GEF or bilateral donors. A US$9.5m GEF-financed project * it outlines cost-effective sectoral mitigation designed to evaluate the costs and effec- options available to the country if it seeks tiveness of GHG mitigation measures in 12 to limit its GHG emissions in that sector. Asian countries started in mid-1994. In addition, bilateral donors are also financ- Overlay analyses for climate change quantify ing mitigation studies. The most extensive only climate change externalities, i.e. the program is the U.S. Country Studies emission or sequestration of greenhouse gases Initiative, which is guided by an interde- associated with a particular sector development partmental working group and comprises strategy. Sectoral analyses for other focal areas mitigation components for over 50 coun- (such as biodiversity losses or impacts on tries. Technical advice for its mitigation international waters) are being implemented work is coordinated by Lawrence Berkeley separately. Laboratory. The U.K., in co-operation with the Bank, is funding a study on It should be noted that an extension of Bank Greenhouse Gas Emissions from the India sector work is already called for by recent Power sector in two states, which is changes in its operational policy. Specification nearing completion. The Norwegin, of global environmental externalities is called Danish, German and Dutch governments for by Bank Operational Policy (OP) 10.04 have also been active in funding GHG (September 1994). The document notes that mitigation studies and projects. global externalities are "normally identified in the Bank's sector work or in its environmental Global Overlays assessment process". Hence, Bank sector studies should now specify the global environ- Bank experience and resources warrants mental implications of proposed sectoral significant involvement in the rapidly evolving development plans. OP 10.04 requires that field of GHG mitigation work. For example, the "global externalities [ ... ] are considered in the Bank's economic and sector work deals with economic analysis when (a) payments related to many of the same sectors, e.g energy, forestry the project are made under an international and transport, that are analyzed in GHG agreement, or (b) project and project compo- mnitigation studies. Extending this work to nents are financed by the Global Environment cover climate change externalities offers the Facility. Otherwise, global externalities are prospect of capturing analytical economies of fully assessed (to the extent tools are available) scale and scope as well as enhancing consis- as part of the environment assessment process tency between sectoral development strategies and taken into account in project design and and global environmental protection. Such an selection". extension has been spearheaded by the Global Environment Division (ENVGC) under the Sector work is generally a precursor for identifi- "Global Overlay" program4. cation of projects. Overlays may be used for identification of both regular Bank and bilateral Similar to a graphic overlay, which attaches a lending and GEF investment operations. For new layer to an already existing surface, a the GEF and similarly motivated donors, Global global overlay adds a global dimension to the Overlays are an investment screening tool that sector studies that the Bank undertakes on a allow reliable identification of cost-effective regular basis for its client countries. An overlay mnitigation options or important investments in for the climate change area is likely to consist of climate change mitigation, in addition to two components: 4 Enviromrnent Department Papers  biodiversity conservation or the protection of Global Environmental Analysis, and Mexico international waters that are eligible for GEF Greenhouse Gas Assessment). An initial study funding. on the Mekong Power Integration Scheme has been completed, which may develop into a full To date, the World Bank has been actively Global Overlay. A Phase 1 Global Overlay for involved in the China GHG Emissions Control the Russia Oil Sector has also been completed. Study. In addition, three Global Overlays have A study on Greenhouse Gas emissions in the already been completed by the regions in urban transport sectors of Mexico City and coHlaboration with Bank clients (Argentina Santiago, examining how control cost curves for Carbon Sequestration, Ukraine Energy Options local pollution reduction are altered by the presence of a globally motivated agent (e.g., the Global Carbon Initiative [GCI]), is also nearing compietion. A study on GHG-friendly alterna- tives to 2-stroke engines in South Asia has just commenced. 1 Art. 4.6 grants economies in transition "a certain degree of flexibility" in meeting their requirements. 2 At this writing, five Bank client countries, Poland, Hungary, the Czech Republic, the Slovak Republic, Estonia, and Latvia, have presented their national communications to the interim secretariat. The Environment Department commissioned an external study to evaluate the consistency of Bank operations in the Czech Republic, Poland, and Hungary with the plans and measures outlined in the communications. The final report is available from ENVGC. 3 The expressions abatement and mitigation are often used interchangeably. Since mitigation is the term favored by the FCCC, mitigation studies is used here to denote efforts developed in relation to the FCCC process that are designed to inform national policymakers about different GHG reduction options. The cost imposed by implementing these options is termed abatement cost in line with the conventional use of the term in economic parlance. 4 Guidelines for Climate Change Global Overlays have now been developed, and are available from ENVGC. Overlays are not restricted to climate change externalities. Similar work has been developed for the biodiversity conservation area, and has been planned for international waters focal areas. Climate Change Series 5  Toward Integrating Climate Change Externalities in Bank Sector Work 6 Envirorunent Department Papers  Typology of GHG Abatement ost Studies Top-down vs. Bottom-up Intermediate and long-term top-down models are intertemporal general equilibrium models of GHG mitigation studies form a particular class various degrees of complexity. Global models of abatement cost studies for GHG emnissions such as the Second Generation Model reduction5. The latter come in different types, (Edmonds et al. 1992) or GREEN (Burniaux et al shapes and sizes. They are also designed to 1992) focus on the broad impacts that different serve various and often very different objectives. policy options would have on world regions For example, some studies are used to evaluate and have limited sectoral or regional detail. the effects of a carbon tax on various sectors National models such as Jorgenson/Wilcoxen and quite often the entire economy, while others (1992) and Goulder (1993) for the US or Blitzer attempt to evaluate the cost-effectiveness of et al. (1993) for Egypt generally offer much more different technologies on a sectoral level. GHG sectoral detail. For example, Jorgenson/ abatement cost studies differ with respect to the Wilcoxen predict impacts of carbon taxes on the geographical and sectoral scope they cover, the output of 35 industries and the supply prices of timescale, the level of disaggregation and the their products. Following from their objective to modeling techniques. Perhaps the most impor- ascertain the impact of climate policy choices, tant distinction is between two different profes- top-down models concentrate on capturing the sional and methodological approaches, which links and feedbacks between the energy sector are aptly called top-down and bottom-up6. and the rest of the economy, with less emphasis on the dynamics of the energy sector itself. The top-down approach comprises short-run macroeconomic and long-run general equilib- A study taking a somewhat different approach rium models. Both model types look at the was completed in 1995, examining the macro- overall impact of changes in climate policy on economic impacts of direct subsidy elimination the economy - particularly changes effected by in the United States. The Jorgenson-Wilcoxen- the introduction of a carbon tax. In these Slesnick model, developed by Dale Jorgenson models, energy is treated as an input for the Associates for the U.S. Environmental Protec- production of a representative good. Since tion Agency, and reported in Shelby, et al. production is assumed to be efficient, any (1995), applies a computable general equilib- restriction of energy consumption is viewed as rium model of the U.S. economy to a policy a cost. entailing the elimination of US$15.4 billion in direct energy subsidies (about 3% of total Important distinctions within the top-down energy expenditures), including subsidies to models relates to their time horizon (long run non-carbon energy sources. Carbon emissions vs. short run) and their geographical scope drop by 4% annually over the base case for the (global vs. national). Short-run models such as period 1990-2050. Removing the subsidies DRI for the USA (Brinner et al. 1991) or MDM contributes 30% of the total reduction in (Barker et al. 1993) usually focus on the transi- emissions required to achieve stabilization of tional effects of policy changes on the economy, emissions from the U.S. overall. If the revenues often based on a Keynesian framework. are recycled via a decrease in capital income Climate Change Series 7  Toward Integrating Climate Change Externalities in Bank Sector Work taxes, Gross National Product will improve by The bottom-up approach is based on engineer- 0.2% over the base case due to the favorable ing estimates of costs for individual measures impacts of increased investment on the or combinations of measures, primarily within economy. If recycling occurs through a reduc- the energy sector. These models are generally tion in marginal labor taxes, growth increases, based on the least-cost planning approach that but only gradually to 0.1%, as the opportunity integrates the traditional supply perspective of cost of consuming leisure increases, causing energy planning with demand side manage- workers to supply more labor and produce ment measures in a consistent framework. more output. Finally, a change in the average Consistent with their use for energy planning, tax rate only results in an increase in consump- bottom-up analyses are used to evaluate tion, and no increase in labor supplied, result- individual mitigation options in the energy ing in a -0.3% change in growth. The model is sector. not capable of analyzing the impacts of a change in the budget deficit as a result of the There are two subgroups within the engineer- decrease in subsidies, so no "worst case ing/ bottom-up approaches, i.e. (i) partial scenario" is available.6 models and (ii) full energy system representa- tions which can be subdivided in (a) integrated In summary, top-down models have some simulation models, and (b) optimization limitations when applied to modeling green- models, mostly based on linear programming. house gas abatement costs. First, most models In partial models, abatement options are do not have enough. technological detail to analyzed separately and usually ranked in model energy sector options very carefully. One technological cost curves in order of their cost- reason for this is that they tend to focus on the effectiveness. Such cost curves can provide a effects of price changes rather than sector- comparison of the direct costs associated with specific investments. Second, the assumption of the implementation of individual measures, but efficient production may lead to an overstate- such a separate analysis ignores ment of GHG abatement costs for the given interdependences between different options structure of the energy system. Inefficient use of that are likely in a comprehensive mitigation energy may be due to a combination of market, scenario. Integrated energy system models can policy and information failures. Often cited overcome this shortcoming. examples are (i) high transaction and informa- tion costs relative to energy saving, (ii) unfavor- Simulation models such as LEAP offer a large able institutional arrangements (e.g. landlord/ degree of flexibility in representing particular tenant) and (iii) regulatory imperfections. To system characteristics and they also represent the extent that inefficiencies persist, energy individual technologies and subsectors in efficiency improvements - and associated GHG much detail. The high degree of complexity has emission reductions - could be achieved at a its drawbacks, however. It demands thorough cost lower than the gain from efficiency im- knowledge of the model system by the indi- provements. This is known as a 'negative' cost vidual user and the consistency of the system abatement option, and results in a net gain to has to be monitored since there is no built-in society. optimization procedure. Linear programming optimization models minimize discounted cost Many top-down modelers acknowledge that over some period, subject to infrastructural and such opportunities may exist, but are generally policy constraints. Frequently used models less sanguine about their size. In addition, they include WASP, EFOM and MARKAL. Such argue that such efficiency improvements should models have been found useful for designing be included in the baseline as well as in any optimal supply systems, but they also have abatement scenario. To a large measure, then, certain disadvantages imposed by the math- the divergence in results between top-down and ematical framework. bottom-up studies can be traced to different assumptions about the baseline rather than differences in methodology7. 8 Environrment Department Papers  Typology of GHG Abatement Cost Studies Figure 2.1. Typology of Greenhouse Gas Abatement Cost Studies Top-Down Bottom-Up Short-run macro Long-run general Partial Modelling Integrated Sectoral models equilibrum modelsModel Models Global National Simulation Optimization Mtgtoni DRI (Brinner) SGM (Edmonda) Jorgenson/ WB/IUNDP UCCEE Wilcoxen China Examples MDM (Barker) GREEN Blitzer et al. LBU CSMT OGverays Objectives information on costs of large-scale information on cost-effectiveness of interventions and policies individual technological options/ investments There are general limitations of the bottom-up but in a wider sense also intangible costs approach as well, as has been repeatedly imposed by less than perfect substitution of pointed out. First, the feedbacks between the energy-consumning devices or underestimated energy system and the rest of the economy are time requirements for the implementation of generally ignored. Since the energy sector may individual measures. be of considerable significance for macroeco- nomic variables such as annual GDP growth, A Typology of GHG Abatement Cost inflation, or the trade balance, this shortcoming Studies is important, particularly for more aggressive mitigation scenarios. Second, hidden costs of Modeling approaches generally have to match abatement measures are often ignored. These are not only the more visible costs of informa- the obgri e of dysasgwelasits red tion campaigns and program administration, Climate Change Series 9  Toward Integrating Climate Change Externalities in Bank Sector Work quired. Figure 2.1. gives a broad overview of the top-down approach. Examples for mitigation types of studies (see also Grubb et al. 1993, studies at the national level are the WB/UNDP UNEP 1992 for more detailed typologies). study on China and work in the US country studies program on the basis of the Lawrence Figure 2.1 shows that the difference in method- Berkeley/CSMT methodology. Sectoral level ology and professional orientation between top- examples are global overlays (e.g., energy/ down and bottom-up approaches is matched by forestry) and the UCCEE studies, even though a difference in objectives. It also indicates that some examples in the latter also include non- mitigation studies combine a bottom-up ap- energy options. proach on a partial modeling basis or sectoral integration basis with a macroeconomic assessment component that is derived with a 5For a more comprehensive comparison of the two approaches see Grubb et al. 1993 and UNEP 1992. 6Shelby, M., R. Shackleton, M. Shealy, and A. Cristofaro. "The Climate Change Implications of Elimnating U.S. Energy (and Related) Subsidies." Paper presented at the Evaluating Energy Subsidies conference spon- sored by the U.S. EPA, Sept. 6-7, 1995, Washington, DC. 1995. See also: Jorgenson, Dale W. and Peter J. Wilcoxen. "Reducing US Carbon Dioxide Emissions: An Assessment of Different Instruments." Journal of Policy Modeling, Volume 15, pp.491-520, October-December 1993. Bohringer, Christoph, "Fossil fuel subsidies and environmental constraints: a general equilibrium analysis of German coal subsidies and carbon emission restrictions, Environmental and Resource Economics, 8:141-55, September 1996. Welsch, Heinz, "Recycling of carbon/energy taxes and the labor market: a general equilibrium analysis for the European Community", Environmental and Resource Economics, 8:141-55, September 1996. 7 The issue here is to what extent win-win options would be implemented in the absence of climate change concerns. If all profitable investnent options are realized, these win-win options would become part of the baseline. Further reductions would then have positive costs. This is also called an efficient baseline. Con- versely, if existing trends are simply projected into the future, the win-win potential in an abatement scenario often is considerable. This is called the business-as-usual baseline. 10 Environment Department Papers  3 2Review of Previous GHG Mitigation Studies in Developing Countries While there are quite a few GHG abatement (UNEP 1994b) and accompanied by a main studies both for the global and the national report that explains the methodology and level in developed countries, there is a paucity compares the results (UNEP 1994a). of such studies for developing countries and for economies in transition. Of this set, only a few Methodological framework are GHG mitigation studies - studies that analyze and evaluate individual mitigation The proposed UCCEE methodological frame- measures and GHG reduction opportunities. work consists of three elements: (i) a definition Probably the most extensive mitigation studies of key concepts, (u) a set of common quantita- are the UNEP Greenhouse Gas Abatement tive assumptions (international fuel prices, Costing Study Project that comprised 10 discount rates, reduction targets), and (iii) an countries and the study entitled "Issues and analysis procedure with essentially four steps: Options in Greenhouse Gas Control in China", 1) Definition of baseline or reference undertaken jointly by UNDP, the World Bank scenario; and two Chinese ministries. These two studies 2) Identification and ranking of mitigation are reviewed below, with a focus on their options; methodological framework. 3) Construction of two abatement scenarios; 4) Macroeconomic assessment and subse- The UNEP Greenhouse Gas quent political and social evaluation. Abatement Costing Study Project A graphical and more comprehensive represen- The UNEP GHG Abatement Costing Project that tation of the analysis structure is displayed in was guided by a team of researchers from the Figure 3.1. UNEP Collaborating Centre on Environment and Energy (UCCEE) at the RISO National This procedure combines features from bottom- Laboratories in Denmark is one of the pioneer- up and top-down approaches. Step (2) is ing efforts to undertake a thorough analysis of equivalent to the partial modeling version of the GHG mitigation options and their costs in bottom-up approach. It consists of separate developing countries. In its first two phases, evaluations of costs and emission reduction the project has focused mainly on CO2 emis- potentials for different mitigation options. The sions from the energy sector, but extensions to ensuing steps (3) and (4) correct some of the other sectors and gases are planned. drawbacks associated with that method. Phase Two encompasses 10 country studies, Step (3) sets up a consistent abatement scenario including developed countries such as Nether- that integrates the technological options lands, France, Denmark and developing outlined in (2), but accounts for possible countries such as Brazil, Venezuela, Egypt, interdependencies between these options. For SenegaL Zimbabwe, Thailand and India. The example, calculations of emission reductions results of these 10 studies are published from demand-side savings are calculated with separatelys, but summarized in one volume current structures of the energy supply system, but they may be inaccurate if they are part of a Clinate Change Series 11  Toward Integrating Climate Change Externalities in Bank Sector Work SOCIAL, POLITICAL AND DATA ANALYSIS MODEL ECONOMIC ASSESSMENT Macroeconomic forecast __________ I____forecast _ Reference scenaro for energy sector, _ agriculture, etc. | Technology and emission data Sector models Technical options tor Ranking of technical __l GHG abatement 4 options for GHG Multicriteria assessment and abatement judgement. Criteria: -complementary environmental effects GHG abatement -secondary economic effects / scenanos -technology preference -social and political considerations | | ~Macroeconomic Marocnoi V / ~~assessment !model Figure 3.1 Structure of applied abatement analysis procedure comprehensive GHG abatementprogramthat Common Results would imply changes in the energy supply system. There are two scenarios in the UCCEE study-a 20% reduction relative to the baseline in the Similarly, step (4) evaluates the feedback of short term (2005/10) and a 50% reduction in sectoral changes on the economy in a macroeco- the long run (2025/30). Figure 3.2 shows nomic assessment. The eventual GHG abate- marginal abatement costs for six out of the ten ment strategy also has to undergo a social and countries in the study. On the x-axis are CO2 political evaluation with multiple criteria, as emnission reductions in the abatement scenario commonly provided by political processes. relative to the baseline. On the y-axis are These extensions seek to avoid the shortcom- marginal abatement costs in the target year. ings of sectoral models with insufficient representation of the linkages to the general Figure 3.2. shows two similarities of the cost economy. curves for developing countries. First, there is a considerable potential of abatement measures A very similar procedure was suggested by with 'negative' costs - or costs that are lower Lawrence Berkeley Laboratory and the US than the gain from efficiency improvements - Country Studies Management Team (CSMT/ which ranges from around 7% in Brazil and LBL 1994). Senegal to 35% in Egypt. The only exception is 12 Environment Department Papers  Review of Previous GHG Mitigation Studies in Developing Countries Figure 3.2 Marginal abatement cost in the long-term target year. 300 Reduction costs in USD Zimbabwe 250 200 Thailand 150 U 100 50 -Venezue O~ ~ ~~~* -r . . . . .j= . . . i, -50 CO2 reduction in -_ _ * - Senegal Egypt percentage of baseline emission -100 0 5 10 15 20 25 30 35 40 45 50 55 % Reduction Venezuela, where the baseline scenario includes reductions are often fairly expensive, which all efficient investments so there is no possibility makes the marginal cost curve rise steeply of obtaining further gains in efficiency. again from that point on. Second, there some similarities in the shape of In comparison, a win-win reduction potential the curves. Roughly, a steep part for the most of 10-35% in developing countries is compa- cost-effective option(s) is followed by a flat part rable in magnitude to that estimated for representing about 20% reduction potential, developed countries, both within the UCCEE which is either slightly above or below zero project and outside of it. In a thorough review abatement cost per tonne of CO2. Further Climate Change Series 13  Toward Integrating Climate Change Externalities in Bank Sector Work Table 3.1. Main categories of CO2 reduction and related marginal abatement cost (MAC) in the participating developing countries for the long-term target. Country CO2 reduction option reduction MAC (%) (US$) Brazil Electricity savings (industry, services and residential) 7 -66 Solar uses in agriculture 1 22 Fuelwood and charcoal for afforestation programmes 21 24 Ethanol, bagasse and electricity generation from bagasse 19 29 Total CO2 reductions 48 Egypt Fuel switching in households 6 -21 Efficient industrial equipment and maintenance 10 -12 Transportation 2 -12 Heat recovery and new industrial processes 9 -8 New raw materials 5 -3 Efficient household appliances 5 -3 Electricity generation 8 -I Efficient stoves 7 2 Total CO2 reductions 52 Senegal Early hydropower implementation 0.1 -210 Agriculture intensification 5 -28 Energy conservation in industry 0.4 -4 Dissemination of improved stoves II 0 Improve carbonisation efficiency 13 1 LPG-charcoal substitution 15 2 Biomass from afforestation 6 3 Total CO2 reductions 50 Thailand Efficient air conditioners 2 -36 Electronic ballast 1 -27 Compact fluorescent lamps (service sector) 6 -14 Compact fluorescent lamps (residential sector) 2 -9 Nuclear electricity 18 20 Highly efficient gasoline cars 1 92 Total C02 reductions 30 Venezuela Reduced flaring and leakage of methane 7 3 Efficient boilers and kilns 10 10 Freight transport 1 13 Efficient electric motors in the industrial sector 2 17 Passenger transport 4 21 Electric sector 0.6 35 Other energy savings in the industrial sector 2 39 Efficient electrical appliances 0.4 52 Total CO2 reductions 27 Zimbabwe Efficient boilers 23 -9 Energy savings in the industrial sector 4 -2 Efficient motors and power factor correction 2 -2 Increased hydropower 5 5 14 Environrment Department Papers  Review of Previous GHG Mitigation Studies in Developing Countries of abatement cost models, both of the top-down options is likely to produce a smoother shape of and the bottom-up variety, Grubb et al. (1993) the marginal cost curve. This might be most estimated a reduction potential for developed relevant for the intermediate cost range that is countries of around 20% at very low or negative usually covered in less detail, while the ranking costs. of the most cost-effective options may not be affected. Still, the unevenness of the coverage of An issue slightly different from the overall cost technological options makes comparisons of an abatement scenario are the individual between countries even more uncertain. mitigation options that comprise these sce- narios. Table 3.1 summarizes the main C02 Modeling tools: Studies also differed with respect reduction options, their reduction potential and to the modeling techniques they employed. All the respective (average) marginal abatement studies used some sectoral integration model to cost. develop the abatement scenario, but as Table 3.2 shows, some teams use simulation models, Table 3.1 shows some similarities across the while others preferred optimization models. developing countries in the study. The least Some country teams developed their own expensive measures are usually energy end-use spreadsheet models, while others used interna- savings in households or industry. By compari- tionally available products such as MARKAL son, electricity supply options tend to be much or LEAP. more expensive and tend to appear only at the high end of long-term abatement efforts. Output: There are considerable differences between the studies in content and presentation Differences between studies of results. For example, a macroeconomic assessment as proposed in the methodology At the same time, there are significant differ- was undertaken only by the France and Egypt ences between the country studies, with respect country teams. While the Egypt study provided to their results and their assumptions and both macroeconomic assessments and the analysis structure. Some of these differences are abatement cost for individual measures, the unavoidable, such as time or resource con- French country study reports only overall costs straints, expectations about the penetration of from different levels of carbon taxes, without new technologies, or those distinctions intro- explicitly listing the abatement costs of the duced by differences in data availability and chosen mitigation options. The India study comprehensiveness. Some other, less inevitable also did not report abatement costs for indi- differences are listed below. vidual options and due to a limited range of options considered, the maximum reduction Range of options considered: The mitigation their model could achieve was below 30%. As a options considered extend from less than 10 in consequence, the India study was not used at the Brazil, Thailand and Senegal studies to 57 all and the France study was only marginally options in Egypt. Clearly, a larger range of useful for international comparison, despite its Table 3.2. Energy Sector Modeling Tools Used in UNEP Country Studies Simulation Optimization Brazil MARKAL (dynamic) Denmark BRUS Egypt L ENPEP France NEXUS India TEESE (static) Netherlands MARKAL (dynamic) Senegal LEAP Thailand own spreadsheet model Venezuela LEAP Zimbabwe own spreadsheet model Climate Change Series 15  Toward Integrating Climate Change Externalities in Bank Sector Work analytical sophistication. Similarly, the Thai- multiple baselines. For example, the three land and the Venezuela studies did not exhaust options (efficient, most-likely, business-as- all abatement options and, as in the case of the usual) could be developed separately. This India study, did not reach the 50 % reduction would provide a range for the baseline, with a objective. central estimate to be chosen by the research team. A somewhat simpler version would be to Baselines: The study teams varied considerably develop two baseline scenarios (high/low). in their assumptions about the changes that would occur in the energy system without an Independent of these options is the question of explicit GHG abatement strategy9. For example, whether the focus on the 'degree of efficiency' of the Egypt team assumed that an adjustment the baseline, i.e. to what extent profitable program with energy price increases of 200- energy-saving investments should be assumed 300% would be successfully implemented over to be taken up without an abatement policy, is a 5-10 year timeframe. As a response to these really adequate. As the WB/UNDP study on price changes, some efficiency improvements China study indicates, factors unrelated to would be undertaken and are integrated into investments in specific energy-saving technolo- the baseline, but a lot of the present inefficiency gies such as structural changes may be of is assumed to continue even in the long run, considerable importance for energy consump- which opens up room for profitable and energy- tion, particularly in rapidly growing economies. saving investments resulting from a GHG These aspects seemed to have received scant abatement strategy. In contrast, the Venezuela attention both from UCCEE and the country team assumed similar price changes as a result teams. of an adjustment program which the country has been undergoing, but the study assumes MacroeconomicAssessment. The UCCEE team that all profitable efficiency inprovements are acknowledges that their methodological design realized under the baseline. Most other coun- was over-ambitious regarding the linkage tries took a position somewhere in between. between bottom-up energy supply and top- down macroeconomic assessment. While all Limitations studies used macroeconomic projections to estimate energy demand, only France and Egypt The UCCEE report lists three issues for further had a formalized macro model they could use to methodological improvement, namely (i) the assess macroeconomic effects of the abatement development of baselines, (ii) the macro- scenario. They note that "a more practical and econornic assessment and its link to the bottom- helpful guideline for macroeconomic assess- up study, and (iii) the issue of implementation ment would have been advice on the assess- costs. ment of a few key macroeconomic indicators on the basis of available national economic Development of baselines. The UCCEE report lists statistics" (UNEP 1994a: 91). in some detail the above-mentioned differences in baseline development across the country Implementation Costs. The implementation cost studies. It is less clear what lessons the team issue is raised most often with respect to draws from these divergences and how they energy-saving programs in the household and could be avoided. industrial sectors, where suspected high implementation or other transaction costs may One possibility would be to align baseline prevent seemingly profitable projects from being development more closely through collabora- implemented. While the report flags the tion between the Coordination Center and the development of a "formalized procedure for the national teams. Even though there are some assessment of implementation costs and guidelines for baseline construction, the report barriers" (UNEP 1994a:124) as an area for emphasizes that the decision rested with the further research, it is not clear how they could national country teams. Closer coordination be measured on a sectoral basis. Similar issues may allow consistent assumptions behind are currently under study in the GEF's PRINCE baseline development. A second and perhaps program for the project level, and some indica- more promising option wouId be to develop tion could perhaps be taken from these results once they are available. 16 Environment Department Papers  Review of Previous GHG Mitigation Studies in Developing Countries The World Bank/UNDP Study on This study put particular emphasis on baseline China development. Future emissions are calculated from an 18-sector energy demand model that This study is a comprehensive two-year analy- allows fairly precise representation of expected sis titled: "Issues and Options in Greenhouse technological developments in each sector. For Gas Control in China". It was funded by the the High Efficiency scenario, energy coefficients GEF, administered by UNDP and executed by for sectoral production were taken from indus- the Industry and Energy Division in the China trialized countries to simulate a transition to and Mongolia Department at the World Bank in currently available technologies. a joint effort with the Environmental Protection Agency and the State Planning Commission in Steps (3) to (5) are based on a partial mode3ing China (National Environmental Protection framework. Abatement costs are calculated as Agency et al. 1994). The sumnmnary report came direct costs from the investment, i.e. investment out in November 1994. and operating costs, relative to the baseline. The evaluated measures for each sector are then Its objectives included an inventory of GHG aggregated for each sector. There is no inte- emissions and sinks, the estimation of future grated energy systems model, however, so that baseline emnissions until 2020, as well as the the alternative energy supply component identification and evaluation of emissions cannot easily complement the High Efficiency reduction options in four sectors: energy use, scenario. alternative energy supply, forestry and agricul- ture. Significantly, they also included the The case studies are geared more towards the identification of potential barriers to the imple- identification of possible implementation mentation of least-cost GHG reduction options barriers rather than the estimation of sectoral and the steps that governments and interna- abatement costs. Since the cases under consid- tional financing institutions could take to eration had been proposed by an expert panel overcome these barriers, including recommen- as the most promising ones and may therefore dations for future GEF funding in China. not be representative of the entire sector, the findings from the case studies were not Methodological framework sectorally aggregated. This affects the ranking of these measures: while individual mitigation In keeping with the various study objectives, the options can be ordered regarding their cost- methodology employed by the joint study team effectiveness, they cannot be aggregated into consisted of different - though partly inter- abatement cost curves since the quantitative connected - components: dimension is missing. 1) Estimation of two baseline scenarios (high/low growth); Baseline Scenarios 2) Estimation of High Efficiency Abatement 2)Escenarion of High Efficiency Abatement hIn order to provide a realistic estimate of the 3) Senar future path of GHG emnissions in China and to 3 agrat sectoral analyses for gyrstry, identify the key factors affecting this path, the agriculture, alternative energy supply; joint study team setup a GHG model that 4) Identification and evaluation of energy use consisted of three components: (i) a macro- reduction opions in various sectors; model to project final demand for goods and 5) Ranking of cost estimates from energy usemoetopjctfnldadfrgosad case studies and sectoral analyses in order services; (ii) an 18 sector input-output table to of cost-effectiveness, represent the structure of the economy; the table includes energy coefficients to determine the This methodology blends elements from top- energy inputs needed for the production of down and bottom-up approaches. Steps (1) and sectoral outputs; and (iii) an emission coeffi- (2) represent top-down procedures, while steps cient matrix to estimate the resulting emissions (3) to (5) are based on a bottom-up partial of GHGs as well as local pollutants. modeling framework. Climate Change Series 17  Toward Integrating Climate Change Externalities in Bank Sector Work The input-output coefficients were projected to intensity falls from 2.7 kilotonnes of coal the years 2000, 2010 and 2020 on the basis of equivalent per doIlar in 1990 to 0.9 ktce/$ in historical data for China and under consider- 2020, in comparable prices. The main reason ation of the I-0 coefficients for industrialized for this shift is structural change, as Figure 3.3. countries in earlier decades. The energy shows. coefficients are of particular importance for the resulting GHG emissions. Their projections According to this estimate, only 21% of the total were based on four factors: (a) expected eco- decline in energy intensity results from techni- nomic growth, (b) scale of new plants relative to cal efficiency gains. The remainder is the existing ones, (c) rate of adoption of new consequence of structural change. The most technologies, (d) increases in product diversity. important change is the shift in the product mix These findings were corroborated by case study within subsectors, which contributes 37% of the results. total decline. This change represents the improvement in product quality and the shift The studies produced two baseline scenarios. A into higher value added products, mainly in the High-Growth scenario expects real economic chemical, machinery, building materials and growth to be 8% p.a. for the period 1990-2020. light industry sectors. Saturation effects among This scenario is not unrealistic given China's residential consumers and the shift from direct historical record and the experience of the East coal use to electricity in this sector contribute Asian economies. The Slower-Growth scenario 12% of the total, and structural changes be- expects 6.6% p.a. real growth. Under the High- tween industry and other sectors as well as Growth scenario, GHG emissions will rise to among industrial subsectors contribute 7% and 2400 mtC equivalent by the year 2020, which is 9%, respectively. These numbers indicate the a threefold increase over 1990 levels. This enormous impacts that structural changes have baseline scenario already implies a consider- on energy intensity. Thus macroeconomic and able drop in China's energy intensity, which is other policies that affect some of the structural one of the highest in the world. In the baseline energy demand scenario, Chinese energy Figure 3.3. Sources of Projected Energy Intensity Decline in China, 1990-2020 37% 1 3 %~~~~11 9